HILDITCH AND JONES FATTY ACIDS AND COMPONENT GLYCERIDES ETC. 76 The Fatty Acids and Component Glycerides of some New Zealand Butters. BY T. P. HILDITCH D.Sc. F.I.C. AND EVELINE E. JONES M.Sc. (Read at the Meeting Febrzcary 6 1929.) WHEN a natural fat composed of a mixture of neutral glycerides is carefully oxidised in acetone solution with potassium permanganate all the unsaturated groups present are broken down into mixtures of lower fatty acids and semi-acidic glyceride derivatives of azelaic acid whilst fully-saturated triglycerides are left unaltered (Hilditch and Lea J . Chem. SOC. 1927 3106; Collin and Hilditch, J . SOC. Chem. Ind. 1928 47 261T). Consequently a natural mixture of fully-saturated and mixed saturated-unsaturated glycerides can be quantitatively resolved into a corresponding mixture of neutral and acidic products.This procedure not only leads to a simple method of ascertaining the proportion of fully-saturated glycerides in a fat but if they are present in reasonably large amounts permits them to be isolated in quantity in the pure condition; analysis of the mixed fatty acids combined in the whole fat and in the fully-saturated portion by the rnethyl ester distillation method (cf. Collin and Hilditch Zoc. cit.; Hilditch and Houlbrooke ANALYST 1928 53 2461 then furnishes considerable information of a semi-quantitative nature with regard to the mode of union of the fatty acids with glycerol in the original fat. This line of attack has been applied to the case of butter-fats the materials selected for investigation being three bulk samples of ordinary deliveries of New Zedand butter.It was necessary however to devote considerable attention to the determination of the composition of the mixed fatty acids since as is well known this is not an easy mixture to deal with on a quantitative basis. Indeed, the literature on the subject justifies this statement by its volume and also by the discrepant observations which it discloses. Whilst allowance should be made, especially in dealing with an animal fat (and perhaps above all with milk fats) for probable variations caused by differences in habit and feeding it seems unlikely that these suffice always to account for the varying data which have been put forward from time to time; this view is perhaps confirmed by the fact that the characteristics which are the analyst’s chief guide in the examination of butter (volatile soluble and insoluble acids iodine value saponification value refrac-tivity etc.) have not been found subject to variations of by any means so wide an order.OF EsTERs.-It is probably unnecessary for the purpose of this communication to refer specifically to more than one or two of the more recent analyses of the mixed fatty acids of butter commencing with that of Crowther and Hynd (Biochem. J. 1917,11 139). These workers converted the mixed acids into methyl THE DETERMINATION OF MIXED FATTY ACIDS BY FRACTIONAL DISTILLATIO 76 HILDITCH AND JONES THE FATTY ACIDS AND COMPONENT esters and fractionally distilled the whole; the fact that even the lowest-boiling fractions contained some methyl oleate indicates that the fractionation method employed was not highly efficacious.Holland and Buckley (J. Agric. Res. 1918, 12 719) and subsequently Holland and co-workers (ibid. 1923 24 365) have published data obtained by somewhat similar methods for a large number of butter-fats. The general results of both groups of workers were as follows:-Crowther and Hynd. Per Cent. Butyric acid . . . . 4.6 Caproic , . . . . 1.7 Caprylic , 1-3 Capric , . . . . 1.3 Lauric , 5.4 Myristic , 17.7 Palmitic , 16.0 Stearic , . . . . 3.7 Oleic , . . . . 48.3 Holland and co-workers . Per Cent. 2.2-4.2 (“by difference”) 1.3-2-4 06-1-0 1.2-2.0 4.6-7.7 16.6-22.6 64-22-9 (“by difference ”) 7.8-20.4 25.3-40.3 These results met with severe criticism from Channon Drummond and Golding (ANALYST 1924 49 311) and Elsdon (did.1924 49 423) mainly on the grounds that accuracy to the third place of decimals per cent. was apparently claimed and that the fractionation procedure had shown itself incapable of pro-ducing the binary mixtures of saturated esters necessary for arithmetical inter-pretation. Armstrong Allan and Moore (J. SOC. Cliem. Ind. 1925 44 63T) emphasised the necessity for attention to certain vital points in connection