fil \ M " l W F l MINI ) llslh «MN I I WI 01 MODIFICATION C”? THE MOJONNEE METHOD OF FAT DETERM‘NAWON BY SUBSTETUTION OF SOLVENTS Thesis {or the: Degree of M. S. MICHIGAN STATE COLLEGE Richard Emerson Marland i941 HWWIHIWWWWIWWI 293 00648 6553 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on Of baton due duo. DATE DUE DATE DUE DATE DUE MSU In An Affirmative Action/Equal Opportunity Instituion cWMI ‘..,.., ‘ '17-'57 5. . r: i . | -. 3 C a ! I 1 I I; This is to certify‘that the thesis entitled Modification of the Mojonnier Method of Fat Determination by Substitution of Solvents presented by Richard E.Marland has been accepted towards fulfihnent of the requirements for M.S. deywe n. Dairy.Manufaotures MODIFICATION OF THE MOJONNIER METHOD OF FAT DETERMINATION BY SUBSTITUTION OF SOLVENTS MODIFICATION OF THE MOJONNIBB.METHDD OF FAT DETERMINATION BY SUBSTITUTION OF SOLVENTS by Richard Emerson.Mar1and Mun-O A.TEESIS Submitted to the Graduate School of Michigan State College of Agriculture and Applied Science in partial fulfilment of the requirements for the degree of MASTER OF SCIENCE Department of Dairy Husbandry March, 1944 ITHESIS ACKNOWLEDGMENTS Sincere appreciation is eXpressed to the Dairy Industries Supply Association for the fellowship which permitted the under-- taking of these studies; to Doctor Earl weaver, Head of the De- partment of Dairy Hquandry, for making it possible to complete the work; to Doctor I. A. Gould for valued guidance in conduct- ing these investigations and in the preparation of the manuscript; and to Doctor W. D. Eaten for invaluable aid in making the sta- tistical interpretations. ’3{Jnlrf"I" ‘3 -xra, qjdjf, TABLE OF CONTENTS INTRODUCTION ......................................................... EH OF LITERATURE ................................................ Dry extraction procedures involving ether ........................ Wet extraction procedures involving ether ........................ Use of solvents other than ether ................................. ThGBIOjonniermethOd 0.0.0.000...0......I.OOOOOOOOOIOOOOPIOOOOO... SCOPEOF INSTIGATI()N .0.O...00.00.00.000...0.0.0....OOOOOOOOOOOOOOOOI EEPERIMENTAL PROCEDURE ............................................... RESULTS ...................................t.......................... Replacement of solvents in the Mojonnier method .................. Complete replacement by Skellysolve products ................. Miscibility of reagents ...................................... Partial replacement by Skellysolve products .................. Reduction in the amount of ethyl ether ........................... Possible influence of seasons .................................... Effect of various factors on the efficiency of the modified methOdS 0.0.000...0.0.0....IOOOOOOOOOOOODOOOOO...OOOODOOOOOOOOOOO DISCUSSION 0 U D O O O O O O O O O O O O O O O O O O O O Q 0 O O O O O O O O O I O O 0 O O O O O O O O O O O O O O O O 0 O O I O W Alq'D CONCLUSIONS C O O 9 O I O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O I O O O O O O 0 REFERENCES 0 O O O O O O O O O O O I 0 O O O O O O 9 O O O O O O O O O O O O O O I O O O I O O O O O D O D O O O O O O O O I O . APPEIDIX TABLES 0.0.0.0.....0....0.0000000000000000000...OOOOOOOOOODOO page 1 2 2 5 14 16 19 20 24 24 24 27 32 36 39 72 76 82 MODIFICATION OF THE MOJ'ONNIER METHOD OF FAT DETERMINATION BY SUBSTITUTION OF SOLVENTS INTRODUCTION For many years theiMojonnier method of determining the fat content of dairy products has been the standard laboratory control procedure for the dairy industry. This method, which is a modification of the Rose- Gottlieb technique only in regard to apparatus, utilizes ethyl and petro- leum others as solvents. The relatively high costs of these two ethereal agents, along with recent war-time shortages of ethyl ether, have created a need for an in- vestigation into the possibility of replacing these reagents with less ex- pensive and more easily obtainable materials. Although substitutes for these solvents have been suggested before, available data are limited and inconclusive. Furthermore, deSpite the fact that the Rose-Gottlieb (Mojonnier) method has been a standard for several years, little information is avail- able on the efficiency of extraction when variable quantities of solvents or other reagents are used. Similarly, a scarcity of specific data exists concerning certain of the fundamental concepts of the functions of the various reagents, especially in regard to the interactions resulting from the addition of the solvents. In view of the present shortage and economic considerations involv- ing the solvents, and because of the need for additional information per- taining to the principles of the Mojonnier method, investigations along these lines appear desirable. REVIEW OF LITERATURE The existing procedures for the gravimetric determinations of fat in dairy products may be classified into two groups: (a) dry extraction or (b) wet extraction. Dry extraction procedures involve a preliminary drying of the milk, or other product, to an almost anhydrous condition before proceed- ing with the extraction. In wet extraction procedures, the milk, or other product, is extracted without any preliminary dehydration process. 251 Extraction Procedures Involving Ether The Adamngethod: The Adams method as described by Richmond (59) and Croll (12) is purportedly the first gravimatric method developed for the analysis of milk for fat and was adapted by the Society of Public Analysts as a quasi-official method (59). The procedure is as follows: A.weighed quantity of milk is absorbed by blotting paper and the milk and paper dried to constant weight at 100° C. The dried paper coil is placed in a Soxhlet extractor and extracted with ether. The other is then evaporated and the fat dried to con- stant‘weight. Richmond (59) describes the work done by numerous investigators concern- ing the accuracy of this method. These workers agreed that the blotting papers contained ether soluble substances causing too high a value to be obtained. Richmond (59) overcame this difficulty by extracting a piece of blotting paper as a blank along with the milk samples. Later studies by Hale and Klykksen (24) showed the Adams test to give lower results than the Gottlieb, Werner-Schmid or Gerber tests. Eichloff (l4) objected to the use of the Adams method when.milk was abnormally acid as the results obtained were too high. Thomson (77) modified the Adams method by treating the milk saInple with pepsin and hydrochloric acid before drying. He found this treatment improved the results obtained when skim.milk and butter milk were analyzed. Timpe (78) recommended the addition of 25 per cent of sulphuric acid to the ether used to dehydrate it, thus eliminating the water soluble residue from the fat. RiJnefeld (62) compared the Soxhlet extraction method with the Bonnema. The results indicated that both.methods were equally good. The Bonnema method, as described by Croll (12) was based on principles similar to the Rose Gottlieb method, but included the addition of gum trage- canth (originally suggested by Busting (67)) to gelatinize the aqueous layer, making separation easier. Other workers (3, 6, 16, 24) have found the Adams method to be inaccurate in the analysis of certain dairy products, chiefly concentrated products, and have recommended other methods. Stg;ch.Methog: The Storch method of analysis as described by Richmond (59) was another early development. This method calls for the drying of milk on pumice at 100° C, grinding to a very fine powder, and extracting the fat by percolation using ether as an extraction.medium. The grinding and extraction were repeated, and all other washings were collected in a tared flask, dis- tilled, and the fat dried and weighed. Richmond (59) commends the accuracy of the Storch method. However, due to its length and the difficulty of grinding with pumice, he preposes the use of Kieselguhr for pumice and the use of the Soxhlet extractor. This worker states that plaster of Paris and kaolin also could be used satisfactorily as grinding substances. Other Methods:. The Somerset House method of Fat Extraction, which was developed by Bell of England (59) differs from that of Storch in that the milk is neutralized with N/ll strontia solution using phenolphthalein as an indi- cator. The neutralized product is dried on a hot plate with continual mixing. 'When the texture is such that it can be broken up, 20 ml. of’methylated ether are added and rubbed up with the dried powder. The other is decanted through a weighed filter. This extraction and maceration process is continued for eight extractions. When the ether extractions are finished, the same pro- cedure is followed using alcohol in place of the ether for two extractions. These are followed by eight more ether extractions. The filter is washed with other, the ether distilled off and the fat dried and weighed. The author claims that the slight error due to phenolphthalein and strontia in the resio due can be ignored. A somewhat similar method has been preposed by Mitchell and Alfend (47) for the analysis of butter. This method is commonly called the Sand- Gooch method and was found to be inferior to the Kohman technique (38). Richmond (59) describes a method devised by Babcock in which the milk is dried in a perforated metal cylinder, using specially prepared asbestos as the drying material. The extraction is carried out in the same container us- ing anhydrous ether. The results using this method agreed closely with those secured by using the Adams procedure. The refractometer has been used in estimating the fat in ethereal ex- tracts of milk by some workers. Such a method is the Wealney technique (12, 23). In this method, an aliquot portion is taken, and the fat calculated by use of the refractometer. The value obtained is used for calculation of the fat in the sample. Hale and Gregg (23) found the method to lack rapidity and accuracy. Several miscellaneous dry extraction methods have been reviewed by Richmond (59). All of these methods were proposed by some of the earlier 'workers (Wanklyn, Mbore, Piggott, Lieberman, Baynes, Harpmann, Ganutter, IDuclaux,'Wiley, Johnson, FernandezéKing, Ramps and Froidevaux). wanklyn extracted the fat from dried milk solids without any other 1nedium.being present except the ether. MOrre, Piggott and, later, Liebermann dehydrated the milk by adding to it anhydrous cOpper sulphate. The dry resi- due was extracted with petroleum ether. Baynes employed powdered glass as a grinding medium in his modification of the Storch method. IMarpmann used cot- ton wool, Ganutter used wood-fiber, Duclaux used sponge, Wiley and Johnson used asbestos as absorbing agents in their adaptations of the Adams method. Fer- nandezéKing and Hampe dried their milk samples by use of anhydrous sodium sul- phate in conjunction with dried Kaolin. The anhydrous product was extracted directly with 25 ml. of ether, and an aliquot portion taken for analysis. Froidevaux precipitated the protein and fat by use of a calcium phosphate solution containing acetic acid. The fat is extracted from this precipitate as in the Ritthausen process. Wet Extraction Procedures Involving Ether The earliest method using the wet extraction technique was originated by Horsley (29) and was known as the Meigs method (46). The procedure is as follows: ten ml. of milk are weighed and poured into a 100 ml. graduated cylinder, using 20 m1. of water to wash the weighing dish and dilute the sam- ple. Twenty ml. of ether are added and the cylinder shaken violently for five minutes. Twenty ml. of alcohol are then placed in the cylinder and the whole shaken again for five minutes. On standing, the ethereal layer sepa- rates and is drawn off with a pipette. Several five ml. portions of other are shaken in and removed to wash out any remaining fat. The ether is boiled off and the fat is weighed. Meigs states that this method gives distinctly higher results than the dry extraction methods, and a determination may be conducted in less than one hour. Croll (12) made numerous comparisons of Meigs' method with the Soxhlet method and found an average deviation of 0.023 per cent in 24 samples. He further proceeded to modify'Meigs' method by adapting a new type of mixing cylinder and other layer removing tube which resembles the presently used wash bottles. His results show that this new apparatus shortened the length of time necessary to complete the extraction without incurring any greater error than the original Neigs method. gobling Solution as Precipitant: Fehling's solution has been used to precipitate the proteids in milk with the precipitate trapping the fat. (Gudeman (21), Ritthausen (59), Broeman (8), Rieter (61), Sutton (74), and Szilasi (75)). Broeman (8) and Sutton (74) treated the precipitate with hydrochloric acid whereas Szilasi (75) treated the precipitate with neon. However, Gudeman (21), Reiter (61) and Ritthausen (59) extracted the precipi- tate directly using ether and alcohol. Gudeman (21) made extensive inveStiga- tions on the efficiency of this technique and concluded that in a high fat product, a preliminary extraction with petroleum ether should be conducted. Richmond (59) quotes workers as stating that the Ritthausen method cannot be used for the estimation of fat in sterilized or condensed milk, but Broeman (8), Sutton (74), and Szilasi (75) found their modifications to give excellent results when used on concentrated milks. Apid.Method§: Various acids, notably sulphuric, hydrochloric and acetic, have been used as agents to facilitate extraction. The Richmond and Rosier ‘method (60) was developed to provide a more rapid extraction method which sub- stituted petroleum ether for methylated ether because of the undesirable water soluble material often carried over by the latter. The procedure is as follows: Nine ml. concentrated sulphuric acid are placed in a 50 ml. tube having a con- striction at the 20 ml. mark. Ten grams of milk are added being careful to prevent mixing with the acid. This is followed by the addition of 0.9 ml. of amyl alcohol and the mixture is then thoroughly shaken. After cooling to about 25° C, 20 ml. of petroleum ether (b. p. 60° C.) are added and the sample again thoroughly shaken. When the layers separate, an additional shaking is made. The separation and shaking are repeated the second time and the other layer than decanted into a beaker containing 20 ml. water. After separation from the ago, the ethereal layer is removed to a tared flask where the fat is dried and weighed. Richmond explains the function of amyl alcohol as causing the surface energies of the surface of the fat globule to be broken down, allowing the ether to penetrate to the fat globule. Richmond and Rosier's data show good agreement with those secured with the Storch and Ritthausen methods, but according to Richmond (59) the good results are due to a compensation of errors, since the fat is affected by the sulphuric acid even as it is in the Gerber test from which this test was adaptedv Cochrane (10) developed a modification of the Babcock test similar to Richmond and Rosier's (60) modification of the Gerber test. Cochrane's method followed the usual Babcock method until the centrifuging step, at which point he added ether for dissolving the fat freed by the action of the acid. The ether layer is drawn off, and the fat determined. This method is not ex- tremely accurate because of the action of sulphuric acid on ether, as ex- plained by Richmond (59). A.test devised by Cobleigh (9) is very similar to Cochrane's, and is especially adapted to homogenized milk. - Richmond and.Musgrave (57), finding that all other available methods failed in the analysis of malted milk, developed their own method which is as follows: one gram.malted milk powder is stirred with 25 ml. of water till homogeneous, acidified with five ml. of one per cent acetic acid and then heated on steam bath until precipitation is complete. The precipitate is transferred to a Gooch crucible where it is extracted in a continuous extrac- tor’using petroleum ether as a solvent. The fat is determined by loss of weight in the crucible. They describe a similar method developed by Trillat and Sauton wherein acetone is utilized instead of petroleum ether for extrac- tion. Richmond and.Musgrave (57) found their*method to agree closely with Cochrane's (10) modified Babcock test, which they consider to be a valuable, accurate method. The Werner-Schmid method is described by Richmond (59) and'McLellan (44). The original method was applied to milk, and was adapted to cheese by Bondzynski, resulting in the Schmid-Bondzynski method (52, 72). Werner- Schmid's procedure is as follows: ten m1. of milk are placed in a graduated tube of 50 m1. capacity. Ten ml. of hydrochloric acid are added and the mixture boiled until dark brown. After the sample is cooled, 30 ml. of other are added and shaken. After standing, the volume of the ethereal solution is measured and ten ml. are removed with a pipette. The other is evaporated and the fat dried and weighed. woosnam (87) deve10ped an apparatus for use with this method which consisted of a boiling flask equipped with a ground glass top and a stopcock. A.graduated cylinder, also equipped with stopcocks for withdrawing the sample, was fitted into the flask after boiling, and the sol- vent added. When inverted, this apparatus could remove as much of the ethe- real layer as desired. Van Lennep and Ruys (79) modified the Werner~Schmid.method by substi- tuting trichlorethylene for the ether, and by filtering the aliquot before evaporation. They employed a correction factor as follows: Per cent'Eat . The authors found this to give perfect agreement 500 x .993 - Per cent Fat with the Rose-Gottlieb test. Grossfeld (20) modified the Werner-Schmid method for the analysis of cheese by substituting chloroform for the other as extraction agent and re- fluxing the mixture to attain better dissolving. He also used a correction factor secured by the following equation: 92 x drypresidue . Per cent fat. 32 - dry residue Sutton (74) modified the Werner-Schmid method as applied to condensed milk to eliminate the charring of the sugar. His method requires the precipi- tation of proteins by copper sulphate and washing this precipitate to free it of sugars. The hydrochloric acid is then added and the procedure continued as in the Werner-Schmid method. Weibull and Smetham, according to Scheringa (68) and Richmond (59), devised a special extractor for extracting the fat from the acid-treated milk. Scheringa (68) advised against dilution of the milk as it caused inconsistent results. Several workers have studied the Werner-Schmid method from a critical point of view with the objections being centered on the action of the solvent on the non-fat constituents. (Dyer (13), Fleming (16), Hale and Klykksen (24), Richmond (58), McLellan (44), and Ardenghi (2)). The Werner-Schmid method was criticized by Dyer (13) for not extracting the fat completely, but this error was balanced by the inclusion of non-fat substances in the residue. However, in the case of skim milk and condensed products the error was magnified, giving high results. Fleming (16) also criticized the Werner-Schmid.method because of the impure fat residue. Hale and Klykksen (24) in a series of extensive comparisons of methods of analysis found that the Werner-Schmid test gave higher results than the Adams or’Rose-Gottlieb methods. Richmond (58) considers the Werner—Schmid test as an accurate, but not particularly rapid method. McLellan (44) conducted a comparison of methods of analysis for fat in dried milks. He found the Werner-Schmid method gave results which were superior to other methods tried. ‘His work includes a description of the 'Wernerchhmid method with data showing its superiority in this type of anal- ysis. 10 Ardenghi (2) after a review of the current standard methods for fat determination decides that the Werner-Schmid is most suitable for laboratory control work. Richmond (59) discusses additional modifications of the Werner- Schmid.method including weighing of the sample and withdrawal of the aliquot for determination. The Schmid-Bondzynski method was adopted as the official method for analysis of cheese by the Association of Official Agricultural Chemists (31). Another acid.method making use of acetic acid for dissolving the pro- tein was devised by Harding and.Parkin (25) for the analysis of evaporated milk and milk powders. The method consists of treating a small sample with 25 per cent acetic acid to dissolve protein material, then shaking it with alcohol and carbon tetrachloride to suspend the fat and finally, extracting the fat with petroleum ether. Three extractions are used, and the fat is dried and weighed. A.Nessler jar with a bloweoff tube similar to Croll's (12) is employed in removing the ether. 2 Data presented by Harding and Perkin (25) average 0.43 per cent higher than those secured by the Rose-Gottlieb method and average 0.166 per cent higher than results secured by Richmond's modification of the Rose-Gottlieb (58). On trials using purified milk fat, their extractions were 99.9 per cent efficient. Harding and Perkin (26) further adapted the method to ice cream. Their data average 0.11 per cent higher than results by the Rose- Gottlieb method. Hortvett (82) recommended that the.Association of Official Agricultural Chemists tentatively adOpt the Harding Parkin.method for analysis of ice cream. alkali Methods: An entirely different principle is employed in an alkali method preposed by Rose (65). In this method the sample is shaken with ammonia in contrast to the acids of previous techniques. His procedure is as 11 follows: About 20 g. of milk are mixed with two ml. of ammonia, 45 ml. of alcohol and 120 ml. of mixed ethyl and petroleum ethers are added, and the solutions mixed by shaking in a 230 ml. graduated burette. The ethereal volume is measured, 25 ml. taken, and the fat in this amount is determined. He found that it was necessary to add 0.015 per cent to account for the fat remaining in the aqueous layer. Schreib (69) describes extensive comparisons which show the Rose method to be as accurate as any existing method. The Rose method was studied and modified by several workers, with important modifications being preposed by Pepp (53) and Gottlieb (18). Adaptations of the method were applied to all types of dairy products. Popp (55) modified the Rose method by weighing the milk in accurately, and by changing the extraction tube. He found no saponification of the fat to be caused by the ammonia. Gottlieb (18) further modified the Rose method by changing the quan- tities of reagents used. This modification became known as the Rose-Gottlieb method. The revised procedure was as follows: ten g. of milk are weighed into a cylinder graduated to 0.5 ml. The sample is then shaken, in turn, with one ml. of 10 per cent ammonia, 10 ml. of ethyl alcohol, 25 ml. of ethyl ether, and 25 ml. of petroleum ether. The mixture is allowed to stand for six hours to complete separation of the aqueous and ethereal fractions. The upper layer is then completely removed, and the fat obtained from it in the usual manner. Gottlieb states that making the volumes of milk and alcohol equal causes a ,more rapid separation of the ethereal and aqueous layers because of the in- creased density of the aqueous layer. Gottlieb (18) later presented a more rapid adaptation which utilized pipettes for placing the sample in the tube, for withdrawing the ethereal 12 layer from the tube. He states that this increase in rapidity is gained without loss of accuracy, and his analyses show the fat to be pure, and not contaminated with other ether soluble substances. In early work, McLellan (44) found the Rose-Gottlieb method to be in- ferior to the Werner-Schmid method in the analysis of dried milks. Fleming (16), on the other hand, criticized the Werner-Schmid method as giving an im- pure fat residue, whereas the Rose-Gottlieb method gave a pure fat, but was a more tedious procedure. Bigelow and Fitzgerald (6) found the Rose-Gottlieb technique to be the only method which gave good results on concentrated milk products, when compared to dry extraction.methods and rapid volumetric methods. Many early workers used the Rose-Gottlieb test as a control when con- ducting comparisons of other methods. (Hale and Klykksen (24), Netbohm and Angerhausen (49), Rammstedt (55), Windisch (85), and Olson (50)). Weibull (as), Kuhn (40), Popp and Siegfeld (54), and Gottlieb (18) showed the Rose-Gottlieb test to be applicable to skimmed milk, buttermilk, cream, butter and cheese. Eichloff (l4) and Eichloff and Grimmer (15) devised adaptations of the Rose-Gottlieb test for use in the analyses of cream, butter, cheese, and dried milks. Hesse (28) deve10ped an adaptation for use in butter analysis. Richmond (58) outlines an adaptation of the Rose-Gottlieb method for analysis of dried milk in which the amount of all reagents used were reduced 50 per cent. A.micro-analytical method was deve10ped by Kurz (41) which employed all the reagents used in the Rose-Gottlieb method and in similar prOportions. Improved extraction tubes were deve10ped by Rohrig (63) and Rieter (61), and Adorjan (1) deveIOped a burette for drawing off the ethereal layer. This burette, graduated in 0.1 ml., was equipped with a ground glass stOpcock, and was attached to the extraction tube in such a way that the junction of the 13 ethereal layer and the aqueous layer was level with it. Biesterfeld and Evanson (5) modified the Rose-Gottlieb method by mak- ing use of acid in conjunction with the ammonia. They found this liberated a slightly greater amount of fat in the case of melted milk. Roop (64) re- placed all the ammonium hydroxide with one m1. of 1:1 sulphuric acid. 'He agreed with Biesterfeld and Evanson that this use of acid gave a true picture of the fat content of the milk. The Rose-Gottlieb method and its adaptations for various dairy prod- ucts received much attention from the referees of the Association of Official Agricultural Chemists. In 1906, after comparisons of all available methods for the analysis of cheese, Olson (50) states that the true picture is shown by the Gottlieb method. Also in 1906 Well (as) reports that the Gottlieb re- sults were higher, but believed more accurate than other methods, and recom- mends that the Gottlieb test be made provisional by the Association. In 1915, Hortvett (30) recommended that the Association adOpt provisionally the con- tinuous method of Paul. (The method of Paul is described by Bigelow and Fitz- gerald (6) as being a dry extraction method very similar to the Adams method.) The procedure gave higher results than the Rose-Gottlieb test. However, in 1917, this investigator recommended the adaption of the Rose-Gottlieb method for'use on milk, sweetened and unsweetened condensed milk, and later (32) recommended the adOption of the Rose-Gottlieb method for the analysis of ice cream and malted milk. Keister (36) reports that the acid medium modification of'RDOp (64) gave good results if the procedure was followed carefully. Rupp and Maller (66) further modified the Rose-Gottlieb method by the addition of 0.4 g. of gum tragacanth after the ethereal additions, a treatment based on earlier findings of Rusting (8). This procedure gelatinized the tnyleous layer, causing a more rapid and clearer separation of the two layers. 14 Oven (51) found that less tragacanth could be used with equal benefits, and studied the method in comparison to others. His claim is that the Rupp- Mfiller~modification is just as accurate as the Gottlieb test, and is more rapid. IKrOpat (39) applied the Rupp-Muller modification to cheese, cream and butter analysis. Koenig (37) found that in butter analysis, more trage- canth was necessary, and he advised three washings with petroleum ether after decantation. The Use of Solyents other than Ether varioussolvents-have been suggested to replace ether in fat extractiOn methods. Certain workers (13, 16, 60) have shown that errors are encountered when petroleum.ether or methyl ether were used as solvents. Broderick and Pittard (7) have also found that petroleum ether dissolves the fatty acids with greater difficulty than ethyl ether. The use of carbon tetrachloride as a solvent in fat extraction has been the object of much study. velrath (81) mentions that carbon tetrachloride is superior to ether, benzene, and carbon disulphide as a solvent, due to the fact that it carries no water on condensation. werkers for the Greishm Elec- trochemical works (19) found that carbon tetrachloride was superior to gaso- line as a solvent, and was less dangerous to use. Rammstedt (55) ran exten- sive comparisons of the existing ether extraction and carbon tetrachloride methods, and found the results obtained by using carbon tetrachloride were about four per cent higher than by the ether extraction method. For this reason he did not recommend the use of carbon tetrachloride. Carbon tetrachloride was used as solvent in Leithe's pycnometer method for the determination of fat in dairy products (43),-and in Hackman's method for determination of fat in cheese by measurement of the Specific gravity of 'the solvent before and after treatment of the cheese (22). 15 The Consortium far Electrochemische Industrie (11) compared the chlorine derivatives of ethane and ethylene with benzene, carbon tetrachlo- ride and carbon disulphide. A preference was indicated for the derivatives because of their freedom from toxic effects and explosion. The belief was‘ expressed that dichlorethylene is an efficient solvent and can be used as a substitute for ether. Taufel and Standigl (76) conducted a comparison of trichlorethylene, benzene, carbon disulphide, acetone, ethyl ether, and chlorofonm. They state that the ether and benzene gave clear, light-colored extracts whereas the other solvents yielded darker extracts. The trichlorethylene extracted the fat slowly, but gave the greatest yield in the end. The composition of_the fatty residues varied only slightly, the acetone having extracted a greater amount of phosphatides. The authors believe this increase in phosphatide content is due to a higher percentage of lecithin in the acetone extract. Acetone was used by Beida (4)and obtained French patent rights on his use of this solvent. Other fat solvents have been used to replace ether by still more work- ers. Harris (27) describes a method in which ortho-dichloro benzene is used as the solvent, and the fat percentage found by the change in Specific gravity of the solvent. DeWaal (82) reports that the petroleum ether in the Rose-Gottlieb test can be replaced by gasoline without any loss of accuracy in the method. Gaso- line has also been used along with petroleum other by Kohman for butter analy- sis (38) and.Mandal substituted benzene for petroleum other as a solvent with- out loss of efficiency. Carbon disulphide was preferred to ethyl and petro- leum others by Visern and Guillot (80). Jaffe (34) describes monoethyl ether of ethylene glycoll as being a 'very good solvent, but the work was not conducted on butterfat. 16 various alcohols have also received attention in fat extraction studies. Gerber's (l7) patented the process of extracting fat by use of ketones as methylethyl ketone alone or combined with alcohols of the fatty series - e.g. amyl, butyl, or methyl alcohol. Lampestrasse (42) received patent rights on the use of a butyl alcohol fat-dissolving agent in combi- nation with a fat-soluble coloring agent. Wendler (84) used butyl alcohol with sodium salycilate, and Sichler (7) used isobutyl alcohol with a fat- soluble coloring agent in addition to alkali salts of pyrotartaric acid. The Moionnier Methgfid The Mojonnier method of fat extraction (48) has in recent years been accepted by the dairy industry as being the standard method for lab- oratory control work. The extraction principles are exactly the same as in the original Rose-Gottlieb method, the changes being made only in the apparatus used. Hortvett (33) states, "The Mbjonnier method is nothing more or less than the Rose-Gottlieb method. The term relates specifically to an apparatus or machine designed sepecially for the purpose of shorten- ing the time required for carrying out a determination." The Mojonnioraap- paratus consists of a compact piece of equipment containing an analytical balance, vacuum drying ovens, hot plates, dessicators, a centrifuge, re- agent burettes, extraction flasks and tared evaporating pans. 'Mojonnier and Troy (48) and Supplee and Bellis (73) made exhaustive studies as to the ac- curacy of this method as compared with the official Rose-Gottlieb method. Their data show rather conclusively that there is no significant difference in the results obtained by using either method, but the Mojonnier method re- quired only about one-sixth to one-seventh the time necessary to conduct the Rose-Gottlieb test. 17 Mojonnier and Troy (48) also present data showing the effect of vary- ing the amounts of reagents. A portion of their results are shown in table I. Table I. The Effect of varying Amounts of Solvents on the Efficiency of Fat Extraction Product anntity of Solvent 4%Fat Remarks Normal Amount 3.13 Normal Milk 50% Reduction 3.0? Cell formed on adding ether 50% Increase 3.12 No advantage Normal Amount 8.09 Normal Evaporated Milk 40% Reduction 8.07 Extraction incomplete 20% Increase 8.09 No advantage Normal Amount 13.51 Normal Ice Cremn Mix 40% Reduction 13.50 Extraction incomplete 20% Increase 13.53 No advantage Table I shows that reducing the quantities of the solvents resulted in incomplete extraction of the fat, whereas increasing the amounts failed to improve the extraction. These investigators further show that in some cases large amounts of ethyl ether lowered the dividing line, whereas large amounts of petroleum ether and reduced quantities of ethyl ether had the Opposite effect. In summarizing their work, Mojonnier and Troy (48) state that variations in the quantities of water and alcohol have a greater effect on the accuracy of the method than do variations in the other reagents. Since the development of the Mojonnier tester only one modification has been proposed, - that being by Johnson (35). This modification replaces 18 both the ethyl and petroleum ethers with isopropyl ether. The method in- volves the dilution of the sample with water, the use of ammonia and alco- hol much as in the original method, but replaces the 50 ml. of solvents by 25 ml. of iSOprOpyl ether. The data average 0.02 per cent higher on all products than the Mojonnier test. In many products the tests were in per- fect agreement, but in evaporated milk, the Mojonnier test is nearly 0.10 per cent lower than the modification. Johnson (35) states that is0pr0pyl other is advantageous over ethyl ether because of its higher boiling point which reduces evaporation losses. He states further that isoprOpyl ether is free of non-volatile residue, is an excellent solvent and does not ab- sorb water, or carry over milk solids-not-fat. The approximate cost is~ stated to be about 60¢ per gallon, with no drum deposit required. It has the disadvantages of forming peroxides more readily than ethyl and petro- leum ethers, and also tends to form an emulsion with water, necessitating the use of 50 per cent alcohol in raising the volume of the lower layer after the final extraction. Johnson concludes that the replacement of ethyl ether and petroleum ether by isopropyl ether results in a method which is more rapid, more con- venient and cheaper than the standard Mojonnier method. 19 SCOPE OF INVESTIGATION This study was conducted with the following intentions: 1. 2. 3. To investigate the possibility of replacing the ethyl and petroleum ethers, either wholly or in part, by some less expensive and more easily obtainable solvent. The greater part of the investigation was devoted to this phase of the study. To ascertain the efficiency of fat extraction as influenced by variations in the quantity of the reagents. The chief consideration in this connection pertains to the use of re- duced amounts of solvents. To gain further knowledge of the interactions of the several reagents used in the Mojonnier procedure. 20 MERIMENTAL PROCEDURE general: Determinations in this study were made using whole milk, homogenized milk, ice cream mix, evaporated milk, skimmilk, and churned buttermilk. The milk, skimmilk, and buttermilk were obtained from the College creamery. The homogenized milk was processed at 2,500 pounds using a piston- type viscolizer. Ice cream samples were obtained partly from commercial dairies and partly from the College creamery. Comercial samples of evap- orated milk were obtained for the trials devoted to their analyses. All products were warmed to room temperature before the samples were weighed and each sample was well mixed by pouring from one container to an- other immediately prior to weighing. When partial churning was observed, or the sample was otherwise found to be non-homogeneous, it was discarded. Control procedure: The Mojonnier method, conducted under carefully controlled conditions and in accordance with the directions of Mojonnier and Troy (48), served as the control method. In the case of milk, homogenized milk, skimmilk and butter milk, the general procedure is as follows: An accurately weighed ten gram sample is treated successively with 1.5 ml. ammonium hydroxide, 10 ml. ethyl alcohol, 25 m1. ethyl ether, and 25 ml. petroleum other, being shaken for 30 seconds after each of the additions. The mixture is then centrifuged and the other layer decanted into tared pans. A second extraction is made consisting of 5 ml. ethyl alcohol, 15 ml. ethyl ether and 15 m1. petroleum ether. After complete decantation of the second ethereal layer, the other is volatilized, the fat dried at 155° c. under 20 inches of vacuum for 5 minutes, cooled in a dessicator to room temperature, and the fat determined by weighing. For the analysis of evaporated milk and ice cream mix the procedure was varied slightly by using a 5 gram sample, and diluting this with 4 ml. of 21 water for evaporated.milk and 5 m1. of water for ice cream mix. These two products also received 25 ml. of each other on the second extraction instead of the 15.m1. used for the analysis of the other products. Ezgcedure for the replacement of solyents: For the purpose of substi- tution for the ethyl and petroleum ethers, two petroleum.naphthas were ob- tained from the Skelly Oil Company, Lyman, Oklahoma. These naphthas were Skellysolve "F" and Skellysolve "A”. The first of these, Skellysolve "F”, is described by the Skelly Oil Company as being essentially petroleum ether, having a boiling range of 30° - 60° C, a non-volatile residue of 0.0015 per cent, and A. P. I. gravity of 90° at 15.70 c, and a Reid vapor pressure of 14.2 pounds at 57.90 C. The actual boiling point was determined at 39.0 - 489 C. This material was used experi- mentally to replace the petroleum ether which was found to have a boiling point of 48-570 C. The second of these naphthas, Skellysolve EA", was used to replace the ethyl ether in pr0portions of 25 per cent, 40 per cent, 50 per cent and 100 per cent respectively. This reagent is described by the Skelly Oil Company as being essentially normal pentane with a boiling range of 28.50 - 58° C, an A. P. I. gravity of 91.7 at 15.70 C, a Reid vapor pressure of 14.8 pounds at 37.90 C, and a non-volatile residue of 0.0015 per cent. The actual boiling point of this material was found to be 33.0 ~ 35.5° C. These solvents were used in the same amounts and manner as the materi- als they replaced in the standard Mojonnier procedure. Procedure for reducing the quantity of ethyl ether: In order to find the efficiency of Skellysolve “A" when used in the ethyl ether mixtures, two experiments were conducted in which reduced amounts of ethyl ether alone were used as the first solvent. In the first of these trials, 50 per cent of the normal amount of ether was used, and in the second series 75 per cent of the 22 normal amount was used. Both of these series were conducted in two sets of experiments, one set utilizing petroleum ether as the second solvent, the other set using Skellysolve ”AP as the second solvent. Except for necessary reductions in the quantity of alcohol used in the second extraction, no other changes were.made in the original method. Ergcedure for studying the interactions of the various reagents: Pre- liminary trials in which the ethyl ether was replaced, either partially or entirely, by Skellysolve showed a distinct difference in the volume of the lower layer of liquid in the extraction flask when compared with the control procedure. It was found that a smaller quantity of alcohol, or none at all, could be added in the second extraction because of the high dividing line which prevented efficient decantation of the ether layer. It was also ob- served that in all trials in which ethyl ether was replaced by Skellysolve, there appeared a flocculant yellowish layer of fat-like appearance at the Junction of the two layers in the flask. Because of these observations, an experiment was conducted for the purpose of determining (a) the miscibilities of each of the reagents in various combinations and concentrations, and (b) the effect of each reagent on\the volume and appearance of the lower layer. Several 100 ml. cylinders, graduated to 1.0 ml. were equipped with tightly fitting steppers, and into certain of these were placed the reagents used in the Skellysolve method. Control cylinders contained the Mojonnier reagents. Shaking treatment was administered as in other trials. Observa- tions were made on the variation in the volume of the lower layer, as affected . by changes in the concentrations of the various reagents. The color and ap- pearance of the lower layer was observed, and changes in the flocculant, fat- like mass, occurring at the Junction of the two layers, were noted. 