with the preparation and fractionation of the esters (notably the isolation and charac-terisation of fractions of individual esters) and claimed that with due precautions an accuracy to within a unit per cent. was attainable. We agree with the criticisms of the earlier attempts to apply the fractionation method quantitatively and with Channon Drummond and Golding in their statement that the procedure “ can yield very valuable information if its limitations are recognised”; but we cannot go so far with these authors as to affirm that “as an exact quantitative method it is of little value.” Obviously the strain upon the fractionation process should be lessened as much as possible by preliminary division of the acids into groups of varying character.Due attention to pre-liminary separations of this kind coupled with precautions in the fractional dis-tillation on the lines suggested by Armstrong Allan and Moore and subsequent workers render an accuracy of within a unit per cent. quite attainable. This is shown by numerous results published from several independent sources within the past two or three years.Whilst dealing with the fractionation method from a critical standpoint two matters of detail may be mentioned: (i) The isolation of absolutely individual esters as postulated by Armstrong, Allan and Moore is somewhat of a counsel of perfection when mixtures of small quantities of successive homologues are encountered ; in such circumstances we prefer to collect a series of small fractions covering the region in which an in-dividual member may be expected to predominate. This effectively minimise GLYCERIDES OF SOME NEW ZEALAND BUTTERS 77 any inaccuracy due to a possible error in the nature of the minor components of the mixture; the tables on pp. 80 81 illustrate the general composition of final fractions as determined by our present method of operation.(ii) It is our custom to collect the primary distillates in fractions as large as is consistent with the gradual rise in boiling-point and the general proportions of the mixture of esters known to be present. Each primary fraction is then submitted to refractionation as a separate entity; we do not in any circumstances add a subsequent primary fraction to the residue from the redistillation of its pre-decessor. This in our opinion makes the ultimate series of calculations more accurate whilst it avoids the necessity for the employment of “corrected” weights in the latter. In connection with the earlier data for the acids of butter-fat which we have discussed we have a further criticism to make-all unsaturation seems to have been calculated on the assumption that oleic acid alone is present.The data which we give in this paper (cf. p. 84) show definitely that the unsaturated acids present contain small quantities of an acid less saturated than oleic. We have confirmed this by examination of the hydroxystearic acids prepared from the acids of the fractions concerned by alkaline permanganate oxidation when although we did not succeed in isolating any pure tetrahydroxystearic acid the dihy-droxystearic acid produced melted somewhat indefinitely about 5” below the true melting-point of 9 10 dihydroxystearic acid from pure oleic acid. Further a very small trace (insufficient for recrystallisation) of brominated acids insoluble in ether was obtained from the “liquid” fatty acids of the butter; this substance blackened without definitely melting at 170-180”.We conclude that the un-saturated acids of butter-fat (like those of most other fats) contain small proportions of linoleic acid and even traces of linolenic acid. Allowance for this in the older analyses may reduce the proportion of un-saturated acids considerably in not a few cases (since calculated on iodine value, two molecules of oleic acid appear as one of linoleic acid) ; consequently the amounts of some of the saturated acids will be correspondingly low. In this connection Holland’s results with butter from cows fed on varying rations may be significant the “oleic acid” figures for example for cows fed respectively on a general ration on one including coconut oil and on one including soya bean oil were 30.8 29.5 and 45.8 per cent.whilst the corresponding palmitic acid figures (determined “by difference”) were 20.2 17.1 and 8.7 per cent. It is quite possible