23 Statistical treatment of the data; The data in these experiments were subjected to a statistical analysis wherever possible. The standard error, which.measures the dispersion of the cases about the mean, Was de- termined for each series of eXperhments. As a note of explanation concerning the standard error, it should be pointed out that approximately 67 per cent of all cases will fall within the range of one standard error of the mean; therefore, a small standard error indicates a narrow range of disPersion. In data such as these, where an ex- perimental method is compared to a standard control method, a small standard error indicates that the difference between the methods is constant, and a shmilar difference would be obtained in 67 per cent of the cases if the work were to be repeated. In contrast, a large standard error indicates a wider range of differences as obtained by comparison of the two methods. The significance of the average difference between the two methods was found by use of the following formula: Actual difference between the methods - zero Significance = Standard error of the difference between the methods* The ”significance" thus found is compared to a tabular value which correSponds to the number of trials on which the observation was based. This tabular value is called ”t”. If the ”significance" is greater than the corresponding ”t", the difference between the methods is significantly different from zero, and it can be said that the difference obtained is not due to error in obtain— ing samples, but due to the methods themselves. Conversely, a value smaller than the corresponding ”t” means that the methods are not significantly differ- ont. For the purpose of estimating the actual amount of fat which is pres- *Baten, w. D. Mathematical Statistics, p. 220, John Wiley & Sons, New York. 24 ent when only the modified procedure is known, the regression curve was used. Through this procedure, the Mojonnier value may be calculated from the value obtained by the modified method by use of the equation y = a-+ bx where y represents the Mojonnier control value, 1 represents the value as determined by the modified procedure and a and b are constant values which have been de- termined for each.method by use of the regression curve. The curve is constructed by plotting a graph of the results of a series of trials using both the MOjonnier and a modified procedure. The Mojonnier value is plotted along the ordinate and the values obtained by the modified method are plotted along the abscissa. A straight line may be drawn through the points thus determined which will intercept the abscissa near the origin. The values a and b are found from this line, a being the slepe of the line and b being the point of intersection with the y axis. RESULTS Replacement of Solvents in the Mojonnier Method gomplete replacement by Skellysolvegproducts: 1. :Substitution offi§kelly- gglve "F” for petroleum ether. Results obtained by complete replacement of the petroleum ether of the mojonnier procedure by the same quantity of Skellysolve ”F” are pre- sented in Table 2. These results were obtained on samples secured during January, February and July. These data show excellent agreement between the two methods and between their reapective duplicate determinations. In general, the Mojonnier control procedure extracted slightly'more fat than the modified experimental method, although in the four trials conducted on evaporated milk the modified method gave somewhat higher values than the Mojonnier procedure. However, the differ- 25 Table 2. Total Replacement of Petroleum Ether by Skellysolve ”F" Using Ethyl Ether as the First Solvent* Differences Product No.of Fat thaiged Between Methods Between Duplicates Trials Mo onnier Modified Moignnier Modified (a) 1%) (%) (%) (£3’ (%0 Milk 14 3.5484 3.5440 0.0044110.0020 0.0090 0.0109 Homo. Milk 15 3.9889 3.9838 0.0051110.0020 0.0105 0.0077 Evap. Milk 4 7.8927 7.9027 0.0300 - 0.0300 0.0088 Ice cream. 9 9.5948 9.5807 0.0141110.0122 0.0191 0.0243 ibomplete data found in Appendix Table II.- ence in extraction efficiency between these two methods is well within the range of experimental error and is statistically insignificant. There is some indication that considerable more laxity in the shaking of the samples in the extraction is allowable in the case.of petroleum other than when Skellysolve is used with:gffecting the results. In several trials in which the shaking was limited to a SIOW'mixing rather than the prescribed vigorous horizontal shaking, the Skellysolve gave results which were consid- erably lower than the Mojonnier control method, and triplicate samples showed poor agreement. The detailed presentation of these data may be found in Table I of the appendix. ‘ 2. gubstitution of Skellysolvg ”A" for ethyl ether. Effect of reduc- igg the Quantity of alcohol, In trials in which the ethyl ether was replaced by the same quantity of Skellysolve "A” it was found necessary to use reduced quantities of alcohol in the second extraction, since the addition of the usual five milliliters increased the volume of the lower layer to such an extent that efficient de- cantation of the ethereal layer was impossible. To find the effect of this reduction in the quantity of alcohol on the extraction efficiency, trials were conducted in which 2.5 ml. of alcohol were used in the second extraction of 26 the Mojonnier method, instead of the standard five milliliter quantity. No other change in the procedure was made. The results of these studies are shown in Table 5. Table 3. The effect of Reducing the Quantity of Alcohol USed in the Second- Extraction of the Mbjonnier‘Method* Amount of No. of Differences Alcohol Trials Fat Obtained Between.Methods Betwe n cates m1. ' (i0 (75 (ii 2.5 5 4.0738 0.0083 0.0227 melete data found in Appendix Table Iv. These data indicate that no appreciable extraction efficiency is lost by reducing the quantity of alcohol used in the second extraction of the Mojonnier method. Efficiency of Skellysolve ”A" when used in place of ethyl eth_r. Trials were conducted in which the ethyl ether was replaced by the same amount of Skellysolve ”A”. These trials were conducted on samples ob- tained during the months of February and March, and the data are presented . in Table 4. In the analysis of milk, the modified procedure extracted 0.22181: 0.0094 per cent less fat than the Mojonnier control method, a statistically significant difference. It was also found that wide variations occurred be- tween triplicate determinations when this modified procedure was used. In three trials conducted on ice cream.mix, the modified method was found to extract an average of only 0.0665 per cent fat as contrasted to the 12.4217 per cent average obtained by the MOjonnier method. It may be concluded from.these data that this modified procedure not only extracted much less fat than the control procedure, but resulted in poor agreement between triplicate determinations. For these reasons, such a modi- fication is unsatisfactory for laboratory control work. 27 Table 4. Total Replacement of Ethyl Ether by Skellysolve "AF'Using Petroleum Ether as the Second Solvent* _Differences Product No. of Fat Obtained Between Methods Between Triplicates Trials Mo onnier Modified 1M0 onnier Mod‘f ed £95 5%) lib i9} 1%; Milk . 31 4.4913 4.2695 0.2218 * .0094 0.0125 0.0592 Ice Cream 3 12.4212 0.0665 11.7576 0.0125 0.0911 IBeta found in Appendix Table III. Miscibility of zeagents; As noted in the previous trials involving replacement of the ethyl ether with Skellysolve ”AP, only a small amount of alcohol could be added in the second extraction without resulting in a high dividing line. It was further noticed that in these trials a yellow, flocculent fat-like layer ap- peared at the junction of the two liquids in the extraction flask. The volume of this colored layer was reduced by the second extraction but did not entirely disappear. The variation in volume of the lower layer encountered when using Skellysolve was believed to be due to the immiscibility of alcohol and Skellysolve "A”. Accordingly, trials were conducted to find the extent of this immiscibility, and the effect of other Mojonnier reagents upon it. The results of preliminary determinations showed that: a) Ethyl a1- cohol and ethyl other are perfectly miscible in Mojonnier prOportions. b) Ethyl ether and Skellysolve ”A” are perfectly miscible in.Mo10nnier propor- tions. c) Ethyl alcohol and Skellysolve ”AF are not miscible in.Mojonnier prOportions. d) The addition of ethyl ether to a Skellysolve-ethyl alcohol mixture will increase the miscibility of the system. Trials were conducted for the purpose of determining the amounts of ethyl other required to bring about miscibility in a three-component system of alcohol, Skellysolve "A” and ethyl ether, under conditions of varying 28 quantities of ethyl alcohol and Skellysolve ”A”. The results of these trials are shown in Table 5. Table 5. Effect of Ethyl Ether on the Miscibility of an Ethyl Alcohol- Skellysolve ”A" Mixture when the Quantities of Alcohol and Skellysolve Are varied Series Ethyl Ether Required to N9, Ethyl Alcohol Skellysolve ”A” Bying about Miscibility (m1 ) (ml ) (ml ) (%) I. 10 20 4 13.3 15 20 4 11.4 15 35 6 12.0 15 40 10 18.1 15 55 16 22.8 II. 10 ' 20 4 13.3 15 20 4 11.4 15 35 8 16.0 15 50 13 20.0 III. 1 10 2.0 18.1 2 10 1.0 8.3 3 10 .5 3.8 4 10 .0 0.0 5 10 .0 0.0 These trials show that increasing the amount of Skellysolve in the system necessitates an increased amount of ethyl ether to bring about misci- bility. In contrast, increases in the amount of ethyl alcohol do not re- quire increased amounts of ethyl ether to bring about miscibility. These data also indicate that when 20 per cent or the system is ethyl ether, the system is entirely miscible despite variations in the amounts of ethyl a1- cohol. I The addition of water to the above mixture resulted in a feurb component system which responded to the addition of ethyl ether in a manner somewhat similar to the three-component system. Additions of ethyl ether lowered the demarcation line until it was at the level of the amount of water present. It was found that this immiscible layer persisted at the same vol- 29 ume as that of the water added, despite large additions of ethyl ether. The quantity of ethyl other necessary to bring the line down to the level of the volume of water in the system was increased in this four-component system to 60 per cent when reagents are used in.Mojonnier preportions, compared to the 20 per cent required when water is absent. It was likewise noticed that when larger amounts of water are present, more ethyl other is required to bring about miscibility. When larger amounts of Skellysolve are present, less ethyl ether is required. As noted previously, a considerable difference exists between the volumes of the lower layers of the Mojonnier extracts and the extracts in the modified method in which Skellysolve “A" is used in place of ethyl ether. This may not be entirely accounted for by the difference in miscibility of the two materials. In order to study other effects on the volume of the lower layer, several trials were conducted in which the volumes obtained when using other were compared with volumes obtained using Skellysolve ”A". IMixtures of Skellysolve and ethyl etherfwere prepared which contained 0, 20, 40, 60, 80, and 100 per cent ethyl ether reSpectively. Forty milli- liters of the desired mixture were combined with all other Mojonnier reagents except petroleum ether, with the milk being replaced by water in the first trial and skim.milk in the second trial. These results are shown in.Table 6. Table 6. variation in Volume of the Lower Layer as Affected by varying Composition of Ether-Skellysolve Mixture when 1.5 ml. Ammonia, 15 ml. Alcohol, and 40 ml. of the EtherrSkellysolve'Mixture Are Present* velume of’Lower’Layer when the Following ‘ Percentage of Ethyl Ether Is Used Wgter Skimilk Milk 0 20 go 60 _ so 100 (1111.) (ml.) (1111.) (m1.) (m1.) (1:11.) (1:11.), (1:11.) (1:11.) 10 0 - 25 25 25 23 19.5 10.5 - lo - 26 26 25 22.5 17.0 10.0 - - 10 26 26 26 22 16.5 9.0 *Average of four trials. These data show that in the presence of all mojonnier reagents ex- cept petroleum ether, the volume of the lower layer is roughly proportional to the amount of Skellysolve in the system. These trials were repeated, with the exception that petroleum ether was included in its regular amount. The same mixtures of ethyl ether and Skellysolve ”AF were used, and no other changes were made. The results of these trials are shown in Table 7. Table 7. variation in.Volume of the Lower’Layer as Affected by varying Composition of Ether-Skellysolve Mixture when 1.5 ml. Ammonia, 15 ml. Alcohol, 40 ml. of the Ether-Skellysolve Mixture, and 40 ml. of Petroleum Ether are Present* Volume of Lower Layer when the Following Percentage of Ethyl Ether Is Used Watey Skimmilk Mi 1k 0 20 40 so so 100 (m1.) (n1.) (m1.) (m1.) (m1.) (m1.) (m1.) (filo) (ml.) 10 v r. 25.0 25.0 23.5 23.5 22.0 21.0 - lo - 25.0 2500 2500 2300 22.0 20.5 iAverage of three trials. These data indicate that the volume of the lower layer is increased with increased amounts of Skellysolve in the system. However, the differ- ence in volume between the extracts of 0 per cent ethyl ether and the 100 per cent ethyl ether is decreased in these trials due to the presence of petroleum ether. The volume change in the lower layer resulting from.the use of petroleum ether amounts to approximately 11 ml. (66 per cent), and is caused by the removal of alcohol or water from the ethyl ether. Table 8 shows the results obtained when the petroleum ether was re- placed by the same quantity of Skellysolve "F". All other reagents were left unchanged. 31 Table 8. variation in.volume of the Lower Layer as Affected by varying Composition of Ether-Skellysolve Mixture when 1.5 ml. Ammonia, 15 ml. Alcohol, 40 ml. of the Ether-Skellysolve Mixture, and 40 ml. of Skellysolve "F" are PresentIII velume of the Lower Layer when the Following Percentage of Ethyl Ether Is Used Nate; Skimmilk Milk 0 20 40 50 80 100 (mlo) (mlo) (mlo) (ml.) (ml.) (ml.) (ml.) (ml-) (ml.) 10 " " 24.0 2400 23.5 23.0 22.0 2200 - 10 - 2402 2404 24.0 23.0 2105 20.0 - - 10 24.8 26.0 24.0 23.7 22.0 20.0 1"Average of three trials. These results indicate that there is little difference between the effects of petroleum ether and Skellysolve ”F” in the reapect of regulating the volume of the lower layer. In these trials also, the sample with the greater amount of Skellysolve "A” and with less ethyl ether showed a greater volume in the lower layer than when ethyl ether was used alone. These data show that in following the Mojonnier procedure, 3 to 5 ml. of alcohol are carried into the other layer, while little or none remains in the ether layer when the modified procedure is used. Along with changes in the volume of the lower layer, noticeable changes also occurred in the color and opacity of the lower layers as the ethyl other content was increased from 0 per cent to 100 per cent. In the trials in which no second solvent was used (Table 6) trials conducted on whole milk showed the yellowish (fat-like) layer between the ether layer and lower layer in all mixtures, but was not observed when only ethyl ether was used. The lower layers of the O, 20, 40, and 60 per cent mixtures showed a white Opaque color, which was not present to such an extent in the 80 and 100 per cent mixtures. The volume of the fat-like layer was the greatest (1.5 ml.) when the 20 per cent mixture was used. The fat-like layer was present only in the 0 per cent, 20 per cent and 40 per cent mixtures in the trials where petroleum ether was used (Table 8). 32 The colors of the lower layers in these three mixtures were less clear than in the 60 per cent, 80 per cent, and 100 per cent mixtures which showed no yellow, flocculent layers between the other layer and lower layer. The vol- ume and appearance of this yellowish layer and the lower layer did not change on standing 24 hours. Similar results were obtained when comparing the mix- tures in which the petroleum ether was replaced by Skellysolve ”F”. Partial Replacement by Skellysolve Products: I Results of these previous investigations indicate that when the 60 per cent mixture of ethyl,ether and Skellysolve ”AF was used, the alcohol and Skellysolve were miscible, the volume of the lower layer was not too great to permit a second extraction, the color of the lower layer was satis- factory, and there was no evidence of the flocculent, fat-like layer. A 50 per cent mixture was also prepared and this was found to exhibit the same physical characteristics as the 60 per cent mixture. On the basis of these preliminary observations, fat determinations were made using mixtures of 50 per cent, 60 per cent and 75 per cent ethyl ether in Skellysolve ”A”. These mixtures were used as substitutes for ethyl ether in the modified experimental procedures. 