that the apparent increase in “oleic acid” and the extremely low (differential) percentage of palmitic acid in the butter after diet including soya bean or other semi-drying oil. is really due in part to the presence of varying proportions of linoleic acid. We are proceeding with the investigation of butters from cows whose food has included different types of fatty material and we hope thereby to throw further light on this matter. As regards Holland and Buckley’s figures for the volatile acids it may be remarked that good agreement exists between these and the values obtained i 78 HILDTICH AND JONES THE FATTY ACIDS AND COMPONENT the present work and also those arrived at recently by Virtanen and Pulkki (2.anal. Ckem. 1928 74 321) in the case of ten samples of Finnish butter. The latter included the following range butyric acid 3.1492 per cent. and caproic acid 1.4-2.1 per cent. It is permissible to conclude that the general proportions of the four lowest acids of butter are now known within comparatively narrow limits. It has been said that in the fractionation analysis of mixed fatty acids each fatty mixture should be considered on the lines of a separate problem from the point of view of general procedure. As a rule it is sufficient to effect a preliminary resolution by the lead-salt method into two groups: (i) “Solid” acids containing all the stearic or higher acids all but traces of palmitic and most of the myristic and lower saturated acids with small proportions of oleic acid; (ii) “Liquid” acids containing all C18 acids less saturated than oleic most of the oleic minor amounts of myristic and lower saturated acids but only traces of palmitic and no stearic or higher saturated acid.The “volatile” acids of butter render this particular case more complicated. We studied Crowther and Hynd’s method in the first instance and satisfied ourselves that the volatility of methyl butyrate rendered its application unsuitable, apart‘ from the tendency already discussed for methyl oleate to distil to some extent throughout the process. We have found it best to adopt in effect an intensive “Reichert-Meissl” procedure on a somewhat large scale.The fatty acids from 300-600 grms. of butter-fat are subjected to careful steam distillation for about four or five hours, i.e. sufficiently to ensure that the whole of the butyric and caproic acids have been removed. The acids volatile in steam are extracted by ether and fractionated as such mainly a t atmospheric pressure ; the difference in boiling-point between the successive homologues is sufficient to permit binary mixtures to be separated. The extracted aqueous liquors and the recovered distilled ether are titrated with alkali any acid present being calculated as butyric acid. The acids non-volatile in steam are recovered and weighed and a suitable portion is submitted to the lead-salt separation the resulting “solid ” and “liquid ” acids being converted into methyl esters and fractionated quantitatively in the usual way.The satisfactory concordance which we have obtained by this procedure is not likely to be due merely to repetition of the same sequence of processes on similar materials by the same manipulator since the following differences have been introduced: (i) The determination of mixed fatty acids in the whole butter-fats “A” and “B” was carried out on about 600 grms. of each fat and in “C” on 300 grms., whilst those of the fully-saturated glycerides of “A” were made on 200 grms., and of “B” on 140 grms. of material GLYCERIDES OF SOME NEW ZEALAND BUTTERS 79 (ii) In the lead-salt separations of the acids non-volatile in steam the “solid” acids in the case of “A” (56.7 per cent.) had an iodine value of 10-1 of “B” (67-6 per cent.) 23.2 and of “C” (54.4 per cent.) 11.8.Similarly whilst the “liquid” acids of “A” and “B” (iodine values respectively 82.7 and 80.9) consisted of the corresponding soluble lead salts from the original separation united with the mother-liquors from the recrystallisation of the separated lead salts in the case of “C,” the acids from the soluble salts (38-4 per cent. iodine value 83-1) and from the mother-liquors from recrystallisation (7.2 per cent., iodine value 62.1) were methylated and fractionated separately. DETERMINATION OF THE COMPOSITION OF THE FATTY ACIDS OF BUTTER-FAT.