1, USe of a 1:1 mixture pf Skellysolve ”A” and ethyl ether: The original trials conducted during July'using this 1:1 mixture of ethyl ether and Skellysolve "A" were conducted using Skellysolve ”F” as the second solvent. Later work was done in August with the same mixture, using SkellysolveflA” as the second solvent. The results of these trials are pre- sented in Table 9. These data show that the use of the ethyl ether Skellysolve "A" mix- ture results in lower values than were obtained by the Mojonnier control method. When Skellysolve ”F” was used as second solvent the differences were as follows: milk, 0.0495110.0030 per cent; homogenized milk, 0.10321 0.0260 33 Table 9. Comparison of the Mojonnier*Method with.a.Method whicthtilizes a 1:1 Mixture of Ethyl Ether and Skellysolve ”A”* ‘ __ harshness _ Product No. of Fat Obtained Between Methods Between Duplicates ___ _rfT:1als ‘Mo onn a Modified M onnier Modif ed 195') 1%) (i) (96) (5k) Skellysolve ”F" as Second Solvent Milk 38 3.8604 3.8109 0.0495 1: 0.0030 0 .0089 0.0244 Homo.Milk 8 4.1162 4.0130 0.10321=0.0260 0.0077 0.0186 Evap. Milk 6 8.1420 8.0475 0.0945 - 0.0228 0.0180 Ice Cream 5 11.0627 10.8080 0.2547 ‘- 0.0263 0.0343 Skellysolve ”A” as the Second Solvent Milk 9 3.9408 3.8960 0.04481: 0.0060 0.0091 0.0179 Homo. Milk 10 3.8454 3.7741 0.0713 t 0.0036 0.0084 0.0278 Evap. Milk 5 7.9775 7.6992 0.2783 - 0.0196 0.1312 Ice Cream. 9 11.1424 10.9567 0.1857 i 0.0307 0.0235 0.0447 iComplete data appears in Appendix Tables v and IX. per cent; evaporated milk, 0.0945 per cent; ice cream, 0.2547 per cent. When Skellysolve ”AF was used as the second solvent the differences were: milk, 0.0448!=0.0060 per cent; homogenized milk, 0.0713 £0.0036 per cent; evaporated milk, 0.2783 per cent; ice cream, 0.1857i:0.0307 per cent. These trials in- dicate that except in the case of evaporated milk, the Skellysolve ”A” results in values which are closer to the Mojonnier control values. Since this method involving a 1:1 mixture of Skellysolve ”A” and ethyl ether results in incomplete extraction, it was considered advisable to study the effect of a third extraction. Consequently a third extraction was uti- lized which consisted of mixing the residue of the second extraction with a 15 ml. portion of the 1:1 mixture of ethyl ether and Skellysolve "AF and a 15 ml. portion of Skellysolve ”F". The results of using this third extrac~ tion are shown in Table 10. These trials were conducted on samples obtained during JUly. 34 Table 10. Comparison of the Mojonnier Method with.a Modified method in which three Extractions are Made with a 1:1 Mixture of Ethyl Ether and Skellysolve 'A"* No. of Fat _p Differences Method Trials Obtainegg_Between Methods Between Du licate " 2' 175) 1%) (if) Mojonnier 9 3.9037 - 0.0108 Modified 9 3.8458 0.0579 0.0343 (2 extractions) IModified 9 3.8777 0.0326 0.0374 (3 extractions) . I00mplete data found in Appendix Table VI. The third extraction reduced the difference between the modified method and the control method from 0.0579 per cent to 0.0326 per cent, a reduction of 44 per cent. The data in Tables 9 and 10 show the agreement between duplicate determinations is more satisfactory when the Mojonnier procedure is used. However, the agreement between duplicates is satisfac- tory, and within range of normal experimental error in all cases with the exception of the 5 samples conducted on evaporated milk in which Skelly- solve "A” is used as the second solvent. (See Table 9) 2, Using a 3;2 mixture of Skellysolve.”h” and ethy; ethep with Shellysolve ”F" as the secong solven . Table 11 shows the results of trials conducted using a mixture of ethyl ether and Skellysolve ”A" in which the other makes up 60 per cent of the solution, or a 3:2 proportion. Skellysolve "F” was again used as the second solvent, and no other changes were made in the procedure. The determinations were conducted on samples obtained during July. These trials indicate that the use of this method also results in values lower than those obtained by using the Mojonnier method. The differ- ences between the two methods are: milk, 0.02981:0.0025 per cent; homogenized milk, 0.0452: 0.0141 per cent; evaporated milk, 0.0615t0.0375 per cent; ice cream, 0.128920.0496 per cent. 35 Table 11. Comparison of the Mojonnier'Method and a Modified Method which ‘Utilizes a 3:2 Pr0portion of Ethyl Ether and Skellysolve "A” Differences Product No. of Fat Obtained Between.Methods Between Duplicates Trials Mo onnier Modified‘ ngpnni r Modified (5.65 (fl 1%) 70137 (91») Milk 15 3.6 786 3. 6488 0.0298 1‘ 0.0025 0.0121 0.0160 Home. Milk 6 3.8378 3.7926 0.0452110.014l 0.0082 0.0235 Evap. Milk 6 7.9684 7.9069 0.06151 0.0375 0.0211 0.0317 Ice cream 6 11.2093 11.0804 0.12891 0.0496 0.0267 0.0451 *Complete data shown in Appendix Table VII. These data also show that the use of a 3:2 mixture of ethyl ether and Skellysolve "A” results in an improvement in extraction efficiency over the 1:1 mixture. The agreement between duplicate determinations is also greater when this 3:2 mixture is used. In this procedure, as in the preceding ones, it is interesting to note that in nearly all cases the extraction efficiency of the modified procedure is poorest in the trials conducted on high fat samples such as ice cream mix and evaporated milk. 3. The use of a 3:1 mixture of ethylpether and Skellysolvefigg:,. using Skellysolve ”F" as the second solvent. Trials were conducted in which the per cent ethyl ether in the mixture was 75 per cent, resulting in a 3:1 mixture of ethyl ether and Skellysolve ”A". This mixture was used to replace the ethyl ether as in previous trials. Skellysolve "F" was again used as the second solvent. The samples were obtained during August. Table 12 shows the results of these trials. These data show that in the analysis of milk the modified procedure extracts 0.02451:0.0038 per cent less fat than the Mojonnier control method. In homogenized milk this difference amounts to 0.01653 0.0037 per cent and in evaporated milk and ice cream.the differences are 0.0461 per cent and 0.0641 per cent reSpectively. 36 Table 12. Comparison of the Mojonnier Method with.a Modified Method which Utilizes a 1:3 Proportion of Ethyl Ether and Skellysolve "A” ___ Differences Product No. of Est Obtained Between Methods Between DuplicatES ___ Trials Mojonni r Modified Mojpnnier'Modified (ii) (%) (%) MY (7W iMilk 10 4.1066 4.0821 0.0245i:0.0038 0.0087 0.0108 Homo. Milk 8 4.6092 4.5927 0.0165I 0.0037 0.0087 0.0183 Evap.'uilk 3 7.9316 7.8855 0.0461 - 0.0227 0.0251 Ice Cream 3 10.4008 10.3367 0.0641 - 0.0162 0.0428 ‘Complete data found in Table VIII of the Appendix. These results also indicate that the 3:1'pr0portion of ethyl ether and Skellysolve ”A” is the most efficient extractor of all the mixtures tried, but does not equal the extraction ability of ethyl ether alone. Reduction in the Amount of Ethyl Ether 1. The use of 50 per cent of the normal amount of ethyl ethegip In -:___‘ w—~- order to determine the amount of extraction accomplished by the Skellysolve ”A” in the 1:1 mixture with ethyl ether, determinations were conducted in which the Skellysolve "A” was omitted, and the ethyl ether was used to the extent of 50 per cent of the normal amount. For the purpose of comparison, trials were conducted using both petroleum ether and Skellysolve ”A" as the second solvent. The amounts of all other reagents were unchanged and these studies were conducted during August. Table 13 shows the results of these trials. These data show that a 50 per cent reduction in the amount of ethyl ether used in the extraction results in incomplete extraction of the fat when compared to the Mojonnier control method. When petroleum ether is used as the second solvent, the differences between the control and modified methods are: milk, 0.03352 0.0034 per cent; homogenized milk, 0.0365t 0.0072 per cent; evaporated milk, 0.0549 per cent; ice cream, 0.0354 per cent. When Skellysolve 11" is used as the second solvent, the differences between the two methods are: 37 Table 13. Comparison of the Mojonnier.Method with a Modified Method which Utilizes 50 per cent of the Normal Amount of Ethyl Ether* Differences Product No. of Fat Obtained Between.Methods Between Duplicates Trials Mojonnier Modified Mojonnier Modified 1%) (%) ’T%) 120 Petroleum Ether as Second Solvent Milk 9 3.5351 3.5016 0.03351 0.0034 0.0068 0.0263 Homo. Milk 9 3.9354 3.8989 0.03653 0.0072 0.0086 0.0260 Evap. Milk 3 7.8909 7.8360 0.0549 -. 0.0219 0.0183 Ice Cream 3 10.7204 10.6850 0.0354 — 0.0336 0.0365 Skellysolve ”A” as Second Solvent Milk 8 4.2154 4.1960 0.01943 0.0040 0.0124 0.0078 Homo. Milk 8 4.3186 4.2927 0.02591 0.0019 0.0145 0.0227 Evap. Milk 3 7.9922 7.8913 0.1009 - 0.0217 0.0332 Ice Cream. 3 10.5433 10.3771 0.1662 - 0.0218 0.0365 *Complete data found in Tables XI and XIII of the Appendix. milk, 0.0194t 0.0040 per cent; homogenized milk, 0.02591 0.0018 per cent; evaporated milk, 0.1009 and ice cream, 0.1662 per cent. These data show that in the analysis of milk and homogenized.milk, the use of Skellysolve "A” results in more efficient extraction, while the limited data for evaporated milk and ice cream indicate petroleum ether is somewhat superior. A comparison of Table 13 with Table 9 will show that the modified procedure which used 50 per cent of the normal amount of ethyl other is more effective in fat extraction than is the 1:1 mixture of ethyl ether and Skellysolve ”A”. For example, in the case of milk, the use of a 1:1 mixture resulted in a difference of 0.0495 per cent, while this modification which uses reduced quantities resulted in an average difference of 0.0264 per cent. This would indicate that the Skellysolve ”A” itself is not efficient in ex- tracting fat when it is used in mixture with ethyl ether. In fact, these data show that the Skellysolve ”A” in mixture with ethyl ether has a detri- mental effect on the extraction efficiency of the ethyl ether. 38 Egg The use of 75 per cent of the normal amount of ethyl ether: In another series of trials, 75 per cent of the normal amount of ethyl ether was used as the first solvent with normal amounts of either petroleum ether or Skellysolve "A” being used as the second solvent. Table 14 shows the re- sults of these trials which were conducted during August. Table 14. Comparison of the Mojonnier*Method with.a.Modified Method which Utilizes 75 per cent of the Normal Amount of Ethyl Ether* _..- Differences Product No. of Fat Obtained Between Methods Between Duplicates Trials Mojonnier Modified . MojonnieriModified_ 1%? 775) 1%? 1733 7%) Petroleum Ether as Second Solvent Milk 4 3. 7821 5. 7715 0.0106 ' 0.0059 0.0188 Ice Cream 3 10.2645 10.2409 0.0236 0.0187 0.0210 ‘ Skellysolve "A” as Second Solvent' Milk 4 4.3435 4.3333 0.0102 0.0074 0.0071 Homo. Milk 10 3.8529 3.8365 0.0164!:0.0026 0.0137 0.0098 Evap. Milk 3 7.9464 7.9111 0.0353 0.0067 0.0102 Ice Cream 3 10.3947 10.3249 0.0698 0.0303 .0.0488 I00mplete data found in Appendix Tables III and XIV. The use of 75 per cent of the normal amount of ethyl ether gives re- sults which are close to, but significantly lower than the results obtained by the Mojonnier method. When petroleum ether is used as the second solvent the difference between the two methods in the analysis of milk is 0.0106 per cent, and in the analysis of ice cream 0.0236 per cent. The use of Skelly- solve ”A” as the second solvent results in the following differences: milk, 0.0102 per cent; homOgenized milk, 0.01641-0.0036 per cent; evaporated milk, 0.0353 per cent; ice cream, 0.0698 per cent. These data substantiate the findings in Table 13 that the petroleum ether is more effective as a second solvent than is the Skellysolve ”A" in the analysis of evaporated milk and ice cream. 39 A comparison of these results with the data presented in Table 12 again reveals that the Skellysolve ”A" itself does not contribute towards increased efficiency in fat extraction when mixed with ethyl ether. Possiblerlgfluence of Seasons Previous studies involving the replacement of ethyl ether were con- ducted during the months of June to September with no attention being given to possible seasonal influences. However, in later experiments dealing with the influence of certain factors on the efficiency of extraction by the vari- ous modified procedures, it was observed that the differences between the re— sults obtained by the modified and Mojonnier methods were appreciably greater than had been found in the earlier work. This change was noted in both the mixture and reduction modification. TheSe later experiments were conducted during November and December and general observations were made as follows: (a) The average discrepancies between results obtained by use of the Mojonnier and the modified methods is significantly higher than differences obtained in the summer months. (b) The differences between the methods covered a wider range than noted previously, certain samples resulting in smaller differences and other samples resulting in greater differences. (c) The color of the lower layer of liquid in the extraction flasks was noticeably'more Opaque and gelatinous when the modified procedures were used on samples of milk obtained ‘during the winter. This characteristic was sapecially obvious after the sec- ond centrifuging. (d) These larger discrepancies are apparently not caused by minor changes in technique or reagents, since changing and purification of the reagents and careful scrutinization of technique resulted in no appreciable change. Table 15 presents results obtained by comparing the Mojonnier method With the modified method which employs a 1:1 mixture of ethyl ether and Skelly- 40 solve "A" and that which utilizes reduced quantities of ethyl ether in the analysis of milk produced under winter conditions. For the purpose of com- parison, the data from.tables 9 and 13 on the analysis of milk under summer conditions are also shown. Table 15. Comparison of the Mojonnier Method with a Method which'Utilizes a 1:1 Mixture of Ethyl Ether and Skellysolve ”A” and a Method which Employs 50 per cent of the Normal Quantity of Ethyl Ether in the Analysis of Milk Produced under varying Seasonal Conditions* No. of Fat Obtained Differences Medificgtion Trial§_ Mojonnier Modified Between Methods Ran (is) (9'6) (967 (7425 1:1 Mixture Winter 33 3.7867 3.6971 0.0896110.0049 0.049l-0.1522 Summer 47 3.9446 3.8960 0.0496310.0040 0.0133- 0.0960 50% Reduction Winter 25 5.9551 3.8870 0.06611 0.0075 0.0078-0.1207 Summer 17 4.2209 4.1960 0.02491 0.0037 -0.0060-0.0492 :50mplete data found in Appendix Table XV} These comparisons indicate that a definite change has occurred in the relationship of these modified procedures to the Mejonnier procedure over a period of a few months. The difference between the Mojonnier results and the results obtained using the 1:1 mixture has increased from 0.0496 per cent to 0.0896 per cent, an increase of 0.0400 per cent. .4 correSponding increase of 0.0412 per cent is noted in the relationship between the Mojonnier method and the modification which.makes use of reduced quantities of ether. The range of differences obtained by using the 1:1 mixture is increased from 0.0827 per cent to 0.1031 per cent, with.minimum and maximum values both being higher for the winter analyses than for the correSponding summer values. In the results obtained by use of the 50 per cent reduced quantity method, the range of dif- ferences is increased from 0.0552 per cent to 0.1129 per cent. This change, itself, is minor, but it is in line with other findings. 41 The lower layer of liquid in the extraction flasks containing the win- ter mdlk showed an opaque color and a gelatinous nature in all trials con~ ducted using the modified procedures. These characteristics were not present during the summer. The appearance of these lower layers suggested incomplete dissolving of the protein materiel. Effect of various Factors on theggfficiency_of the Medified Methods ;L__§ffect of increased shaking on the efficiency of fat extraction: In an attempt to bring about closer agreement between the Mojonnier method and the modified methods, the milk being analyzed by use of the modi- fied procedures was subjected to a 60 second shaking interval after the addi- tion of each solvent, instead of the regular 30 second shaking time. This variation was administered to both the modification which uses the 1:1 mixture of ethyl ether and Skellysolve ”A" and the modification which utilizes 50 per cent of the standard quantity of ethyl ether. The amounts of reagents used, and other techniques remained the same, and these determinations were con- ducted during January'Table 16 shows the results of these trials. Table 16. Effect of’Lengthening the Shaking Period on the Efficiency of the Modification Using the 1:1 Mixture and Reduced Quantities of Ethyl Ether as Compared to the Mojonnier Method in the.Analysis of Winter Milk* Fat Obtained :Deviation from.Moionnigg Method No. of Regular Modified with Regular Modified with Trials Mo onnier'Modified 60 sec.shaki Medif ed 60 sec.shakin i%5 1%) 1%} (%5 (%5 1:1 Mixture 11 3.5446 3.4594 3.5037 0.085220.0091 0.04362 0.0060 50% Reduced 6 3.7643 3.7262 3.7368 0.0380 0.0275 i00mplete data found in Appendix Tables XVI and XVII. These data show that increased shaking causes the modified methods to extract more fat than is accomplished by normal shaking. The difference be- tween the l:1 mixture procedure and the MOJonnier control is reduced from 42 0.0852 3%, per cent to 0.0436 t0.0060 per cent due to the effect of addi- tional shaking with the solvents. This difference of 0.0436 1‘ 0.0060 is well within the established range of differences which were obtained by use of this modified method during the summer months. This technique of increasing the shaking period from 30 seconds to 60 seconds when applied to the method which uses 50 per cent of the normal amount ethyl ether reduces the difference be- tween this modification due the Mojonnier from 0.0380 per cent to 0.0275 per cent. This limited number of trials indicates that the extra shaking period affects this modification much as it does the modification utilizing the mix- ture of ethyl and Skellysolve ”A”, although the increase in efficiency is less pronounced. 2, Extraction efficiency_of the method which uses a lzlgmixture when used fog the analysis of milk obtained from animals on fat feeding trials, To ascertain if the change in character of the fat as caused by seasonal variation may be a factor in the change in extraction efficiency observed dur- ing the seasons, feeding trials utilizing linseed oil were conducted during November and the milk then analyzed for fat. In this feeding trial, conducted with the view of obtaining a softer, lower melting-point fat, COW'NO. A 24 from the College Experimental herd was selected and fed linseed oil at the rate of 0.7 lbs. per day, the 011 being blended with silage to insure complete assimilation. COW'NO. 77 was used as a control, and received a normal winter ration supplemented with corn. Samples for analysis were obtained before feed- ing the oil and after 4 and 7 days respectively. These samples were analyzed by the modified method which utilizes a 1:1 mixture of ethyl ether and Skelly- solve "A". This work was conducted during November and the results are shown in Table 17. 43 Table 17., Cmmparison of the‘MoJonnier Method with the Modified.Method which.Utilizes a 1:1 Mixture of Ethyl Ether and Skellysolve "A" in the Analysis of Milk Obtained when Feeding Linseed 011 to the Cow Fat_thained__s Difference between Days of oil ngonnier Modified Methods £eeding_ Cow A24 Cow #77 COW'A24, Cowgfiz7 COW’A24 Cow £27 (963— W) (757) T70) W (967 0 days 3.3009 5.1156 3.6753 4.9917 0.1256 0.1239 After 4 days 4.3559 - 4.2233 - 0.1326 - After 7 days 4.7186 7.5092 8.7522 7.2665 0.9664 0.2427 These trials show that softening the fat by means of feeding linseed oil does not result in improved efficiency in the modified procedure. In fact the reverse situation prevailed. The large increase in difference be- tween the control and modified methods from 0.1256 to 0.9664 is due to the change in ration, and shows that significant increases in the difference be- tween the methods can be caused by changes in feeding. The large increase in fat content exhibited by control OOW'NO. 77 is-an occurrence which is not ascribed to its ration. Another observation was that in the flasks containing thelnodified re- agents, a white gelatinous emulsion was present particularly after the second extraction, although all reagents were used prOperly. This characteristic was noted in the milk from both cows. Eh“ Effect of low temperature storage on efficiency of the modified procedure which employs the 1:; mixture of ethyl ether and Skellysolye ”A". In other trials, the fat hardening as affected by holding for long periods at a low temperature was studied to find its effect on the extraction efficiency of this modified procedure. Pasteurized samples were obtained from the creamery immediately after cooling, and raw samples obtained directly from the pail at the time of milking for the 0 hour samples. The samples were examined immediately, the raw samples were still warm, and were then aged at 4° C for varying periods up to 72 hours, when each sample was again subjected to analysis. The results of these determinations which were obtained during November and January are shown in Table 18. Table 18. “Effect of Storing Milk at LoW'Temperature on the Comparison between the MOJonnier Method and the Modified Method which'Utilizes a 1:1 MUxture of Ethyl Ether and Skellysolve "A7 Sample cFat_thained No MO onnie Modified Difference 07.) W (9%) History of Sample 1 3.8861 3.8184 0.0677 0 hours from.bottler 3.8604 3.7922 0.0682 24 hours from bottler 3.8640 3.7698 0.0942 48 hours from bottler 2 3.7385 3.6633 0.0752 0 hours from bottler 3.7556 3.6741 0.0815 24 hours from.bottler 3.7541 3.6889 0.0652 72 hours from bottler 3 3.4030 3.3523 0.0507 0 hours from cow 3.3806 3.3315 0.0491 48 hours from cow 4 3.9283 3.8634 0.0649 0 hours from cow 3.8932 3.7888 0.1044 24 hours from cow 5 5.6465 5.4819 0.1646 0 hours from cow 5.6774 5.5336 0.1438 24 hours from cow 6 5.0725 4.9206 0.1519 0 hours from cow 5.0502 5.9149 0.1353 24 hours from cow 7 3.4995 3.3703 0.1292 0 hours from cow 3.4590 3.3334 0.1256 72 hours from cow These data indicate that the difference between the Mojonnierlnethod and the modified mixture procedure is not significantly affected by low tem- perature storage. Sample No. 4 is the only sample showing a significant in- crease in the difference; this increase amounting to 0.0395 per cent. Other samples show Slight decreases in the difference between the two methods after storage. Low temperature storage slightly decreased the results secured by the Mojonnier method in the majority of the trials. Each sample in this series shows a decrease of 0.025 per cent to 0.040 per cent except sample No. 2 which shows an increase of 0.017 per cent after storage. 4;» Effect of heating_the sample to 1000 F immediately_before analysis In another experiment, the fat was softened by heating the milk to 100° F for a few minutes Just prior to weighing the sample into the extraction flask. As a control, analyses by the Mojonnier procedure were made on samples which did not receive the heat treatment. The modified procedure used here employed a 1:1 mixture of ethyl ether and Skellysolve "AF. These trials were conducted during December, and the results are presented in Table 19. Table 19. Effect of Heating'Milk to 1000 F Immediately before Test- ing on the Efficiency of the Method Utilizing the 1:1 Mixture of Ethyl Ether and Skellysolve ”A”* _fi Fat Obtained Deviation from Mo onnier No. of Regular' Modified Regular Modified Tr ale Mo onnier ‘Modified plus Heat gMpdified plus Heat 1%) M (a) 6%) 1%) 8 3.7365 3.6583 3.6778 0.07821I0.0088 0.0587 0.0098 fcbmplete data found in Table XVIII of the Appendix. These data show that warming the milk to 1000 F Just before the anal- ysis is made, reduces the difference between the modified and control pro- cedures by 0.0195 per cent. This slight increase in efficiency did not occur in each case as demonstrated by sample No. 4 (Appendix Table 18). In this case this technique resulted in an increase of 0.0511 per cent in the differ- ence between the Mojonnier method and the modified procedure - an unexplain- able abnormal variation. The average increase in efficiency found here, 0.0195 per cent, is of doubtful significance and does not account for the seasonal difference which has been noted. 46 5, Effect of variation in the amounts of ammonia and alcohol on the efficiency of the modified procedures, In order to decrease the difference between the results obtained by the Mojonnier control method and the reduced modification and mixture modification procedures, and thereby obtain a method which resembles the Mojonnier more closely, certain variations in alcohol and ammonia were conducted. As stated previously, a gelatinous condition exists in the flasks con- taining the modified reagents in the trials conducted during the winter. This white, gelatinous texture in the lower layer is believed to be due to undis— solved protein, and inferior extraction efficiency is caused by the trapping of fat in this gelatinous layer. Results obtained by use of these variations in the quantities of am- monia and alcohol are shown in Tables 20 and 21. All these trials were con- ducted during January. Table 20. Effect of'Variations in the Quantities of Ammonia and Alcohol on the Efficiency of the Modified Procedure which Employs a 1:1 Mixture of Ethyl Ether and Skelly- solve Fat Obtained * neyiation from.ngonnier 1 3.7385 3.6633 3.6464 3.6574 0.0752 0.0921 0.0811 2 3.7556 3.6741 3.6400 3.6478 0.0815 0.1156 0.1078 3 3.4030 3.3523 3.3473 3.3290 0.0507 0.0554 0.0740 4 3.3806 3.3315 3.3075 3.2836 0.0491 0.0731 0.0970 5 3.9283 3.8634 3.8504 3.8554 0.0649 0.7790 0.0729 Ave. 3.6412 3.5769 3.5583. 3.5546 0.0643 0.0829 0.0866 *Variation I consisted of the regular modification which makes use of a 1:1 mixture of ethyl ether and Skellysolve "A". variation II consisted of using 1 ml. ammonia and no alcohol in the second extraction together with the mixture procedure. variation III consisted of using 1 ml. ammonia and 21-m1. alcohol in the second extraction together with the mixture procedure. 47 These variations in alcohol and ammonia served to increase the average difference between the Mojonnier results and the regular modified results by 0.0168 and 0.0223 per cent respectively. For this reason these variations are not practical in increasing the efficiency of the regular modified pro- cedure. Table 21 shows the results of still another variation in the quantity of ammonia. This variation is accompanied by the use of heat, which has pre- viously been shown to have some effect in increasing the efficiency of the modified methods. These determinations were conducted using both the mixture and reduced quantity modifications, and are compared to the Mejonnier method and the regular modified procedures. These data show that heating the sample to 100° F just before analysis combined with the use of additional ammonia in the second extraction serves to decrease the difference between the Mojonnier procedure and both of the two modified methods. ”The application of these techniques to the method which uses the 1:1 mixture of ethyl ether and Skellysolve ”A” resulted in an in~ crease of 0.0384 per cent in extraction efficiency, leaving a difference of only 0.0581t 0.0085 per cent between this technique and the Mojonnier in the fourteen trials conducted. This treatment increased the efficiency of the reduced method by 0.0319 per cent, resulting in an average value for six trials of 0.0584 per cent lower than results obtained by the Mojonnier method. In the six analyses in which both.modified procedures were used, the modification III resulted in an average of 0.0761 per cent below the Mojonnier results, while modification IV gave a value which was 0.0584 per cent lower. This re- lationship between these two modifications is in line with earlier findings. Comparison of these results with the data presented in table 9 shows that this application of heat and extra ammonia causes the results_obtained .eanossw Hwnoapficum use amen Mo was he coasme manwaaaam HH qoapeeahfiuos mo mamfimnoo_>H noapmoahwoo: .noapomspwo cnooom one weannc cones ma manoesm mo Janene .mamhamqm osowon h oooa ow unseen ma panama one page mnoaamoowo one news H doapweamaooa a“ we wducoooosm no mpmamnoo HHH soapmofimaoos .eefie 136 so heapnwsw onmcqmpm we» go peso son on mnemoamso venues cowaaeos Headway one go mumamnoo HH qoflpmoayfimo: .onfivwfia and one mafia: compos coawficoe_smdswoa on» mo mpmamnoo H aoapmofikwcose I ewmo.o nwoo.o.uammo.o nomo.o omno.onwmomo.o m¢¢¢.v Oboo.¢ on~¢.¢ mmmm.n ammo.¢ .o>¢ ommo.o onnc.o momH.o ammo.o mbnm.n Nomm.n mmn¢.n prom.n moan .n «H mmeo.o nch.o awoa.o ammo.o nmmm.n mesm.n mowm.n Hsem.n Hmnw. n ma mnso.o mnmo.o womo.o wmmH.o name.n 000$.» mma¢.n norm.» nom¢.n ma mmmo.o nnmo.o memo.o emHH.o voom.n nmm¢.n owm¢.n mmo¢.n ommn.n HA wHwo.o nmnH.o aboa.o mama.o HHmm.¢ onnm.¢ mmsm.¢ momm.« nNm0.n 0H owec40a onwo.o mnno.o mvwa.o meow.n mon.n ommm.n mame.n no¢m.n m omeo.o embo.o mHNm.m #mmb.n mflum.n m mHmo.o momd.o nmem.n ecom.n oomm.n r Homo.o ¢¢0H.o Humm.n mama.” meow.» o nmoo.o memc.o onmm.n ewom.n ammo.» n oaoo.o Havo.o Oman.n n~mn.n womn.m w Nmoo.o somo.o mama.» nmmn.n owo¢.n n owwo.o mamo.o onmm.n demo.» onnm.n N nmoo.o . mmu¢.0 onsm.n name.» mmwb.n H “iv a &v “en E .83 HHH .82 HH .8: up .32 .8 3.0.8: .02 Houndowoz_eonm soapmwbon coawwpno ash Heaps compo: Abandonoz on» o» wonwgaoo mm .mohdcooosm unawaeos 039 Ho manageamwm one do manose< Mo pqnoa< on» ma nowpwwnmb Mo peekhfi .Hm canes 49 by use of the modified method which uses the 1:1 mixture to fall within the range of values established for this method in the analysis of summer milk. The efficiency of the modified method which uses reduced quantities of ethyl ether was not increased by this treatment to a point where it was equivalent to the results obtained during the summer months. The combination of heat treatment and extra ammonia also served to destroy the Opaque, gelatinous appearance which the lower layer of'liquid in the extraction flasks had developed during the winter months. Attention is called to the differences obtained by using modification III on samples 3 and 5. These samples were analyzed immediately after being drawn from.the cow, and the combination of heat and added ammonia in the sec- ond extraction resulted in differences of 0.0052 and 0.0053 per cent respec- tively from the Mbjonnier control method. These remarkably slight differences were not obtained by this procedure when the milk was allowed to stand for 24 hours at 4° C, as is shown by samples 4 and 6 which are the same respective samples after being held for this length of time. Similarly other samples obtained directly from the cow did not give similar results as is shown by trials 9-14, all of which were analyzed directly after being drawn from the cow. In two separate preliminary trials the combined effects of heat treat- ment and extra ammonia on the efficiency of the modified method which used reduced quantities of ethyl ether were studied. These preliminary studies gave evidence that this added technique resulted in values which were 0.0346 and 0.0299 per cent higher than the'Mojonnier control results. The normal reduced procedure gave results in these studies which were typical of other values obtained in the winter, their average being 0.0492 per cent lower than the Mojonnier value. As is shown by trials 9-14 of Table 21, these results 50 could not be repeated, leaving the explanation of these results and thoseob- tained in trials 3 and 5 a matter of conjecture. 6, The effect of milk preservatives on the efficiency of the modi- fiied method which uses reduced quantities 9f etgyl ether, To study the effect of milk preservatives on the extraction efficiency of the reduced modified method, trials were conducted in which this method was compared to the Mojonnier method in the analysis of milk which had been preserved by 0.0017 per cent of corrosive sublimate in 250 ml. of milk, and held for 7 days at 40 C. For each sample treated with corrosive sublimate a duplicate sample was carried along for a control which contained no preservative. The data which appear in Table 22 were obtained during January. .These data show that, in general, the relationship between the Mojonnier’method and the modified method which utilizes 50 per cent of the standard amount of ethyl ether is not affected by the use of such a preservative as corrosive subli- mate. The average difference between those two methods at 0 days is 0.0822 per cent, and after 7 days of holding both with and without the presence of corrosive sublimate the difference between the results obtained by the two methods is 0.0806 per cent. Although the relationship between the reduced method and the'MOjonnier is unchanged through this treatment, it can be seen that both methods ex- tracted less fat in the samples, both preserved and unpreserved, which had been held 7 days than in the same samples at 0 days. This decrease in ap- parent fat content of the samples may be due to a deveIOpment of rancidity. Samples 6 and 7 were the only pasteurized samples, and these show only a slight change in fat content due to aging, whereas sample 10 had deve10ped a strong rancid flavor after holding and this sample shows the greatest decrease in fat after storage, a decrease of 0.1463 per cent. 51 .mHeEp .30.“ $3 no smegma... mamo.o *Hmme.n osos.n aomo.o «can.» some.» mamo.o same.» owes.» .o>< ammo.o enne.n oeme.n «noo.o damn.n nmee.n msHH.o meme.m meom.n oa seeo.o seem.» «sow.n smeo.o eomo.m ease.» msoa.o seem.n meme.» 0 memo.o smes.n annm.n mmoo.o suns.n mama.» mHoH.o name.n mmmm.n m mmmn.o nfleo.n Haom.n ammo.o ends.» same.» som~.o msmm.n owem.n s enmm.a oseo.o nmom.H nmsm.a nmeo.o Heme.” eoem.a o maam.n ammo.o Hana.n menm.n memo.o memm.n nemm.n a News.» «smo.o mmsm.n Hee«.n oneo.o moms.» mmmm.n e seam.» ammo.o eons.n smem.n ammo.o Hesm.n eeno.e n mmmn.¢ ooafl.o oomm.¢ ooen.e memo.o seoe.e omee.e m mamm.¢ emoa.o momm.e mmwm.¢ neoa.o swam.¢ eofla.e a emmwweos_ummmwowoa me%NWmsw eM%mweoa amaquvos .oz non. 2593.33 ally—.8 dmbfinomoamnn omoaponfiya IIWHdEmm cobaom.m....H-m.l .ee s - eoaaapno use .. escapes .4qqnm.nnqmqummq paw define ooooaowean causaapsm mpfimonnoo_so eopnemoam seem mmm.noaes sass no mammaeq< as» a“ Heaps Heaps co mmeHPemsa eooseom maoHasm soaa3.mnseooonm assesses on» spas eczema noana0n02.onp so nowansosoo .mm capes 52 Other trials were conducted in which a 3:2 mixture of ethyl ether and Skellysolve "A” was cmmpared to the Mojonnier method in the analysis of milk which had been preserved by the use of 0.0017 per cent corrosive sublimate. These data are presented in Appendierable X and indicate that the preserve- tive together with 7 days of aging resulted in a difference of 0.0621 per cent between this modification and the Mojonnier*method. The Use of the MOdified methods in the analysis of LowaFat Egodpctg In order to find the comparative efficiency of the two modified pro- cedures in the analysis of products which contain small amounts of fat, sev- eral trials were conducted in which the modified methods were compared to the MOJonnier in the analysis of skimmilk and churned buttermilk. The two modi- fied procedures used here are those which (a) utilize reduced quantities (50 per cent) of ethyl ether and (b) employ a 1:1 mixture of ethyl ether and Skellysolve‘”A7. Table 28 shows the results of these trials which were conducted dur- ing August and January. From.these data it can be seen that in the analysis of churned butter- milk the method which uses only 50 per cent of the standard quantity of ethyl ether extracted only 0.0152 per cent less fat than the Mbjonnier method. Howe ever, the method which utilizes the 1:1 mixture of ethyl ether and Skellysolve ”A” is much less efficient in the analysis of buttermilk, extracting less than 55 per cent as much as the MOJonnier technique, the difference between these methods amounting to 0.2711 per cent. In the case of skimmilk, however, the inefficiency of this procedure is less pronounced, the difference between it and the Mojonnier control amounting to only 0.0130 per cent. This represents 25' ;;:i5.per cent efficiency when compared to the results obtained by the Mejon- nier method. 53 Table 23. Comparison of the Mojonnier Method with Two'MOdified Methods which Use Reduced Quantities of Ethyl Ether and a 1:1 Mixture of Ethyl Ether and Skellysolve "A" in the Analysis of Low-Fat Products* No. of ___ Fat Obtained__~g_ Deviation from.Mojonnie; groduct Trials Mpignnierngixture Reduced 50%» Mixture Reduced 50% (73 '*'""T%D (%) (%J’ (%3 Buttermilk 3 0.6345 0.6193 0.0152 Buttermilk 0.6014 0.3303 0.2711 Skimmilk 0.1411 0.1281 0.0130 Powdered Buttermilk 5.9115 4.8405 5.5105 1.0710 0.4010 Powdered Skimmilk 0.7707 0.4968 0.4883 0.2739 0.2824 When these same modified procedures are compared to the Mojonnier technique in the analysis of powdered skimmilk, they give somewhat similar results. In these trials the 1:1 mixture gave values which were 0.2739 per cent lower than the Mojonnier, representing 64.5 per cent efficiency, while the reduced quantity procedure resulted in a difference of 0.2824 per cent from the Mojonnier, an efficiency of 63.4 per cent. The two modified methods give dissimilar results when powdered buttermilk is analyzed. .In these trials the reduced quantity procedure resulted in a difference of 0.4010 per cent from.the Mojonnier, representing an efficiency of 93.22 per cent, while the mixture procedure resulted in a difference of 1.0710 per cent, an efficiency of 81.88 per cent. *Complete data found in Appendix Table 11X. 54 DISCUSSION Of primary importance in evaluating a method is a demonstration of its accuracy when compared to a well-known standard procedure. In this particular study involving fat determination in dairy products, the control method uti- lized is the Mojonnier technique. Results obtained by the modified methods are compared thereto to ascertain the comparative accuracies of these proced- ures. To demonstrate the relationship between the Mojonnier and modified methods, statistical treatment of the data involving the use of regression curves may be used satisfactorily. Such treatment has been administered in this study wherever sufficient data were available. By an application of such curves, the per cent fat extracted by the modified method may be used to esti- mate the per cent fat which would be extracted by the Mojonnier method. This calculation is accomplished by the use of the equation y = a O bx where y represents the unknown Mojonnier value, x stands for the value obtained by use of the modified procedure, and a and b are constant values representing the slepe and intercept of the regression curve. Results obtained in these studies of the various modifications of the Mojonnier method indicate that the only modification entirely satisfactory is that in which the petroleum ether is replaced by Skellysolve "A" or ”F". The data obtained by use of this substitution reveal that if the technique is preperly executed, there is no significant difference between the results ob- tained by its use and the Mojonnier procedure. This is illustrated by Figure I which shows the relationship between the‘Mojonnier'method and the modified method in which petroleum ether is replaced by Skellysolve ”F". This regres- sion curve reveals practically perfect correlation between the two methods. __ MOJONN/ER PER CENT FAT 55 11451»———— 40+.