-The method adopted in these investigations will be illustrated by a detailed account of the experimental work on the New Zealand butter-fat “A.” The fat (approx.600 grms.) was hydrolysed by prolonged boiling with excess of alcoholic caustic soda after which as much alcohol as possible was removed from the soap by distillation. The flask containing the residual soap was then con-nected to a condenser for steam-distillation an efficient spray-trap being inserted between the exit from the flask and the,inlet to the condenser. In order to effect smooth distillation without frothing it was found best first of all to remove residual alcohol from the soap by cautious steam-distillation ; the soap solution was then cooled in the flask until it commenced to set to a jelly. At this point sufficient aqueous sulphuric acid was added to provide a slight excess of mineral acid after a,ll the alkaline base present had been neutralised and the fatty acids were then submitted to steam-distillation for about five hours; during this time about 5 litres of aqueous condensate was collected.After cooling the contents of the flask were extracted with ether and from the washed ethereal solution the mixed non-volatile fatty acids were carefully recovered and weighed (551.1 grms. mean equivalent 259.9 iodine value 42.2); the quantita-tive analysis of the non-volatile acids is described below. COMPOSITION OF STEAM-VOLATILE AcIDs.-The aqueous condensates were extracted five times with ether; the united extracts were dried over anhydrous sodium sulphate (which was subsequently washed with fresh dry ether to remove any adherent fatty acids) and the bulk of the ether was removed by distillation on the steam-bath.Portions of the ether-extracted aqueous solution (5345 c.c.) and of the recovered distilled ether (2394 c.c.) were titrated with standard alkali, and the acidity found calculated as bzltyric acid (2.10 grms. and 0.24 grm. in the respective liquids). The remaining volatile acids were slowly distilled from a Willstatter fraction-ation bulb and collected in small fractions the greater part of the distillation was carried out at atmospheric pressure but a moderate (water-pump) vacuum was used in the final stages. The first four fractions contained (diminishing) quantities of ether and came over below the boiling point of ut-butyric acid; the acid present in these was calculated as butyric and that in the remaining fractions was calculate 80 HILDITCH AND JONES THE FATTY ACIDS AND COMPONENT from the mean equivalents on the assumption that only binary mixtures were present : No.B.pt. Grms. O C . In aqueous solution In recovered ether 1 43.64 35/83 2 3.66; 83/86 3 1-96 881166 4 5.11 1661165 5 6.30 165/170 6 8.28 1701183 7 1.47 1421166 8 4-23 166/167 9 2-75 Residue Mean capric Pressure. equivalent. Butyric. Caproic. Caprylic. acids. Grms. Grms. Grm. Grm. Atmospheric # # #> 1 ) 8 ) Reduted 9 ) 96.4 97.8 111.8 119.0 164.4 2-10 0.24 0.97 0-75 1-69 4-64 4.27 2-03 4.84 344 0.17 1.30 3.68 0.66 0.66 2-10 Totals 19.67 10.46 1.20 2.10 COMPOSITION OF ACIDS NON-VOLATILE IN STEAM.-A portion of the non-volatile acids (303 grms.) was treated with lead acetate (212 grms.) in boiling alcohol (2430 c.c.).The deposited lead salts were separated and recrystallised from an equal volume of alcohol. The recrystallised lead salts were re-converted into fatty acids (solid acids S 170.3 grms.) whilst the alcoholic solutions were united and the dissolved lead salts contained therein also converted back to fatty acids (liquid acids L 130.3 grms.) : Presentin Mean Iodine Per Cent. 561.1 grms. equivalent. value. Solid acids S 56-7 312.2 269.2 101-5 Liquid acids L 43.3 238.9 267.3 82-7 Each group of acids was converted into neutral methyl esters and these were fractionally distilled in the usual way from a Willstatter bulb under high vacuum; in both cases it was only necessary to re-fractionate the first fraction: METHYL ESTERS OF SOLID ACIDS S.No. s1 Primary fractionation. Saponi-B.pt./ fication Iodine 1 mm. equivalent. value. No. I-L \ -Grms. O C . 