— 3.05—— 2.5wr——-—- ___L,.l l 1 l l i I J I 2.0 "‘ - 2.5 3.0 3.5 4.0 4-5 PER CENT £417.. MOD/F/ED figure I. Relationship between the Mojonnier Method of Fat Extrrtion the r modified Method in which the Pe- troleum Ether is Repl:ced by Skellysolve "F" 56 0f the types of modifications involving the first solvent, ethyl ether, none exhibited the high degree of accuracy demonstrated in the previous sub- stitution modification. Those modifications which utilized reduced quantities of ethyl ether resulted in incomplete extraction to a greater or less degree when compared to the Mojonnier, the efficiency of extraction being directly related to the quantity of ether used. Figures II and III show the regression curves which were established for the reduction modifications using 50 per cent and 25 per cent less ether reSpectively in the analysis of summer milk. Fig- ure IV shows the relationship between the 50 per cent reduced method and the 'Mojonnier method in the analysis of winter milk. These graphs reveal that when the 25 per cent reduction method is used in the analysis of summer milk, a difference of 0.0104 per cent can be expected between the modified procedure and the Mojonnier, whereas the 50 per cent reduction method under these condic tions results in a difference of 0.0269 per cent. This latter value is in- creased to 0.0661 per cent under winter conditions. Mojonnier and Troy (48) report that a 50 per cent reduction in ethyl ether resulted in a value which was 0.06 per cent lower than the Mojonnier, a value which is in close agree- ment with the value herein reported for winter milk. Substitution for ethyl ether of an equal volume of a mixture of ethyl ether and Skellysolve ”A" also yielded inferior results when compared to the standard Mojonnier procedure. Figures V and VI depict the relationship which exists between the Mojonnier method and the 1:1 mixture modification when Skellysolve ”F" and Skellysolve ”A? respectively are used as second solvents. Figure VII shows the relationship between the Mojonnier method and the 1:1 modification when the comparison is made on winter milk and the regression curve obtained by the comparison of the Mojonnier method is presented in Figure VIII. These graphs reveal that the efficiency of extraction is directly 57 a b. s t.» F” (o 9’ \. BEE CENT EA] __ M.._0_~_/_Q.NJV_L£E a J 1 1L1} J L... 3.5 1 J 3.7 3.9 4./ 4.3 4.5 PET? CENT fl/fl]__ MOD/E/ED figure II. Relationship between the fiojonnier Method and a Modified Method which Utilizes 50 per cent of the Standard gurntity of Ethyl Ether PER CENT EAT _‘MOJONNIETP 3.8 3.7 3.6 58 J 1 I l 3.7 3-8 PEAD CENT FAT __.' MOD/F/ED Figure III. Felationsbip between the Mojonnier Method of int Extraction 2nd a modified Method which Utilizes 75 per cent of the Stsnderd Quantity of Ethyl Ether A4 OJ CNN/5E? PER CENT EAT *- 59 4.4 —— "'""" 4.2!— ? : 0.0888 2 1.0096 3,41J1J1i1l1l 3.5 3.5 4.0 4.3 4.4 858 QL/LT FA] __ MOD/F/EQ -—Ib-lq.- figure IV. helrtionship between the mojonnier Method of Fat Extraction and t Modified Method which Utilizes 50 per cent of the Stflnderd Amount of Ethyl Ether in the finslysis of “inter Milk 3.9 MOJONN/EP C.» m u N 3.6 3.5 PER CENT FAT _ 3.4 3.3 60 4./‘ 4.0' l l 1 1H Ill 3.4 3.5 3.6 3.7 3.8 3.9 4.0 4.] 4.2 PER CENT FAT __ MOD/F/ED figure V. The Relationship between the mojonnier Method rad a Houified method in which a 1:1 Mixture of Ethyl Ether and Skellysolve "L" is Used as First Solvent, Skelly- solve ”F” being used :3 Second Solvent 4.6 T __ MOJONN/Ex‘? 3.6 61 . : 0.0934 b 2 0.9986 1111J1l1 figure VI. .17 ° are 3.9 47/ 4.3 4. 5 4. 7 Cfflj 1.5-fl... MOD/F/ED Relstionship between the Mojonnier Method of Fst Extraction xnd a Modified Method in which a l;l Mixture of Ethyl Ether and Skellysolve "L" is Used as First Solvent, Skellysolve "A" being Used as Second Solvent 4.0 3.8 MOJONN/fif? F” 9‘ C5122 FAZ __ L” PgR S” m 9’ ‘3 o; C) 3.2 3.4 3.5 3.3 4.0 .351? Qé'lNTF/LT... MOD/F/ED Bigure VII. Relttionship between the Mojonnier Method and a Modified Method which Utili7ee 8 1:1 Mixture of Ethyl Ether and Skellysolve "A” in the Analysis - of “inter Milk 63 3 .0 i u 1 J 1 1 l 3.5 4.0 4.5 4.8 PER CENT 54f __ MOD/F/£:_:_D_ Figure VIII. Reletionehip between the Mojonnier Method 6nd h Method which thee Uee of a 3:1 Mixture of Ethyl Ether and Skellysolve "E" 64 related to the preportion of ether in the mixture. In the analysis of sum- mer milk the 1:1 mixture modification averaged 0.04761?0.0046 per cent lower than the MOjonnier and the 5:1 mixture was lower by 0.0245310.0058 per cent. Although no graph is shown for the 3:2 mixture modification, this technique gave results 0.02981 0.0025 per cent lower than the Mojonnier. In the case of the 1:1 mixture method the difference from the Mojonnier is increased in the winter time to 0.08962:0.0049 per cent. Statistical analyses of all data obtained by use of these mixture modifications show that there is a signifi— cant difference between their results and the Mojonnier. Hewever, the differ- ence between Skellysolve ”A” and "F" as second solvents in the 1:1 procedure is not significant. The results obtained by use of the modification which uses only 50 per cent of the normal amount of ethyl ether, without any Skellysolve except as second solvent, indicate that poorer results are obtained by using 40 ml. of the 1:1 mixture than by using 20 m1. of ethyl alone. The decreased efficiency is probably due to the fact that deepite the apparent miscibility of the mix- ture with alcohol, the Skellysolve still exerts an inhibiting action on the penetration ability of the ethyl ether and alcohol. The results of both the reduced and mixture modifications were notice- ably improved by certain innovations in the technique. It was found that in analyzing winter milk, increasing the shaking time'from 50 seconds to 60 sec- onds reduced the average error in both types of modification by about 0.04 per cent. Thus, by creating this extra length of eXposure of the fat to the solvent, the inefficiency created by winter conditions was nullified. A simi- lar increase in winter efficiency was obtained by means of heating the sample to 100° F Just prior to analysis and using extra amounts of ammonia in the sec- ond extraction. Figures IX.and X show the reapective relationships of these .4:ng Q MILE CENT PER 2.8 65 1111111111 3.0 3.2 3.4 3.6 3.8 PER QQALT FAT __ MOD/1575.0 figure IX; Relationship between the iojcnnicr Method (nd a Method which uses a 1:1 Mixture of Ethyl Ether and Skellysolve "A” vhen the thking Perioc is Lengthened to Sixty Seconds (OJOA’N/EP PER CENT FAT __ 4.2* 4.0‘ 3.6 3.6‘ 3.4 3-2 66 _ fl b— w —. —-—* h— p— m t‘ ‘1 0.0535 b 2 1.0012 .1 1 [.1 11.! 3.4 3.6 3.8 4.0 4.2 PER CENT F/47____ fi/I'OD/F/ED Eigure X. Reletionship between the Hojonnier Method and a Modified Method which Uses a 1:1 Mixture of Ethyl Ether and Skellysolve ”A" fiith a further Treatment with Best rnd Additional Ammonia 67 variations in technique to the Mojonnier. The action Of heat alone tends to cause a slight reduction in the discrepancy, but both effects are needed for most efficient results. The effect of a third extraction with the 1:1 mixture procedure was also studied and resulted in a decrease of 0.0253 per cent in the discrepancy from the Mojonnier method. As has been stated in the results, certain procedures which gave in- ferior results likewise exhibited a gelatinous condition and Opaque color in the lower layer. This condition was first noticed when Skellysolve ”A” was substituted in entirety for ethyl ether, and later in the winter trials using both the mixture and reduced quantity modifications. The most pronounced case Of this type was observed when the final sample Of milk from.the fat feeding trials was analyzed. In general, there seems to be a relationship between the degree Of Opacity and gelatination and the efficiency Of fat extraction. This relationship would be expected if the protein surrounding the fat globules were not completely dissolved, resulting in a flocculent emulsion which prevents complete extraction Of the fat. In other trials, the observation was made that this Opacity and flocculence was destroyed and a normal appearance created by the use of additional ammonia and heat, indicating a protein effect. Since this trouble Of gelatination was present when winter’milk was analyzed by the modified procedures, but was not there during the summer, it was felt that the winter milk had undergone some slight, natural change in protein stability. It is believed that the fat thus trapped in the gelatinous lower layer represents the difference Observed between winter and summer effi- ciency of the modified procedures. This is indicated by the fact that varia- tions in technique which destroyed the Opaque color and gelatinous condition gave results which showed deviations from the MOjonnier results which were no greater than the normal summer discrepancy. In connection with the Observa- 68 tions relative to this seasonal influence, studies were made on the effect of low temperature storage Of milk and on the effect of changes in ration. The storage Of milk had no effect on the method, but the change in ration greatly altered the efficiency of the modified technique. For example, in the fat feeding trials, the final series of analyses conducted on this milk resulted in the greatest inefficiency Obtained in any trials by the mixture modification, and the gelatinous condition and Opaque color Of the lower layer were present at their greatest intensity. If such great changes in fat extraction and in the physical appearance Of the samples during analysis by the modified pro- cedures are induced by the change in ration herein made, then surely normal Winter practices of feeding may be expected to be a factor in influencing the seasonal changes in certain Of these extraction methods. Another characteristic was noted which is similar to the gelatinous condition and also occurred when poor extraction efficiency was Obtained. This difficulty was the presence of a yellowish, fat-like layer at the junc- tion of the two layers in the flask, and was encountered when Skellysolve "A" was substituted entirely for ethyl ether and when mixtures of ethyl ether and Skellysolve "A" were used which contained less than 50 per cent ethyl ether. This condition is also believed to be due to the trapping of fat by undissolved protein, the yellow color being due to a greater prOportion Of fat in the pro- tein mass. The reason that this condition should exist only when less than 50 per cent ethyl ether is present in the mixture is probably because the solvent is immiscible with alcohol under these conditions. Immiscibility of the solvent with alcohol has a profound effect on the efficiency of the extraction since the extraction efficiency is due to the complete penetration of the solvent to all particles of fat, any substance which is entirely immiscible with the transporting agent would likely be less 69 effective than ethyl ether. In the Mojonnier method of fat extraction, the alcohol acts as a transporting agent, carrying the solvent to the aqueous suspension of fat. ‘The ethyl ether though only slightly miscible with water, is readily brought in contact with the fat by the alcohol which is miscible with both water and ethyl ether. As indicated in the results, Skellysolve itself is completely immis- cible with alcohol. However, the miscibility Of alcohol and Skellysolve is improved by the addition Of ethyl ether, and a 50 per cent.mixture of ethyl ether and Skellysolve is completely miscible with alcohol, even when water or milk is included in the system. The greater quantity of ethyl ether required when water or milk is present is probably due to the additional volume through which the alcohol must be distributed. The water or milk serves to partition a certain amount of the alcohol, leaving a smaller amount remaining in the sol- vent phase. As small quantities Of alcohol reduce miscibility with Skellysolve to a greater extent than large quantities, it is lOgical that greater amounts Of other are required to keep the mixture in a miscible state. The volume of the lower layer of liquid in the flask is also affected by the miscibility of the reagents. In the process Of conducting the standard MOjonnier method, the volume of the lower layer after the addition Of ethyl ether is only 48.8 per cent as great as the volume of non-ethereal liquids which have been added. The difference correSponds approximately to the amount Of alcohol which has been added and it is likely that it is included in the upper layer at this P01Dt. In the case Of the substituted procedure, when Skellysolve is added, and the emulsion allowed to settle out, the volume Of the lower layer is even greater than the amount of milk, ammonia and alcohol which are thought to compose it, indicating that certain small amounts Of Skellysolve are retained in the lower layer. The presence Of small amounts of Skellysolve at this stage of the extraction procedure will cause an in- crease in the volume of the lower layer, even though the miscibility of the ether-Skellysolve mixture, alcohol, and water is complete. However, these small variations in the volume of the lower layer are of minor importance so long as miscibility is maintained. Upon the addition of the second solvent, petroleum ether, to the mix- ture in the standard Mojonnier procedure the volume of the lower layer is still less than the total volume of milk, alcohol and ammonia by approximately the amount of alcohol which has been added in the second extraction. Thus, petroleum ether does not occlude all of the alcohol or other non-ethereal sub- stances from the ethereal layer. In the case of the modified procedure, using 100 per cent Skellysolve, little change is affected in the volume of the lower layer by the addition of the second solvent, indicating that the Skellysolve does not permit the inclusion of alcohol or other liquids in the upper layer. Further evidence that some non-ethereal substances are present in the ether layer of the Mojonnier samples is indicated by the fact that in all trials, the Mbjonnier extract took a considerably longer time to become completely evaporated, and there was some indication that the boiling temperature of the Mojonnier extract was higher than that of the modified procedures. FUrther reasons for the inefficiency of the several Skellysolve modifi- cations may be revealed by the data obtained in the analysis of low~fat prod- ucts. These data show that a great difference exists between (a) the methods using ethyl ether as the first solvent and (b) the methods using mixtures of ethyl ether and Skellysolve ”A" as the first solvent. In the analysis of Skimmilk and powdered Skimmilk, both the mixture modification and the reduced quantity modification give similar results, in- dicating they are about equally efficient. HOwever, in the analysis of churned 71 buttenmilk, the 1:1 mixture procedure extracted only 55 per cent as much fat as the Mojonnier method, Whereas the 50 per cent reduced quantity procedure was 92.13 per cent efficient. It may be Speculated that these differences in results obtained by the two methods in the analysis of these products may be due to a difference in the composition of the products. Apparently there is some ether-soluble material in the buttermilk which is not extracted by the Skellysolve-ether mixture. It is known that churned buttermilk contains a relatively high pro- portion of phosphatides, and that skimmilk contains a relatively low percent- age. It is further known that these substances are rather efficiently ex- tracted by ether. It it were found that the phosphatides are not extracted by the Skellysolve-ether mixtures, then a good portion of the discrepancies found between these mixture methods and the Mojonnier would be explained. A final important consideration in the comparison of the Mojonnier method with these various modifications is the cost of the several reagents. The present price of ethyl ether is quoted at from $1.80 to $2.25 per gallon, and petroleum ether at $2.50 per gallon. In contrast, Skellysolve ”A" and "F" cost $0.65 per gallon. On this basis, the complete replacement of petroleum ether by Skellysolve results in a saving of 74 per cent f the cost of second solvents. The replacement of ethyl ether by a 1:1 mixture of ethyl ether and Skellysolve "A” results in a saving of 30 to 36 per cent, and use of the 3:1 mixture gives a saving of 12 to 19 per cent. The use of reduced quantity modi- fications have obvious effects on the cost of reagents. The availability of ethyl ether has been affected somewhat by war production demands, but similar circwmstances have not diminished the supply of Skellysolve. Naturally, the economic factor is of no consideration unless the efficiency of the procedure is satisfactory. 72 SUMMARY AND CONCLUSIONS This study was conducted for the purpose of deve10ping modifications of the Mojonnier method which employ less expensive and more easily available solvents. The modifications which were attempted are as follows: (a) The complete substitution of Skellysolve ”F" and Skellysolve "A" for the petroleum ether of the Mojonnier method. (Skellysolve "F" is essentially a mixture of hexanes with a boiling range of 390 - 48° C., and Skellysolve ”A" is essen- tially normal pentane with a reported boiling point of 33° - 33.50 C.) (b) The use of various mixture of ethyl ether and Skellysolve "A” in place of the ethyl ether of the Mojonnier method. (c) The use of reduced amounts of ethyl ether used alone as first solvent. The graphic relationships between these modifications and the MOJonnier control procedure are presented in Figures I - Ilof the discussion. The most successful modification is the one in which the petroleum ether is replaced by Skellysolve "A” or ”F”. This substitution gave results which were practically identical to those obtained with the Mojonnier proced- ure. Mixtures of Skellysolve "A" and ethyl ether were substituted for ethyl other as first solvent in prOportions of 1:1, 8:2, and 3:1 parts of ethyl ether and Skellysolve "A” respectively. The extraction efficiency was found to be directly related to the prOportion of ethyl ether in the mixture, but all of the modifications gave values which were lower than those obtained by the Mojonnier method. In the analysis of summer milk the procedures involving the 1:1, 5:2 and 3:1 mixtures resulted in the following respective discrepan- cies from the Mojonnier method: 0.04761 0.0037 per cent, 0.0298$0.0025 per 75 cent and 0.0245330.0038 per cent. Analysis of milk produced during the win- ter time gave even lower recovery of the fat, the difference between the 1:1 'mixture and the MOjonnier procedure amounting to 0.08961:0.0049 per cent. The finer dispersion of fat caused by homogenization, and the presence of greater amounts of fat resulted in greater differences between these modi- fications and the Mojonnier control method. With the 1:1 mixture the follow- ing products resulted in the accompanying average differences: homogenized milk, 0.0713310.0036 per cent; evaporated milk, 0.2783 per cent; ice cream, 0.1857110.0307 per cent. The 3:2 mixture gave the following average differ- ences: homOgenized milk, 0.04521=0.014l per cent; evaporated milk, 0.06151: 0.0375 per cent; ice cream, 0.12892t0.0496 per cent. The 3:1 mixture re- sulted in the following average discrepancies: homogenized milk, 0.01653: 0.003? per cent; evaporated milk, 0.0461 per cent; ice cream, 0.0641 per cent. The third modification utilized smaller quantities of ethyl other than are used in the standard Mojonnier procedure. Reduction in the quantity of ethyl ether by 50 and 25 per cent gave summer extraction values which were re- apectively 0.02491:0.0037 and 0.0104 per cent lower than the Mojonnier method. Winter conditions affected these methods to the extent that the 50 per cent reduced method resulted in a difference of 0.0661 per cent. Homogenized milk and high-fat products gave greater discrepancies which for the 50 per cent re- duced method are as follows: homOgenized milk, 0.0259330.0019 per cent; evap- orated milk, 0.1009 per cent; and ice cream, 0.1662 per cent. In the case of the 25 per cent reduction these differences are: homogenized milk, 0.01643? 