62-71 70-130 254.2 2.6 ! S14 lii j E: S17 ' S18 Refractionation. Grms. 0-79 1-44 5-46 9.94 11-23 4.63 9-39 5-71 1 Saponi-B.pt./ fication Iodine 1 mm. equivalent. value. OC. 86-105 217.5 -106-116 231.6 -116-120 242.0 -120-123 248-2 1.2 123-130 265.6 1.2 130-132 264.1 1.9 132-136 270.2 2.7 Residue 277.6 9.8 48-88 S2 4807 130-132 273.2 6.2 S3 13-01 132-136 280.6 12.0 S4 13-22 135-136 287.6 18.0 S5 21.14 136-144 294.6 22.0 S6 884 144-148 296.4 21.6 S7 7-02 Residue 306.0 20.6 164.6 GLYCERIDES OF SOkaE NEW ZEALAND BUTTERS METHYL ESTERS OF LIQUID ACIDS L.Primary fractionation. Saponi-B.pt./ fication Iodine No. lmm. equivalent. value. t-A > I Gnns. OC. L1 33-61 44-130 232.8 34.1 Refractionation. 81 No. Grms. 0.64 1.06 2.34 2.84 3.67 4.48 6-20 9.62 B.pt./ 1 mm. OC. 39-60 60-66 66-09 69-92 106-114 1161 16 Residue 92-106 L2 16.56 130-140 288.0 76.2 L3 40.91 140-160 296.3 96.8 L4 12.67 160-162 297.7 98.0 L6 11-44 162-166 296.6 99.8 L6 8.99 Residue 326.6 108.4 29-66; Saponi-fication equivalent. 161.4 171.8 186.4 198.2 217.2 236.2 243.6 276.4 Iodine value. 8.6 8.6 14.2 12.6 12.1 22.7 26.0 67.7 124.07 The primary residues were saponified and any unsaponifiable matter present extracted by means of ether from the aqueous solution of the potassium salts.The fatty acids freed from unsaponifiable matter were recovered and their equivalents and iodine values re-determined. These figures which are more accurate than those for the small proportions of unsaponifiable matter present, have been employed in calculating the amount of the latter. The amounts of each ester present can now be calculated (in the esters of liquid acids L the assumption is made that the unsaturated q8 esters present in L1 and L2 had the same relative composition as the first pure Gs ester fraction, i.e. possessed an iodine value of 9643). METHYL ESTERS OF SOLID ACIDS S. S11. S12. S13. S14. S15. S16. S17. S18. Total. Grms. Gms. Grms.Grms. Grms. Grms. Grms. Grms. Gnns. Laurate 0.68 049 - - - - - 1.17 Myristate 0.11 0-96 5*46 7.66 6-03 0.97 0.16 - 20.92 Palmitate - - - 2.14 6.44 3.66 8.96 4-01 24-10 1.06 1.06 Stearate Oleate - - - 0.14 0.16 0.10 0.29 0.66 1.34 - - - - - - -Per cent. Per cent. as as * a 9 esters. a d s . 2-4 43.1 494 2.2 2.7 S1 S2. S3. S4. S6. S6. S7. 1-21 0.8 0.8 Laurate 1.27 - - - - - -Myristate 22.70 - - - - - - 22.70 13.8 137 Palmitate 26.16 4143 7434 4.46 2.01 0.30 - 82.18 49.9 49-9 Stearate 1.14 2-96 365 6.00 13.70 6-33 3.60 57.28 22.7 22.8 Arachidate - - - - - - 2.19 2.19 1.3 1.3 Oleate 1.46 3.48 1.82 2.77 6.43 2.21 1.83 18-99 11.6 11. 82 HILDTICH AND JONES THE FATTY ACIDS AND COMPONENT L11. GrmS. Caproate 0.08 Caprylate 0460 Laurate -Myristate -Oleate 0.06 Linoleate 0.01 Caprate -METHYL ESTERS OF LIQUID ACIDS L.L12. L13. L14. L16. L16. L17. L18. Total. Grms. Grms. Grms. Grms. Grms. Grms. Grms. Grms. 0.08 047 0.67 - - - - - 1-84 0.30 1.33 2.20 0.49 - - - 4-32 - - 0.27 2.63 2.30 1-72 0.44 7.36 1.13 2-08 2.43 644 0.08 0.30 0-32 0.39 0.92 1.22 6.81 9-09 0.01 044 0.05 0.06 0.13 0.18 0.84 1-32 - - - - - - -- - - -Per cent, Percent. as 85 f a 9 esters. wds. 03 6.2 14.6 24.8 19.0 30.7 4-4 L1. Caproate 0-09 Caprylate 2-08 Caprate 4-88 Laurate 8.32 Myristate 6-37 Palmitate -Oleate 10.28 Linoleate 1-49 Unsaponifiable -L2. -0.94 2-48 11.38 1-66 -L3. L4. L6. L6. - - - -36-74 10.89 9.60 6-87 6-17 1-78 1-84 1.31 0.81 - - -0.09 2.08 4.88 8.32 7.31 2-48 84.76 13-24 0.81 0.1 0.1 1.7 1.6 3.9 3.8 6.7 6.6 6.9 6*9 2.0 2.0 68.3 88.6 10.7 10.7 0.7 0.7 The whole of the experimental data for the original mixed fatty acids are then combined as follows: NEW ZEALAND BUTTER-FAT “ A.” Acids non-volatile in steam.Acid. Butyric Caproic Caprylic capric Lauric Myristic Palmitic Stearic Arachidic (?) . . Oleic Linoleic (Unsaponifiable) . . Volatile acids. Grms. 33-32 19-67 10.46 1.20 2-10 -Solid acids S. GrmS. 312:3 -2.37 42-77 166.80 71-04 4.14 36.18 -Liquid acids L. GrmS. 238.8 -0.16 3-86 9.16 16.78 13-97 4-77 - -163.87 26-80 1-64 Total. Gms. 684.42 19.67 10.61 6.06 11-26 18.16 66-74 180.67 71-04 4-14 200.06 28-60 1.64 Per cent.excluding. unsapom-fiable matter. 3.4 1.8 0.9 1.8 3.1 97 27.6 12.2 0.7 94.3 4.4 -The New Zealand butter-fats “B” and “C” were investigated by the same method except that in the case of “C” the quantity of fat employed was only about half of that used in the other instances. The final results are given in the next tables Acid. Butyric Caproic Caprylic Capric Lauric Myristic Palmitic Stearic Arachidic (?) Oleic Linoleic (Unsaponifiable) . . GLYCERIDES OF SOME NEW ZEALAND BUTTERS NEW ZEALAND BUTTER-FAT " B." Acids non-vohtile in steam. Volatile acids. Grms. 32.71 18.09 11-07 0.86 2-70 - - - - --I Solid acids S.Grms. 369-2 - - -1.10 7-82 48.86 162.64 49-12 6.60 94-1 7 -1 Liquid acids L. Grms . 