0.0026 per cent; evaporated milk, 0.0355 per cent; and ice cream, 0.0698 per cent. Both the reduced and mixture modifications are less efficient in the analysis of winter milk. This is believed to be caused by a change in the con- 74 stituents of the milk, particularly protein, resulting in a condition whereby the solvent extracts the fat with.more difficulty. The seasonal effect was nullified in both the reduced and mixture techr niques by either of two expedients: (a) Lengthening of the shaking period from 30 seconds to 60 seconds and (b) Application of heat at 100° F. Just prior to analysis, tOgether with the use of additional ammonia in the second extraction. Three extractions were also found to reduce the error in the 1:1 mixture by 0.0255 per cent. In the analysis of milk which had been preserved with corrosive subli- mate or milk which had been subjected to low temperature storage, neither the mixture nor the reduced procedure was affected to any greater extent than the MOjonnier2method. However, in the analysis of milk from cows on fat-feeding experiments, the 1:1 mixture procedure extracted 0.9664 per cent less fat than the Mojonnier, indicating that feeding conditions have a pronounced deleterious effect on the ability of the ethyl ether-Skellysolve mixture to extract fat. In the substitution of a solvent in fat extraction procedures, the efficiency of the solvent is doubtless affected by its miscibility with alco- hol. Skellysolve "A”, not being miscible with alcohol, will not be efficiently dispersed through the milk, thus resulting in incomplete extraction. The addition of ethyl ether to Skellysolve ”A? increases its miscibility, and a mixture of 1:1 prOportions of ethyl ether and Skellysolve ”A" is completely miscible with alcohol. The results indicate that the efficiency of this miscible mixture is greatly improved over the efficiency of the Skellysolve "A? alone. The miscibility of the reagents not only affects the extraction ef- ficiency of the method, it plays a role in governing the volume of the lower, non-ethereal layer of liquid. When Skellysolve is one of the reagents, the upper layer contains no aqueous or alcoholic liquid and the result is that the volume of the lower layer is slightly greater than normal. However, in the standard Mojonnier procedure, the volume of the lower layer is decidedly lower than normal, indicating that it does not include all of the non-ethereal solutions which are present. Therefore, certain of these aqueous or alcoholic substances are included in the ethereal layer. Analysis of Skimmilk and buttermilk by the method involving substitu- tion of a 1:1 Skellysolve-ether mixture for ethyl ether and by the method using 50 per cent of the normal quantity of ethyl ether revealed wide differ- ences in the extracting ability of these methods. When compared to the Mojon- nier in the analysis of churned buttermilk, the 50 per cent reduced method was found to be 97.5 per cent efficient whereas the 1:1 mixture modification was found to be only 54.6 per cent efficient. In a similar comparison to the Mojonnier method conducted on Skimmilk, the modification using the 1:1 mix- ture of ethyl ether and Skellysolve “A" extracted 92.13 per cent of the fat. 1. 2. 4. 5. 6. 8. 9. 10. ll. .14. 15. Adorjan, Iosef 1906 Ardenghi, A. 0. 1938 Bartlett, I. M. 1909 Bieda 1906 76 REFERENCES Apparatus for Fat Determination Landw. Vers. - Wes. 9, 1063-66. Chem. Absts. l, 1151. Review of the Current Standard Tests for Fat Determin- tion. Rev. facultad. cienc. quim. 13, 131-34. Chem. Absts. 33, 5076. A.Method for the Determination of Fat in Sweetened Condensed Milk. U. S. Bur. Chem. Bul. 122, 152. French Patent. Chem. Absts. 3, 256. Biesterfeld, C. H. and Evanson, 0. L. Estimation of Fat in Condensed 1917 Milk and Milk Powder. Jour. Ind. Eng. Chem. 9, 1111-13. Bigelow, W. D. and Fitzgerald, F. F. A Review of Methods for the 1915 Brodrick-Pittard, 1916 Broeman, F. C. 1913 Cobleigh, W. M. 1916 Cochrane, C. B. 1905 Analysis of Evaporated Milk. Natl. Canners Assn. Res. Bul. 5, 23. N. A. waiters Metteilunv zur Methodik der fettbestim- mung im.Kase. z. Nahr. Benussm. 32, 354-8. Determining Fat in Sweetened Condensed Milk. 3.01er Ind. mg. Chem. 4, 6729 A Modified Babcock Test Applied to Hemogenized Cream. Am. Food Jour. 11, 457. Estimation of Fat in Infant and Invalid Foods. JOUI‘. Am. Chem. SOC. 27’ 906-7. Consortium far elektrochemische Industrie New Fat Solvents and Extrac- 1907 Croll, W. L. 1913 Dyer, B. 1899 E1 chloff, P. 1907 1909 tion Media. Chem. Ztg. 31, 1095-96. Chem. Abets. 2, 717. An Investigation to Determine the Accuracy of a Modi- fied Meigs Method for the Quantitative Determination of Fat in Milk. Biochem. Bul. 2, 509-18. Minutes of a Meeting of the Society of Public Analysts. Analyst 24, 174-5. Studien fiber die Adamssche Fettbestimmungs methods. and Grimmer. Abegeandertes verfahren zur Bestimmung des Fettgehalts nach Rose-Gottlieb in Milch und Molkerel- produkten. Milchw. Zentbl. 6, 114-121. 16. 17. 18. 19. 20. 24. 25. 29. 77 Fleming, R. S. Netes on the Composition and Analysis of Dessi- 1912 cated Milk and Cream. Jour. Ind. Eng. Chem. 4, 543-4. Gerber's Co. German Patent. Chem. Absts. 2, 1641. 1906 Gottlieb, E. Estimation of Fat in Milk. Landw. Vers.-Stat. 40, 1906 1-27. Abst. Jour. Chem. Soc. 62, 549-550. Greisheim Elektron Chemical Works Studies on the use of Carbon 1905 Tetrachloride as a Solvent. Chem. Ztg. 31, 326-328. Chem. Absts. 1, 1778. Grossfeld, J. A New Method for the Determination of Fat in Cheese. 1923 Pharm. Weekblad. 60, 1069-1078. Chem. Absts. 17, 3897. Guigagn, E. A_New Method for Fat Analysis in Evaporated Milk or Other Concentrated Dairy Product. Jour. Soc. Chem. Ind. 28, 738-40. Hackman, C. The Estimation of Fat in Cheese. Milch. Forsch. 3, Hak, S. and Gregg, H. 'fiber die Refraktometrische Methods der Fett- 1902 bestimmung in Milch nach Prof. Dr. Woolney. Hilch. Ztg. 31, 433. ' and Klykksen fiber Fettbestimmung in kondensierter Milch. 1907 Z. Nahr.-Genussm. 13, 338-45. Harding, E. and Perkin, G. A.Rapid, Accurate Gravimetric Method for 1914 the Analysis of Evaporated Milk and Powdered Milk. Jour. Ind. Eng. Chem. 6, 131. A Rapid, Accurate Gravimetric Method for 1914 the Analysis of Ice Cream. Jour. Ind. Eng. Chem. 5, 843-5. Harris, C. P. The Use of O-dichlorobenzine in the Determination of Fat by Change in Specific Gravity. Jour. Ind. Chem., Eng. Anal. Ed. 2 #1, 410-13. Chem. Absts. 24, 5519. Hesse, A. A Simple Method for the Determination of Fat in Butter. 1904 Z. Untersuch. Nahr. Genussm. VIII, 673. Abst. in Analyst 30, 86. Horsley, J. A Method for the Estimation of Fat in Milk. Chemical 1874 News, May 1874, 224. Abst. in Am. Jour. Pharm. 46, 479. Hertvet, J. Report of the Referee. Assoc. Off. Agr. Chem. Jour. 1915 1, 186. 31. ()3 N 0 CR 03 34. 40. 41. 46. Hertvet, J. 1917 1920 1926 Jeffe, E. 1932 Johnson, J. J. 1937 Keister, I. T. 1922 Koenig, F0 1913 Kohman, E. F. 1919 KrOpat, K. 1913 Kuhn, F. 1898 Kurz, O. 1923 Lamrestrasse 1906 Leithe, W. 1934 MOLEIIBD , ‘Bo Go 1908 Mandel, H. 1917 Meigs, A. V. 1882 78 Report of the Referee. 2, 238. Assoc. Off. Agr. Chem. Jour. Report of the Referee. 4, 201. Assoc. Off. Agr. Chem. Jour. Report of the Referee. Assoc. Off. Agr. Chem. Jour. 9, 255. A Study of the Use of Ethylene Glycoll as a Fat Sol- vent. Ann. Chim. applicata. 22, 426-433. Chem. Absts. 26, 6167. The Use of IsOprOpyl Ether in a Modified Mojonnier Fat Test. Jour. Dairy Sci. 20, 563-6. The Determination of Eat in Melted.Milk. Assoc. Off. Agr. Chem. 5, 176. Application of the Gum Tragacanth Method to the Analysis of Butter. Apoth. Ztg. 29, 223. Chem. Absts. 8, 2008. Jour. A Rapid Test for the Determination of Eat in Butter. Jour. Ind. Eng. Chem. 11, 38. Application of the Gum Tragacanth Method to the Analysis of Cheese, Cream, and Butter. Arch. Pharm. 252, 76-82. Chem. Absts. 8, 2434. Milch. Ztg. 27, 772. Cited by Fisher and Walts, Conn. Agr. Exp. Sta. Bul. 131. A Micro-Analysis for Fat. Chem. Absts. 18, 1718. Microchemie 1, 78-86. British Patents. Chem. Absts. 1, 921. A Pycnometer Method for the Determination of Fat in Dairy Products. Z. Untersuch. Lebensm. 67, 441-446. Chem. Absts. 28, 3490. A Comparison of Methods of Fat Extraction on Dried Milk. Analyst 33, 353. The Use of Benzene in Barthel's Method of Fat Extrac- tion. Svensk Kem. Tidskrift 29, 181-3. Chem. Absts. 12, 1402. A Method for Fat Determination in Dairy Products. Philadelphia Medical Times 2, 660. 79 47. Mitchell, L. C. and Alfend, S. A Lend-Gooch Method of Analysis of 1926 Butter. .Tour. Assoc. Off. Aer. Chen. 9, 209-20. 48. Mojonnier, '1'. and Troy, 11. 0. The Technical Control of Dairy Products. 1925 2nd ed. Pub. by Mojonnier Bros. Co., Chicago, Ill. 936 pp. illus. 49. Notbohm, I. r. and Anigerhausen, .T. Nachprdnmg der "Hansel-Methods 1910 you Dr. Wendler" zur Fettbestimnung in Milch. z. Nahr. Genussm. 20, 495-8. 50. Olson, G. A. A Conparison of Ether Extraction Methods in the 1906 Analysis of Cheese. Proc. Assoc. Off. Aer. Chem. 23. 51. Oven, A. V. A Review of Various Old and New Methods of rat Ex- 1915 thtiono “IChe mtble 43, 285-95. Chm. nets. 9, 1080. - 52. Patrick, G. E. Review of the Current Methods for let Determination 1913 in Dairy Products. U. 8. Dept. of Agr. Bur. Chm. Cir. 66. 53. Popp, M. Observations on Rose's Method of Fat Deteminetion. 1904 Milch. Ztg. 33. Cited by Fisher and Melts, Conn. 381'. Exp. Sta. Me 151, 19250 54. Pepp, M. and Siegfeld Zeit. far Untersuchung Nehr. Genusa. 7, 772. 1904 Cited by Fisher and Melts, Conn. Aer. Expt. Sta. . Bul. 131. 55. Rmatedt, 0. Comparison of Ether Extraction and Carbon Tetra- 1909 chloride Extraction. Chan. Ztg. 33, 93-4. Chen. Abets. 3, 1097. 56. - Comparison of Rapid Methods and Gravinetric Methods 1912 of Fat Analysis. Z. Anger. Chan. 25, 754-8. Chm. Abstl. 6, 1938. 57. Richond, G. and Musgrave, W. A Special Extraction Method for Horlick'e 1908 Melted Milk. Phillipine Jour. Sci. 3, 87-90. 58. Richmond, H. D. Modification of the Rose-Gottlieb Method for the 1908 Analysis of Dried Milk. Analyst 33, 389. 59. Dairy Chemistry. 3rd ed. revised. Pub. by Cher1es . 1920 Griffin 8: Co. Ltd. London. 490 pp. illus. 60. and Rosier, C. H. The Estimation of Pat in Milk Using 1599 Petroleum Ether as s Solvent. Analyst 24, 172-175. 61. Rieter, E. A New Apparatus for the Rose-Gottlieb Test. Ann. Chin. 1909 Applicatu. 14, 54-57. Chan. Abate. 3, 1222. 62. 65. 66. 67. 68. 69. 71. 72. 74. 75. ’76. Rijneveld, L. 1911 Rohrig, 1905 Roop, J . 1918 Rose, B. 1888 Rupp, E. 1912 Rusting, 1898 80 Comparison of Banana, Dekker and Soxhlet Method of Fat unetione Phame weekblade 48’ 12040 Chas Abate. 6, 517. Apparatus for the Determination of Pat in Dairy Products. mre Z‘ite Nahl‘. Genus“. 9, 531-33e Cited. by Fisher and Walta, Conn. Agr. Expt. Sta. Bul. 131. The Use of an Acid Medium in the Rose-Gottlieb Test. Chen. Analyst 26, 21. The Determination of Pat in Milk. 1, loo-107. z. Angewande. Chem. Abst. in Jour. Chem. Soc. 54, 1135. and 11611.: A. bar on beachleunigtes Verfahven der Gottlieb- Scheringa, K. 1931 Schreib, ices Si chler, 1903 H. a. #1904 ainner, 1940 Supplee, 1922 G. W. C. and 3.1118, BO Sutton, R. W. 1925 6211881 , J. 1914 Taufel, I. end Standigl, L. 1930 Rose'schen Iettbeetimung in Milch. 25' 358'9. Z. Nahr. Genussm. The Use o'fI Tragacanth in Ether htraction Methods. Nederl. Ty dschr. Pham. 10, 163-70. Abst. In 0115. Centr. 2, 393. . ‘ Observations on the Weibull-aethan Method of rat Ix- tmctione Chas “Oekblade 28’ 6540 ch“. AbatBe 2‘, 1038. Comparison of the Rose-Gottlieb and Adams Methods of Fat EnrQCtiono :0”. Ohm. SOC. 54, 1135. German Patent. Chan. Abate. 2, 347. German Patent. Chan. Abate. 1, 2435. Methods of “817818 Of the A8800. Off. A315 Ch“. 551! Ode P'ube by ASSOC. Off. Aer. Chas, “Bantan, De 0. Fat Analysis of Milk Powders. Joure Mr! 8010 5. 39-500 A Modification of the Werner-Sclmid Test for Condensed Milka. Analyst 50, 17-18. Analyses of Human Milk. Chem. Zeit. 14, 1202-3. Abate 1n Joure Ghae Soc. 62, 18920 Conlpa A118”. son of Solvents Used in the Soxhlet Method. 1. Fettzg. 27, 127-29. CDC. Abate. 2‘, 41%. '77. 79. 80. 82. 83. 86. 87. Thomson 1905 Tilpe 1907 Van Lennep, D. B. P. and Rays, J’. 1912 Vizern and Guillot Comparison of Various Solvents. 1932 Vollrath, I. 1905 de Veal, D. C. 1922 leibull , M. 1907 Iendler, O. 1909 Hindi sch, Karl 1908 W01]. 9 Fe We 1906 Ioosnsn, R. W. 1897 81 Fat Extraction from Milk. Tidskrift fur Landoeconomi 1905 #2, 101. Abst. in Analyst 30, 128. Dehydration of Ether. Abate. 2, 715 Chem. Ztg. 31, 1107. Chan. Use of Trichlorethylene in Extrac- tion. Chan. Weekblad. 9, 654. Chan. Abate. 6, 3477. Chemie and industrie. Special No. 580, 1-March. Chan. Abate. 26, 3687. Observations on the Use of Carbon Tetrachloride as a Solvent. Chm. Ztg. 31, 398. Chan. Abate. 1, 1778. A New Method for Fat Analysis. Tijschr. Vergelijk. Geneskunde 8, 138. Ohm. Abate. 17, 441. line Beobachtung bei der Gottlieb'schen Methode der Pettbeeti-ung. Z. Mahr. Genussm. 17, 443-5. - U. S. Patent. Chan. Abate. 4, 2528. Ueber das Sindcidver rahren and dae Salverfahren sur Iettbestimung in der Milch. Milch. Zentbl. 5, 344-52. Report to the Assoc. Off. Agr. Chen. Pros. Asso. Off. A810 Chem. 230 A Modified Schnid Process. Analyst 22, 91. APPENDIX 82 Table I. Effect of Replacing Petroleum Ether with Skellysolve "F” when a Slow Shaking Procedure 18 Used Difference Difference between Triplicates Trial Fat Obtaingg between Mojonnier Modified SK}? Product No. Mojonnier Modified Methods Maximum Average Maximum Average percent percent percent percent percent percent mrgent 14111: 1 4.1052 4.0709 0.0343 0.0142 0.0094 0.0155 0.0110 2 3.9294 3.7999 0.0295 0.0038 0.0025 0.0191 0.0127 3 3.3459* 25.8086" 0.0373“ 0.0532* 0.0384* 0.0270 0.0145 4 5.5315 5.4852 0.0454 0.0280 0.0185 0.0457 0.0304 5 3.5212 3.5129 0.0083 0.0151 0.0151 0.0157 0.0097 5 4.4425 4.4043 0.0382 0.0189 0.0125 0.0093 0.0052 7 3.5383 3.5155 0.0218 0.0183 0.0122 0.0183 0.0122 8 3.5599* 3.5388* 0.0311* 0.0390! 0.0254? 0.0190 0.0125 9 5.3737 5.3391 0.0345 0.0247 0.0247 0.0179 0.0109 10 3.8291 3.8277 0.0014 0.0313 0.0208 0.0225 0.0150 11 3.2542 3.2259 0.0283 0.0070 0.0045 0.0251 0.0157 12 3.8980“ 3.8505* 0.0375* 0.0348“ 0.0232* 0.0175 0.0115 13 3.4434 3.4077 0.0357 0.0238 0.0158 0.0084 0.0055 14 5.5720 5.5335 0.0384 0.0208 0.0138 0.0471 0.0314 15 5.5550 5.5457 0.0093 0.0200 0.0133 0.0308 0.0205 15 4.3273* 4.2783* 0.0490* 0.01581" 0.0280“ 0.0141 0.0094 17 2.5710" 2.5458“ 0.0252" 0.0687" 0.0454“ 0.0154 0.0102 18 4.4745 4.4557 0.0189 0.0145 0.0130 0.0288 0.0155 19 4.4702“ 4.4475* 0.0227“ 0.0343" 0.0225“ 0.0138 0.0092 20 3.7040 0.0259 0.0055 0.0043 0.0312 0.0218 3.5781 *Not included in averages. 83 Table II. Effect of Replacing Petroleum Ether with Skellysolve "F” when a Vigorous Shaking Procedure 18 Used Difference Difference between Duplicates Product Trial Fat Obtaippg between . __ No. Mojonnier Modified SKF Methods Mojonnier Modified $3 percent percent peppent percent Apercent Milk 21 3.4927 3.9388 0.0039 4 0.0107 0.0065 22 4.0904 4.0955 0.0051 0.0154 0.0110 23 2.9748 2.9710 0.0038 0.0002 0.0032 24 2.9813 2.9775 0.0038 0.0003 0.0027 25 2.4156 2.4144 0.0012 0.0050 0.0045 26 3.3841 3.3775 0.0066 0.0151 0.0153 27 4.1977 4.1851 0.0126 0.0125 0.0165 28 3.8953 3.8909 0.0044 0.0201 0.0118 29 3.4867“ 3.5012* ”-0.0145“ 0.0370“.' 0.0204 30 3.7953 3.7803 0.0150 0.0058 0.0098 31 4.9186 4.9200 -0.0014 0.0194 0.0105 32 4.2985 4.2896 0.0089 0.0038 0.0203 33 4.2233 4.2067 0.0166 0.0127 0.0050 34 3.4957 3.5068 -0.0111 0.0038 0.0035 35 1.0637 1.0623 0.0014 0.0010 0.0232 3555- ‘ 35 3.8844 3.8758 0.0086 - - genized ZMilk 37 3.8765 3.8886 -0.0121 0.0105 0.0075 38 3.7816 3.7687 0.0129 0.0138 0.0073 39 3.7668 3.7644 0.0024 0.0121 0.0094 40 3.3829 3.8642 0.0187 0.0111 0.0057 41 3.7255 3.7200 0.0055 0.0098 0.0048 42 4.5581 4.5535 0.0046 0.0072 0.0169 43 4.5607 4.5501 0.0106 0.0145 0.0245 84 Table II. Continued. Difference Difference between Duplicates Product Trial Fat thgineg between No. Mojonnier Modified SCF Methods Mojonnier Modified SEF percent percent .percent gppercent pegcenp 44 3.8770 3.8743 0.0027 0.0065 0.0037 Homo- 45 3.8923 3.8843 0.0080 0.0046 0.0045 genized IMilk 46 3.7059 3.7118 0.0059 0.0289 0.0018 47 3.7865 3.7736 0.0129 0.0142 0.0056 48 3.9909 3.9902 0.0007 0.0110 0.0096 49 4.0324 4.0274 0.0050 0.0095 0.0013 50 4.5121 4.5095 0.0026 0.0032 0.0045 Ice 51 10.4080" 10.2292* 0.1798“ 0.0125 0.2570* cream ' Mix 52 10.4457" 10.3507" 0.0850* 0.0299 0.1472* 53 10.2014* 10.0395* 0.1519* 0.0238 0.1259* 54 10.9576 10.9248 0.0328 0.0209 0.0314 55 11.0834 11.0544 0.0290 0.0069 0.0037 56 0.6503 0.5373 0.0130 0.0046 0.0124 57 10.4772 10.5279 -0.0507 0.0124 0.0208 58 10.5534 10.5189 0.0345 0.0353 0.0241 60 10.5835 10.6071 -0.0236 0.0316 0.0097 61 10.7220 10.6567 0.0653 0.0240 0.0691 62 10.7479 10.7079 0.0400 0.0085 0.0234 rated Milk 64 7.8833 7.8924 -0.0091 0.0510 0.0044 65 7.8784 7.9377 -0.0593 0.0170 0.0004 66 7.8764 7.8747 0.0017 0.0430 0.0118 *Not included in averages. 85 Table III. Effect of substituting Skellysolve ”A" Entirely for Ethyl Ether when Petroleum Ether Is Used as the Second Solvent —*7 Difference Difference between Triplicatgs Product Trial Fat Obtained between.___Mojonnier Modified SKA No. Mojonnier Modified m Methods Maximum Average Maximum Average per cent _nercen§ pgpcenp percent percent percent percent Milk 1 3.6556 3.3070 0.3486 0.0086 0.0086 0.0650 0.0433 2 3.4911 3.2520 0.2391 0.0104 0.0104 0.0638 0.0425 3 5.2919 4.9188 0.3731 0.0036 0.0036 0.2829 0.1886 4 5.2567 5.0523 0.2044 0.0069 0.0046 0.0234 0.0156 5 .2328 2.1532 0.0796 0.0067 0.0044 0.0322 0.0181 6 4.5530 4.4117 0.1413 0.0303 0.0202 0.0424 0.0282 7 4.5472 4.2777 0.2695 0.0115 0.0076 0.0006 0.0006 8 3.9432 3.6006 0.3426 0.0023 .0'0015 0.0713 0.0475 9 3.9187 3.6083 0.3104 0.0234 0.0156 0.0364 0.0243 10 4.0106 3.4665 0.5441 0.0177 0.0118 0.2945 0.1963 11 3.9970 3.6280 0.3690 0.0108 0.0072 0.1235 0.0823 12 4.0659 3.7779 0.2880 0.0352 0.0234 0.2385 0.1590 13 4.0899 3.6966 0.3933 0.0248 0.0165 0.0309 0.0462 14 4.3803 4.0509 0.3294 0.0370 0.0243 0.1538 0.1538 15 4.0315 3.8715 0.1600 0.0142 0.0094 0.0490 0.0327 16 3.9817 3.8622 0.1195 0.0138 0.0092 0.0087 0.0087 17 4.0227 3.6854 0.3373 0.0029 0.0019 0.1326 0.0884 18 4.0866 4.0241 0.0625 ‘0.0143 0.0095 0.0198 0.0132 19 4.0250 3.9488 0.0762 0.0013 0.0013 0.0353 0.0235 20 4.0181 3.9099 0.1082 0.0188 0.0125 0.0282 0.0423 21 3.1437 3.0478 0.0959 0.0362 0.0241 0.0520 0.0350 22 3.9127 3.7867 0.1260 0.0397 0.0265 0.0416 0.0277 86 Table III. Continued. Difference Difference between Triplicatps Product Trial Fat Obtained between. Mojonnier Medified SKA No. 'Mojonnier~Modified SKA Methods Maximum Average Maximum.Average Apercent _percent _percept percent percent percent percent Milk 23 5.2875 5.2479 0.0396 0.0233 0.0155 0.0395 0.0263 24 3.4747 3.3105 0.1642 0.0126 0.0084 0.0166 0.0111 25 4.1090 3.8703 0.2387 0.0217 0.0144 0.1144 0.0763 26 4.1655 4.0904 0.0851 0.0193 0.0128 0.0707 0.0471 27 4.0739 3.9454 0.1285 0.0243 0.0162 0.0223 0.0145 28 3.9078 3.8894 0.0184 0.0010 0.0010 0.0532 0.0354 29 4.1646 4.0872 0.0773 0.0127 0.0085 0.0427 0.0288 30 3.7715 3.6844 0.0871 0.0300 0.0200 0.0202 0.0134 31 4.1457 4.0923 0.0534 0.0146 0.0097 0.1316 0.0874 Mix 32 13.8208 0.6084 13.2124 - 0.00114 0.1548 0.1032 33 11.6327 0.4798 11.1529 - 0.0048 0.1164 0.0776 34 11.8101 0.9054 10.9047 - 0.0212 0.1388 0.0925 87 NHOH.0 0000.0 0000.0 0050.0 0HH0.0 0H¢¢.0 0HOH.¢ Homa.¢ 0 0000.0 00H0.0 u 0Ho0.0 0000.0 0050.0 «mvn.¢ ¢m¢u.¢ 0 0000.0 Hm¢o.o 0000.0 0000.0 0m00.0 0000.0 0¢NH.¢ ¢p¢H.¢ 0 0000.0 50H0.0 00N0.0 000H.o 0000.0: 0000.0 0000.0 0000.0 0 0000.0 0000.0 a «000.0 m¢00.0 0000.0 000H.v vao~.¢ A Md“: pqeqmmm $000960 $000909 ammonoq‘ pdmohmq‘ paoenom. pdmohmn pqoonoqw headache: headaches“ poanQOnosfl .02 4% 000.3002 013.3005 Amandoho: a wonhauofi omnhacofi .mwpwouaqsn nomkvop monopohawn no powhpwfi ma 00000600“ «mm 883% e328“ 93.308: 3980 350.8 emefieppo pee .3 e323: e930: “3883: a as. eofioz 833326 5... ofiomefioxm 23 no 8229.8 ego H9282 .3 5393.0 2:. mfioeeom obH oanwa 88 Thble V} The Use of 6 Proportion of Skellysolve ”A” and Ethyl Ether when Skellysolve ”I" Is Used as the Second Solvent Fat Obtained Difference D fference between Du licates Product Trial Mojonnier 50%.Mixture between .MOjonnier 55;‘Mixture Tested No. Methods percent percent percent percent _percent ‘5112 1 4.0550 3.9895 0.0635 0.0252 0.0505 2 4.0075 5.9644 0.0429 0.0059 0.0159 5 4.0029 5.9501 0.0528 0.0176 0.0160 4 5.5426 5.4645 0.0783 0.0054 . 0.0050 5 5.7079 5.6465 0.0616 0.0164 0.0246 6 5.6156 5.5990 0.0166 0.0162 0.0111 7 5.9627 5.9564 0.0263 0.0056 0.0059 8 5.9557* 5.5900* 0.0457* 0.0560* 0.0228 9 3.8923 5.6565 0.0558 0.0057 0.0045 10 5.7652 5.7465 0.0549 0.0045 0.0207 11 5.5565 5.4995 0.0388 0.0012 0.0060 12 5.4155 5.5604 0.0551 0.0197 0.0028 15 5.8161 5.7750 0.0411 0.0077 0.0295 14 _ 5.9746 5.6972 0.0776 0.0018 0.0521 15 3.7696 5.7550 0.0546 0.0158 0.0260 16 5.6160* 5.6250* 0.0090* 0.0460* 0.0041 17 5.7204 5.6640 0.0564 0.0127 0.0025 18 3.7489 5.7266 0.0225 0.0289 0.0200 19 5.5959 5.5455 0.0504 0.0027 0.0517 20 5.9207 3.8267 0.0940 0.0024 0.0456 21 4.6261 4.5565 0.0698 0.0011 0.0422 22 5.7525* 5.7459* 0.0066* 0.0525* 0.0255 25 5.6965 5.6595 0.0566 0.0057 0.0221 Table V. Continued; Fat Obtained Product Trial i0jonnier 50%IM1xture between ‘Mojonnier 50%2Mixture Tested No. Methods __ percent ppercent _percent percen§_ _p9rcent Milk 24 4.1680 4.1156 0.0524 0.0125 0.0587 25 3.7087 3.6427 0.0660 0.0211 0.0629 26 3.7874 3.7553 0.0321 0.0076 0.0147 27 3.8607 3.8129 0.0478 0.0088 0.0088 28 3.3508 3.3219 0.0289 0.0030 0.0049 29 3.7435 3.6919 0.0516 0.0018 0.0408 30 3.7983 3.7452 0.0531 0.0093 0.0318 31 3.3977 3.3604 0.0373 - 0.0243 32 3.7674 3.7102 0.0572 0.0066 0.0032 33 4.1500 4.0890 0.0610 0.0037 .0.0103 34 4.0454 4.0104 0.0350 0.0092 0.0505 35 3.7137 3.6377 0.0760 0.0014 0.0312 36 3.4863 3.4347 0.0516 0.0024 0.0454 37 3.7737 3.7089 0.0648 0.0054 0.0119 38 3.9620 .9122 0.0498 0.0147 0.0732 IEmmOgen- 39 3.8763 3.4476 0.4287 0.0082 0.2475 :13: 40 4.8409 4. 5985 0. 2424 0.0077 0.1363 41 4.2625 4.1669 0.0956 0.0037 0.0484 42 4.1921 4.0646 0.1275 0.0096 0.0076 43 4.0855 4.0158 0.0697 0.0038 0.0187 44 4.0886 4.0286 0.0600 0.0176 0.0035 45 3.9683 3.8398 0.1285 - 0.0195 46 4.0177 3.9045 0.1132 0.0048 0.0177 J 1 Difference Difference betflgggkggpgicates 1 90 Table 7. Continued. FatrObtained Difference Difference between Duplicgtes Product Trial MOJonnier 50% Mixture between Mojonnier 50% Mixture Tested No. Methods _f 4percent percent percent Aperpent percentu‘ Hbmogen— 47 4.1448 4.0351 0.1097 0.0045 0.0558 iii: 48 4.2107 4.0888 0.1215 0.0095 0.0092 EVapor- 49 7.8839 7.8642 0.0297 0.0346 0.0058 :11: 50 7.9378 7.6544 0.0654 0.0256 0.0160 51 8.2958 8.1957 0.1001 0.0100 0.0037 52 6.5666* 6.2126* 0.1562* 0.0556* 0.0162 53 8.2679 8.1333 0.1346 - 0.0473 54 8.3244 8.1897 0.1347 - 0.0147 Ice 55 12.9125 12.2866 0.6259 0.0241 - Cream Mix 56 9.9319 9.6986 0.2333 0.0307 - 57 11.4014 11.2179 0.1835 0.0370 0.0555 58 11.6112 11.2785 0.3327 0.0302 0.0310 59 12.2395 11.8559 0.3836 0.0068 - 60 10.1296 9.9882 0.1414 0.0290 0.0165 * Net included in averages. 91 ¢Hm0.0 mmmo.o #mHo.o nm¢o.o nnm0.0 mnmo.o nNNo.o «Hoo.o o powhpNfi can nnmo.o mmmo.0 mmw0.o ¢b¢o.o ammo.o ano.o mmoo.o OHSPHfi2,om\om 6600.0 mmoo.o omoo.o enao.o mono.o neoo.o *meno.o omoo.o ¢§Ho.o headache: mopwowansnhmmezpon codenameaa weaves mm¢o.o mmpo.o ammo.o ammo.o mnmo.o mnno.o omno.o #eoo.o ¢nao.o aboo.o nowo.o emeo.o mHmo.o ombo.o mmno.o «000.0 n¢Ho.o obmno.o doapomnpxm onecooonm chm uwaemom avenues noospop coconomman .memepobw a“ cousaona p02 * 6566252 om\on Hedm.» nomm.n mama.» m Harm.» moam.» whom.n m Hmmn.¢ bnon.¢ mam¢.¢ b boeo.¢ nnHo.¢ bmpo.¢ m unbn.n mobn.n nmon.n n mmea.¢ HnNH.¢ mmba.¢ v *mnmn.n *eoon.n 4¢m&¢.m n n¢np.n memm.n nom§.n w Hmmn.n 8mm¢.n fimmn.n H Maw: aofipowpvmm onscooonm cospos cum Heaswom .oz copmoe nmaqeonos define aoeeonm .Mmeqfimpno «we eqopaom eeooom 56 gm. opaommafloem apes ewes.on¢.nonpm Hmepm one :d: opaomhaaoxm no opapxaa.aua a Mo wnoHpownpxm omega non? hho>ooom new mom: pooghm .HP oanee 92 Table VII. Use of a 2:3 ProPortion of Skellysolve ”A" and Ethyl Ether when Skellysolve ”F" Is Used as the Second Solvent Fat Obtained Difference Difference between Duplicates Product Trial M0jonnier 40/60 Mixture between Mojonnier 40/60 Mixture Tested No. Methods .percent percent .percent percent percent Milk 1 3.4181 3.3628 0.0553 0.0136 0.0172 2 3.4070 3.3813 0.0257 0.0078 - 3 3.5017 3.4813 0.0204 0.0154 0.0024 4 4.1251 4.1010 0.0241 0.0139 0.0089 5 3.4533 3.4027 0.0506 0.0197 0.0121 6 5.0084 4.9725 0.0359 0.0144 0.0119 7 3.8200 3.7764 0.0436 0.0094 0.0077 8 3.8047 3.7940 0.0107 0.0228 0.0108 9 1.7224 1.7171 0.0053 0.0067 0.0131 10 3.2284 3.1962 0.0322 0.0093 0.0321 11 4.3395 4.3283 0.0112 0.0070 0.0094 12 4.1610 4.1187 0.0423 0.0154 0.0478 15 4.0957“ 4.0559* 0.0616* 0.0512“ 0.0152 14 3.9775 3.9537 0.0238 0.0004 0.0147 15 3.5332 3.4972 0.0360 0.0134 0.0226 Ebmogen-16 3.9562 3.8991 0.0571 0.0087 0.0113 iéfig 17 3.6098 3.5724 0.0374 0.0127 0.0322 18 3.9363 3.9121 0.0242 0.0128 0.0150 19 3.9930 3.9422 0.0508 0.0075 0.0156 20 3.9097 3.8638 0.0459 0.0068 0.0349 21 3.6218 3.5657 0.0561 0.0005 0.0320 93 Table VII. Continued. Fat Obtained Difference Difference between Duplicates Product Trial Mojonnier 40/60 Mixture between Mojonnier 40/60 Mixture Tested No. Methods percent percent percent percent _percent ___ EVapo- 22 7.9292 7.8715 0.0557 0.0523 0.0433 £11: 23 7.8992 7.8897 0.0095 0.0053 0.0193 24 7.9929 7.9141 0.0788 0.0330 - 25 7.9981 7.9088 0.0893 0.0100 0.0079 26 8.0247 7.9503 0.0744 0.0051 0.0551 27 6.0772“ 6.0122‘“ 0.0650'“ 0.0705“ 0.0529 Ice 28 11.6374 11.3287 0.3087 0.0237 0.0243 Cream Mix 29 11.4239 11.3452 0.0787 0.0449 0.0930 30 10.1281 10.0741 0.0540 0.0055 0.0240 31 11.1384 11.0058 0.1326 0.0318 0.0844 32 12.9728 12.8695 0.1033 0.0320 0.0163 33 9.9552 9.8593 0.0959 0.0222 0.0283 * Not included in averages. 94 Table VIII. The Use of a 1:3 Proportion of Skellysolve ”A” and Ethyl Ether when Skellysolve “F” Is USed as the Second Solvent .Fat Obtained Difference Difference between Duplicates Product Trial Mojonnier 25/75 Mixture between Mojonnier 25/75 Mixture Tested No. Methods 11. <_pergent pergept percent percent percent Milk 1 5.4575 5.4020 0.0555 0.0045 0.0141 2 4.4562 4.4557 0.0205 0.0161 0.0026 5 5.4054 5.3784 0.0570 0.0044 0.0200 4 5.1661 5.1510 0.0551 0.0172 0.0020 5 4.0646 4.0591 0.0257 0.0092 0.0197 6 5.7009 5.6620 0.0169 0.0067 0.0069 7 4.6107 4.5656 0.0269 0.0054 0.0150 6 4.6096 4.5920 0.0176 0.0154 0.0100 9 5.0447 5.0270 0.0177 0.0052 0.0077 10 4.5501 .5295 0.0206 0.0068 0.0096 Ebmogen-ll 4.7655 4.7609 0.0244 0.0069 0.0295 1143: 12 4. 7601 4. 7766 0.0015 0.0052 0.0065 15 4.1655 4.1404 0.0229 0.0100 0.0216 14 4.0976 4.0650 0.0146 0.0066 0.0150 15 4.7642 4.7625 0.0219 0.0066 0.0555 16 4.6590 4.6564 0.0026 0.0217 0.0045 17 4.6766 4.6570 0.0196 0.0091 0.0022 16 4.5271 4.5025 0.0246 0.0009 0.0516 Ice 19 10.1626 10.1079 0.0549 0.0109 0.0907 Cream ‘Mix 20 10.0225 9.9514 0.0909 0.0141 0.0549 21 11.0174 10.9709 0.0465 0.0256 0.0027 Evapo- 22 7.9042 7.6577 0.0665 0.0556 0.0510 rated Milk 25 7.9576 7.9180 0.0596 0.0159 0.0069 24 7.9551 7.9006 0.0525 0.0165 0.0175 Table IX. The Use of a 1:1 Proportion of Skellysolve ”A" and Etlwl Ether when Skellysolve ”A" Is Used as the Second Solvent 95 Fat Obtairied Diffgrgrice D fference between licate‘L Product Trial'Mojonnier 50/50 MixtdEE‘ between Mojonnier 50/50 Mixture Tested No. Methods percent __ percent___ percent percent percent Milk 1 2.5552 2.2646 0.0706 0.0056 0.0042 2 5.9295 5.6915 0.0562 0.0049 0.0095 5 5.4471 5.4091 0.0560 0.0001 0.0146 4 .2105 4.1996 0.0109 0.0204 - 5 4.7514 4.6772 0.0542 0.0064 0.0529 6 5.7664 5.7506 0.0556 0.0054 0.0515 7 5.5526 .2756 0.0570 0.0049 0.0120 8. 5.5576 5.2751 0.0625 0.0152 0.0191 9 4.3569 4.5205 0.0564 0.0185 0.0270 10 5.6669* 5.6622* 0.0067* 0.0557* 0.0106 Homogen-ll 3.8785 3. 7897 0.0888 0.0060 0.0611 .951: 12 3.8468 ' 3.7881 0.0567 0.0051 0.0622 15 5.6572 5.7795 0.0779 0.0044 0.0465 14 5.6515 5.7920 0.0595 0.0122 0.0140 15 5.6420 5.7606 0.0614 0.0119 0.0105 16 5.6551 5.7665 0.0666 0.0050 0.0166 17 5.6456 5.7797 0.0659 0.0045 0.0550 16 5.6515 3.7838 0.0675 0.0126 0.0116 19 5.6414 5.7712 0.0702 0.0021 0.0141 20 3.7870 5.7104 0.0766 0.0205 0.0065 Ice 21 10.5592 10.4765 0.0629 0.0295 0.0098 Cream ‘Mix 22 11.4071 11.2427 0.1644 0.0095 0.0765 96 Table 111 Continued. Fat Obteined Difference Difference between Duplicetes#fl IProduct Trial Mojonnier 50/50 Mixture between .Mojonnier 50/50 Mixture 'Pested No. Methods Percent _percent percent percent percent iIce 23 11.4912 11.2585 0.1327 0.0136 - Cream 121x 24 10.4172 10.4005 0.0167 0.0257 0.1240 25 11.1567 10.8980 0.2587 0.0091 0.0368 26 10.6666 10.5455 0.1211 0.0356 0.0341 27 10.7184 10.5310 0.1874 0.0522 0.0254 28 12.3937 12.2030 0.1907 0.0136 - 29 11.4711 11.0544 0.4167 0.0229 0.0065 Evepo- 30 8.0659 7.8111 0.2548 0.0224 - rated IMilk 31 7.9310 7.6313 0.2997 0.0089 0.2911 32 8.0445 7.7376 0.3069 0.0381 0.1062 33 7.9228 7.7378 0.1850 0.0050 0.0946 34 7.9234 7.5784 0.3450 0.0234 0.0330 Butter— 35 0.5700 0.2322 0.2978 0.0039 0.0245 milk 36 0.6059 0.3220 0.2839 0.0090 0.0058 Skim- 37 0.1725 0.1654 0.0071 0.0046 0.0056. milk 38 0.0636 0.0478 0.0258 0.0012 0.0038 39 0.1873 0.1712 0.0161 0.0004 0.0098 * Not included in averages. 97 Table X. The'Use of a 2:3 Preportion of Skellysolve ”A” and Ethyl Ether when Skellysolve "A" Is USed as the Second Solvent Fat Obtained Difference Difference between Duplicates Product Trial Mojonnier 40/60 Mixture between Mojonnier 40/60 Mixture Tested No. Methods Apercent p$percent percent percent Apercent Compos- 1 3.3945 3.3382 0.0563 0.0141 0.0345 ite Milk Samples 2 3.4925 3.4396 0.0529 0.0005 0.0051 3 4.8728 4.8041 0.0687 0.0165 0.0371 4 3.3523 3.2818 0.0705 0.0058 0.0291 98 Table XI. The Effect of'Using 50 percent of the Standard Quantity of Ethyl Ether when Petroleum Ether Is Used as the Second Solvent Fat Obtained Difference Difference between Duplicates Product Trial Mojonnier' Modified between Mojonnier Modified Tested No. Methods Apercent percent percent ppercent percent Milk 1 3.3885 3.3419 0.0466 0.0018 0.0252 2‘ 3.6131 3.5695 0.0436 0.0035 0.0363 3 3.8442 3.8154 0.0288 0.0077 0.0032 4 3.1951 3.1600 0.0351 0.0046 0.0218 5 3.7357 3.6982 0.0375 0.0212 0.0465 6 3.7109 3.6924 0.0185 0.0041 0.0233 7 3.5461 3.6034 0.0427 0.0041 0.0610 ‘ 8 3.3375 3.3141 0.0234 0.0008 0.0079 9 3.3450 3.3191 0.0259 0.0132 0.0118 Hbmogen-lo 5.0136 4.9644 0.0492 0.0148 0.0130 ized Milk 11 3.8595 3.8188 0.0407 0.0043 0.0445 12 3.9678 3.8800 0.0878 0.0105 0.0480 13 3.8467 3.8217 0.0250 0.0032 0.0109 14 3.8349 3.8182 0.0167 0.0070 0.0346 15 3.8060 3.7769 0.0291 0.0069 0.0202 16 3.8116 3.7926 0.0190 0.0173 0.0295 17 3.8068 3.7778 0.0290 0.0107 0.0174 18 3.4716 3.4401 0.0315 0.0025 0.0161 rated Milk 20 7.9088 7.8529 0.0559 0.0194 0.0319 21 7.8730 7.8322 0.0408 0.0043 0.0099 Ice 22 10.9978 10.9263 0.0715 0.0165 0.0653 Cream Mix 23 10.5079 10.4828 0.0251 0.0426 0.0079 24 10.6556 10.6460 0.0096 0.0416 0.0363 Butter 25 0.6518 0.5315 0.0203 0.0099 0.0004 Milk 99 Table XII. The Effect of Using 75 percent of the Standard Quantity of Ethyl Ether when.Petroleum.Ether Is Used as the Second Solvent Fat Obtained Difference Difference between Duplicates Product Trial Mojonnier’ Modified between Mojonnier Modified Tested No. Methods , __ percent Apercent percent percent percent Milk 1 3.9102 3.8835 0.0267 0.0107 0.0391 2 3.5409 3.5354 0.0055 0.0028 0.0071 3 3.8536 3.8380 0.0156 0.0091 - 4 3.8236 3.8290 0.0054 0.0010 0.0103 Ice 5 10.5470 10.6297 0.0173 0.0110 0.0251 Cream Mix 6 10.4758 10.4762 0.0004 0.0227 - 7 9.6708 9.6167 0.0541 0.0225 0.0168 Table XIII o The Effect of Using 50 per cent of the Standard Quantity of Ethyl Ether when Skellysolve 7A" 16 Used as the Second Solvent 100 Fat Obtained Difference Difference between Duplicates Product Trial Mojonnier .Modified between Mojonnier Modified Tested No. Methods __7 percent Apercent percent percent percent Milk 1 4.2025 4.1819 0.0206 0.0168 0.0058 2 3.9932 3.9772 0.0160 0.0099 0.0071 3 3.7237 3.7132 0.0105 0.0049 0.0030 4 3.7106 3.6866 0.0240 0.0227 0.0054 5 8.5030 8.4739 0.0291 0.0099 0.0050 6 1.9246 1.8942 0.0304 0.0139 0.0247 7 3.6712 3.6769 0.0057 - 0.0039 8 3.9946 3.9643 0.0303 0.0090 - Homogen- 9 3.8336 3.8132 0.0204 0.0186 0.0280 ized Milk 10 4.4970 4.4637 0.0333 0.0141 0.0177 11 3.8615 3.8370 0.0245 0.0006 0.0211 12 4.4918 4.4647 0.0271 0.0114 0.0213 13 4.2957 4.2769 0.0188 0.0255 0.0218 14 5.1840 5.1640 0.0200 0.0272 - 15 4.2130 4.1824 0.0306 0.0009 0.0118 16 4.1719 4.1400 0.0319 0.0178 0.0369 Evapo- 17 7.9136 7.8890 0.0246 0.0012 0.0210 rated Milk 18 7.8934 7.8480 0.0454 0.0468 0.0783 19 8.1695 7.9368 0.2327 0.0170 0.0003 Ice 20 10.5924 10.4749 0.1175 0.0128 0.0425 Cream Mix 21 10.1332 9.9684 0.1648 0.0347 0.0323 22 10.9042 10.6880 0.2162 0.0178 0.0348 Table XIV. Fat Obtained Ether when Skellysolve ”A” Is Used as the Second Solvent 101 The Effect of Using 75 percent of the Standard Quantity of Ethyl Difference “Difference between.Duplicates Product Trial Mojonnier JModified between Mojonnier Modified Tested No. Methods percent percent Apercent ppercent p_percent Milk 1 3.7283 3.7005 0.0278 - - 2 3.6534 3.6483 0.0051 0.0045 0.0002 3 6.2704 6.2738 0.0034 0.0122 0.0114 4 3.7218 3.7105 0.0113 0.0055 0.0096 Homogen- 5 3.8610 3.8362 0.0248 0.0283 0.0039 Mil: 6 3.8508 3.8376 0.0132 0.0258 0.0247 7 3.8609 3.8513 0.0096 0.0189 0.0050 8 3.8631 3.8436 0.0195 0.0044 0.0002 9 3.8640 3.8565 0.0075 0.0098 0.0018 10 3.8736 3.8550 0.0186 0.0129 0.0112 11 3.8572 3.8533 0.0039 0.0177 0.0144 12 3.8670 3.8234 0.0436 0.0052 0.0304 13 3.8619 3.8498 0.0121 0.0076 0.0039 14 3.7698 3.7584 0.0114 0.0064 0.0023 Evapo- 15 7.9832 7.9655 0.0177 0.0119 0.0154 rated Milk 16 7.9367 7.8876 0.0491 0.0058 0.0092 17 7.9192 7.8803 0.0389 0.0024 0.0059 Ice 18 10.0206 9.9338 0.0868 0.0066 0.0500 Cream Mix 19 10.6556 10.6205 0.0351 0.0416 0.0853 20 10.5079 10.4204 0.0875 0.0426 0.0112 102 Table XV. Comparison of the Mojonnier Method with the Modified Methods which Used 1:1 Mixture of Ethyl Ether and Skellysolve ”A” and 50 per cent less than the Standard Amount of Ethyl Ether:in the Analysis of Winter Milk Difference Trial Fat Obtaiped pp,‘ between Difference between Dpplicates No. Mojonnier 1:1 Mixture Methods Mojonnier 1:1 Mixture A_percent percent percent percent ppercent 1 3.9947 3.8953 0.0994 0.0099 0.0162 2 3.8862 3.7874 0.0988 0.0039 0.0156 3 4.3816 4.3003 0.0813 0.0030 0.0005 4 3.9929 3.8407 0.1522 0.0181 0.0175 5 4.0018 3.8665 0.1353 0.0052 0.0205 6 3.7663 3.6360 0.1303 0.0018 0.0265 7 3.5762 3.4634 0.1128 - 0.0241 8 3.8334 3.7435 0.0899 0.0028 0.0188 9 4.0045 3.8608 0.1437 0.0085 0.0363 10 3.0848 2.9805 0.1043 0.0221 - 11 4.9493 4.8244 0.1249 0.0024 0.0229 12 4.1541 4.0782 0.0759 0.0033 - 13 4.1979 4.0886 0.1093 0.0073 0.0156 14 3.8604 "3.7922 0.0682 0.0185 0.0112 15 3.9905 3.9207 0.0698 0.0052 0.0287 16 4.0509 3.9858 0.0651 0.0026 0.0041 17 3.6855 3.6169 0.0686 0.0183 0.0360 18 3.9027 3.8396 0.0631 0.0082 0.0204 19 2.5316 2.4560 0.0756 0.0027 0.0045 20 3.3035 3.2333 0.0702 0.0013 0.0036 21 2.8356 2.7545 0.0811 0.0055 0.0082 22 3.6838 3.5710 0.1128 0.0062 0.0009 23 3.5595 3.5043 0.0552 0.0002 0.0351 24 3.8869 3.8316 0.0553 0.0029 0.0303 25 3.7385 3.6633 0.0752 0.0189 0.0393 26 3.7556 3.6741 0.0815 0.0159 0.0192 27 3,4030 3.3523 0.0507 0.0000 0.0122 28 3.3806 3.3315 0.0491 0.0026 0.0090 29 3.9283 3.8634 0.0649 0.0161 0.0275 30 3.8932 3.7888 0.1044 0.0227 0.0062 31 3.9206 3.8004 0.1202 0.0030 - 32 3.8718 3.7924 0.0794 0.0195 0.0370 33 3.9561 3.8674 0.0887 0.0065 0.0069 103 Table IV. Continued —‘—_— Difference Trial Fat Obtained between Difference between Dpplicates No. Mojonnier' Reduced 50%» Methods Mojonnier Reduced 50% pepcent" percent p_p§rcent percent __pe;gent 1 4.7596 4.7081 0.0515 0.0052 0.0042 2 2.5623 2.5318 0.0305 0.0051 0.0219 3 3.6503 3.5088 0.0415 0.0040 0.0205 4 4.0072 3.9684 0.0388 0.0118 0.0104 5 3.7509 3.7110 0.0399 0.0054 0.0462 6 3.8553 ‘3.8293 0.0260 0.0025 0.0064 7 3.8086 3.6879 0.1207 0.0098 0.0328 8 3.8988 3.7973 0.1015 0.0064 0.0291 9 3.7549 3.6477 0.1072 0.0135 0.0068 10 3.5948 3.4869 0.1179 0.0047 0.0136 11 4.9167 4.8124 0.1043 0.0038 0.0281 12 4.0344 3.9761 0.0583 0.0136 0.0140 13 3.5938 3.5508 0.0430 0.0172 0.0095 14 3.3943 3.3248 0.0695 0.0095 0.0111 15 1.9704 1.9281 0.0423 0.0312 0.0272 16 5.6465 5.5926 0.0539 0.0016 - 17 5.0725 4.9798 0.1073 0.0065 0.0030 18 3.5826 3.4980 0.0846 0.0117 - 19 3.4995 3.4189 0.0806 0.0208 0.0028 20 3.6391 3.5302 0.1089 0.0129 0.0092 21 5.5798 5.4596 0.1202 0.0094 0.0396 22 3.9947 3.9869 0.0078 0.0099 0.0070 23 3.8862 3.8528 0.0334 0.0039 0.0100 24 4.3816 4.3358 0.0458 0.0030 - 25 3.9929 3.9502 0.0427 0.0181 0.0056 Table XVI. Comparison of the Mojonnier’Method with a Modified Method whicthtilizes a 1:1 Mixture of Ethyl Ether and Skellysolve ”A" and a Method Employing the Same Mixture with a 60 Second Shaking Period 104 Fat Obtained Modified with Trial Deviation from Mojonnier Modified with Product No. ‘Mojonnier Modified 60 sec.shaking MOdified 60 sec.ehaking __percent percent percent percent percent Milk 1 3.6855 3.6169 3.5530 0.0686 0.0325 2 3.9027 3.8396 3.8590 0.0631 0.0437 3 2.5316 2.4560 2.4886 0.0756 0.0430 4 3.3035 3.2333 3.2449 0.0702 0.0586 5 2.8356 2.7545 2.8255 0.0811 0.0101 6 3.6838 3.5710 3.5666 0.1128 0.0172 7 3.5595 3.5043 3.5339 0.0552 0.0556 8 3.8869 3.8316 3.8456 0.0553 0.0413 9 4.0018 3.8665 3.9480 0.1353 0.0538 10 3.7663 3.6360 3.7197 0.1303 0.0466 11 3.8334 3.7435 3.7562 0.0899 0.0772 Table XVII. Fat Obtained Comparison of the Mojonnier Method with a Modified Method which Employs Reduced Quantities of Ethyl Ether and a Method Using the Same Modification with a 60 Second Shak- ing Period 105 Deviation f;9m.M01onnig; Trial Modified mfif Modified with Product No. Mojonnier Modified 60 sec.Shaking Modified 60 sec. Shaking percent percent percent percent percent Milk 1 4.7596 4.7081 4.7145 0.0515 0.0451 2 2.5623 2.5318 2.5365 0.0305 0.0258 3 3.5503 3.5088 3.6201 0.0415 0.0302 4 4.0072 3.9684 3.9697 0.0388 0.0375 5 3.7509 3.7110 3.7297 0.0399 0.0212 6 3.8553 3.8293 3.8504 0.0260 0.0049 106 Table XVIII. Comparison of the Mojonnier Method with the Modified Method which Utilizes a 1:1 Mixture of Ethyl Ether and Skellysolve ”A” in the Analysis of Milk which Has Been Treated to 100° F. for a Few Minutes Immediately before Testing Fat Obtained ‘__ Dezjation from nggnnie; Trial Regular Modified Regular' Modified No. Mojonnier iMOdified plus Heat IModified plus Heat percept percent __percent percent percent A__ 1 3.7385 3.6633 3.6656 0.0752 0.0729 2 3.7556 3.6741 3.6992 0.0815 0.0564 3 3.4030 3.3523 3.3523 0.0507 0.0507 4 3.3806 3.3315 3.2804 0.0491 0.1002 5 3.9283 3.8634 3.8968 0.0649 0.0315 6 3.8932 3.7888 3.8238 0.1044 0.0694 7 3.9206 3.8004 3.8699 0.1202 0.0507 8 3.8718 3.7924 3.8345 0.0794 0.0372 Table XIX. 107 Comparison of the Mojonnier Method with.Two Modified Methods which‘Use Reduced Quantities of Ethyl Ether and a 1:1 Mixture of Ethyl Ether and Skellysolve ”A? in the Analysis of LoweFat Products Fathbtaingd Trial Mojonnier Mixture Reduced 50% Deviation from Mojonnier Mixture Reduced 50% Product No. percent percent percent percent percent Churned , 1 0.6517 0.6315 0.0202 Buttermilk 2 0.6281 0.6088 0.0193 3 0.6237 0.6175 0.0062 4 0.6281 0.3866 0.2415 5 0.5700 0.2821 0.2879 6 0.6060 0.3220 0.2840 Skimmilk 7 0.1725 0.1654 0.0071 8 0.0634 0.0478 0.0156 9 0.1873 0.1712 0.0161 Powdered 10 5.898 4.777 5.518 1.121 0.380 Buttenmilk 11 5.925 4.904 5.503 1.021 0.422 Powdered 12 0.770 0.524 0.0450 0.246 0.320 Skimmilk 13 0.7713 0.4695 0.5266 0.3018 0.2447 ‘77“ ‘F‘ i I 14:1 I.“ HICHIGQN STQTE UNIV. LIBRQRIES 31293006486553