178.2 - -3.73 7.71 14.93 12.23 0.14 -116.73 21.39 1.34 NEW ZEALAND BUTTER-FAT " C," Acids non-volatile in steam. Acid. Butyric' Caproic Caprylic Capric Myristic Lauric Palmitic Stearic Arachidic (7) . . Oleic Linoleic (Unsaponifiable) . b Volatile acids. Grms. 17.05 10.11 5-24 0.93 0.32 0.46 - - - - - - -r Solid acids S. Grms. 162.8 - - -1-60 20.10 88.96 29.60 2.50 20.14 - -1 Liquid acids L. Grms. 136.3 - -1-73 6.88 11.46 13.83 0.63 - -84.1 1 16.81 0.97 Total. Grms. 680.1 1 18.09 11-07 4.68 11.61 22.76 61.08 162.68 49.12 6-60 210.90 21.39 1.34 Total.Grms. 316.16 10.11 6-24 2.66 7-18 13.61 33.93 89.49 29.60 2.60 104.25 16-81 0.97 83 Per cent. (excluding unsaponi-fiable matter). 3.1 1.9 0.8 2.0 3.9 10.6 28.1 8 4 1.0 364 3.7 -Per cent. (excludin T unsapom-fiable matter). 3-2 1.7 0.8 2.3 4.3 10.8 28.4 9.4 0.6 33.1 6-4 -In the next table are collected the characteristics of each of the butter-fats, together with the final results of the fractionation analyses and the equivalents and iodine values of the mixed fatty acids and fats calculated from the latter figures : MIXED FATTY ACIDS OF NEW ZEALAND BUTTERS. "C" Sap. Equiv. . . . . . . . . . . . . 247.6 260.7 249.5 Iodine value . . . . . . . .. . . . 38-0 39.4 39.3 Reichert-Meissl value . . . . . . . . . . 25-4 26.8 26.4 Polenske value . . . . . . . . . . . . 1.9 2-3 2.1 Kirschner value . . . . 23.7 204 20.3 Butyric acid (calcul&d frdm &hner &he) . . 4.2 3.7 3.6 w A ## ## B 0* BUTTER-FAT 84 HILDITCH AND JONES THE FATTY ACIDS AND COMPONENT ##A*# ACIDS BY FRACTIONATION ANALYSIS Per Cent. Butyric . . . . . . . . . . 3.4 Caproic . . 1.8 caprylic . . . . . . 0.9 Capric * . . . . . . . 1.9 Myristic . . . . . . . . . . . . . . 9.7 Lauric . . . . 3.1 Palmitic . . . . . . . . . . . . 27.6 Stearid . . . . . . . . . . 12.2 Oleic . . . . . . . . . . 34.3 Linoleic . . . . . . . . . . . . . . 4.4 Arachidic ( i)’ . . . . . . . . 0.7 c6 B ** Per Cent. 3-1 1.9 0.8 2.0 3.9 10.6 28.1 8.6 1.0 36.4 3.7 46 c #’ Per Cent.3.2 1.7 0.8 2.3 4.3 10.8 28.4 9-4 0 5 33.1 6.4 CALCULATED MEAN VALUES (from fractionation analyses) : Sap. equiv. Fatty acids . . . . . . . . . 238.2 238.0 2374 Iodine value. Fatty acids . . . . . . . . 38.9 39.6 39.6 Glycerides . . . . . . . . 36.9 37.6 37.6 Glycerides . . . . . . . . 250.9 250.7 260.6 Consideration of these figures should be prefaced by a word as to the limits of accuracy qualitative and quantitative of our experiments. Since the majority of the fatty acids of butter have been definitely recognised for many years we have only formally identified our products in certain cases. We may point out in passing that individual esters when predominating in the distillates were readily recognised by their boiling-points at the pressure employed and that in particular, the greater part of the methyl palmitate present was obtained in fractions in which it was the main component and which crystallised in the well-defined form char-acteristic of this ester.Methyl stearate was also definitely recognised both as ester and in the form of stearic acid isolated therefrom. The small proportions of fatty acid calculated as ‘‘ arachidic” represent in-crements of a fatty acid of higher molecular weight than stearic (calculated after elimination of any accompanying unsaponifiable matter) ; the amount present was so small that it was not possible to obtain any fraction rich in this material, and we have not therefore been able positively to identify it.In these circum-stances we have calculated it in terms of the next even-member acid higher in the series than stearic i.e. as arachidic acid. The error if any is small owing to the minor amount present . With reference to the linoleic acid content of butter-fat we would point out that the presence of an acid more unsaturated than oleic is consistently borne out by the iodine values of the q 8 unsaturated esters obtained in fractionation of the liquid acids (cJ pp. 81 83). liquid ester fractions from the re-spective butters possessed the following characteristics : The major Sap. equiv. Iodine value. “A” . . . . . . 296.3 96.8 “B” . . . . . . 293.6 99.3 “C” . . . . . . 293.5 97-2 We experienced more difficulty than had been anticipated in preserving the freshly-distilled unsaturated q8 esters from atmospheric oxidation and it is con-sequently probable that the figures for linoleic acid are if anything on the low sid GLYCERIDES OF SOME NEW ZEALAND BUTTERS 86 (especially in butter-fat ‘‘ B ”).Absorption of oxygen and diminution in iodine value set in rapidly when these esters are stored and we have adopted the practice of placing such fractions in rubber-stoppered bottles in an atmosphere free from oxygen immediately they have been collected. A comparison of the data obtained (cf. table pp. 83,84) shows that the mean equivalents of the mixed fatty acids recalculated from the fractionation data, correspond fairly closely with those of the original fats whilst the calculated iodine values are slightly but consistently low probably for the reason just stated.As regards the figures for individual acids it appears probable from the results for “B” and “C” (two almost identical fats) that the mean error is only a few tenths per cent. except in the case of the saturated acids of highest molecular weight and the unsaturated acids. It may be repeated that we believe that this method of analysis is reliable to the nearest whole number (per cent.) (cf. ANALYST, 1927 52 253) and we are disinclined to lay much stress on fractional values. Nevertheless in the case of butter-fats the small proportions of the lower acids present unfortunately render this course necessary to some extent. The data for butyric-capric acids however are concordant in their relation-ship to the observed Kirschner values and are also within the comparatively narrow range for each acid assigned by Holland and Buckley (Zoc.cit.). The relation of the observed butyric acid content to that calculated on the assumption that the Kirschner value is a simple measure of butyric acid would seem to show that the latter registers interms of butyric acid about 15-20 per cent. more than is actually present in the fat. The unsaturated acids present are very similar in composition to those present in tallow and consist of oleic acid admixed with about 12-15 per cent. of linoleic acid; the total unsaturated acid content of the mixed fatty acids now studied was 38.541 per cent. The data for the higher saturated fatty acids in which so much variation occurs in the earlier literature are consistent in showing a content of 9-7-10-8 per cent.of myristic and 27.6-28.4 per cent. of palmitic acid; whilst the observed values for stearic acid are 895 9.4 and 12.2 per cent. We believe that the analyscs establish palmitic acid as the predominating saturated component of these butter fatty acids whilst they also show that myristic and stearic acids are each present to the extent of rather more than one-third of the weight of palmitic acid. Mitchell (ANALYST 1924 49 515) has recently obtained figures from which he concludes that the stearic acid content of butter-fats may range from practically nothing to 22 per cent. at least. I t is noteworthy that in the three samples now examined all otherwise closely similar there is a wider variation in the stearic acid figures than in those of any of the other saturated acids.From the fundamental standpoint it is perhaps more important to compare the relative molecdar q.uantities of the acids combined in a natural fat and therefore it is interesting to tabulate the number of equivalents of each fatty acid present in the three butter-fats and in tallow; for this purpose we have employed figures for a mutton tallow recently obtained in this laboratory and for an Australia 86 HILDITCH AND JONES THE FATTY ACIDS AND COMPONENT beef tallow as recorded by Armstrong and Allan (J. SOC. Chem. Ind. 1924 43, 216T) : Acid. Australian beef Mutton tallow. tallow. Per Cent. Per Cent. Myristic 2.0 4.6 Palmitic . . 26.6 24.6 Stearic .. 22.6 30.6 oleic 49.0 36.0 Linoleic . . - 4.3 Composition of Fatty Acids in Equivalents (per cent.). Tallows. A f \ Lzzz-- New Zealand Butter-fats. Acid. A. B. C. beef. Mutton, - - Butyric 9.2 8.4 8.7 Caprylic 1.4 1.3 1.4 Capric . . 2.7 2.8 3.1 Lauric . . 3.7 4.6 5-1 - -Myristic . . 10.2 11.0 11.2 2.4 6.4 Palmitic . . 25.7 26.2 26.3 28.3 26-3 Stearic . . . . 10.2 7.1 7-8 21.7 29-3 Arachidic 0.6 0.8. 0.6 Oleic and linoleic . . 32-7 33.9 32.4 474 39.0 - - Caproic . . . . 3-7 3.9 3-4 - - - -- -The following points may be noted: (i) One hundred molecules of the mixed fatty acids include in all cases about 26 molecules of palmitic acid (how far this is a general rule in the case of tallows is of course uncertain; the average of four other available analyses of tallows however shows 27.6 mols.of palmitic acid per 100 mols. of mixed fatty acids). (ii) The presence of the fatty acids lower than palmitic in butter-fat is balanced by a lower molecular content of G8 acids as compared with the tallows-the palmitic acid figure remaining about the same. (iii) About one-third of the molecules of the mixed fatty acids of the butters are those of oleic and linoleic acids. (iv) Butyric myristic and stearic acids are present in something approxi-mating to equimolecular proportions in the butter-fats examined. (v) The six lowest-molecular-weight acids of the butter-fats can be arranged as follows in pairs which correspond roughly in their respective molecular proportions : Butyric (CJ 84-9.2 per cent.and myristic (GJ 10-2-11-2 per cent. Caproic (CJ 34-3-9 per cent. and lauric (h) 3.7-501 per cent. Caprylic (c8)# 1-3-1.4 per cent. and capric (Go) 2.7-301 per cent. The circumstance that this approximate relationship may subsist betweell pairs of acids which respectively make up unit groups of 18 carbon atoms together with the comparative deficiency of butter-fat in stearic acid as compared wit GLYCERIDES OF SOME NEW ZEALAND BUTTERS a7 tallow may not be without significance from a biochemical standpoint; but we hesitate in the absence of a much wider series of analyses to do more than draw attention to what is perhaps only a coincidence. stated the procedure adopted has been to oxidise butter-fat until all unsaturated linkages have been converted into free acidic groups leaving only the original fully-saturated glycerides in the form of neutral compounds.The latter have then been freed as completely as possible from acidic products of oxidation and their weight noted; after which the composition of their mixed fatty acids has been determined precisely as in the case of the original butter-fat except that of course the lead-salt separation has been omitted unsaturated acids now being absent. The distribution of the acids in the fully-saturated and the mixed saturated-unsaturated glycerides has then been arrived at by comparison with the analytical data for the original butter-fats direct analysis of the fatty acids in the acidic oxidation products is impracticable owing to the impossibility at present of effecting quantitative removal of nonoic acid (one of the free acidic products) without con-current decomposition of the acidic glyceride compounds.Butter-fats “A” and “B” have been submitted to this treatment which will be briefly described before summarising the analytical figures which have been obtained. The fat (1 part) was dissolved in acetone (10 parts) and powdered potassium permanganate (4 parts) was added in small quantities at a time (with vigorous shaking at each addition) while the solution was gently boiled under a reflux condenser; boiling was continued for a short time after addition of the oxidant had been completed after which as much acetone as possible was removed by dis-tillation and treatment of the residue at 90-100” c. under reduced pressure.In order to minimise any chance of glyceride-hydrolysis in the alkaline solution (by potassium hydroxide possibly liberated from the permanganate) the resulting friable powder was mixed with powdered sodium bisulphite and then with water; the aqueous mixture was warmed and cautiously decolorised by the gradual ad-dition of dilute sulphuric acid with vigorous stirring; after cooling the solution was thoroughly extracted with ether . The united ethereal extracts after washing with water were shaken repeatedly with small quantities of dilute ammonia followed (when the whole was definitely alkaline) by thorough washing with water to remove as much of the organic am-monium salts as possible. The ammoniacal and aqueous liquors were re-extracted with ether to remove any neutral glycerides present in the emulsified condition, and all the ethereal extracts were then concentrated and the residual neutral product dried.If the iodine value of the neutral product was appreciable (e.g. above 04), the material was submitted to a repetition of the oxidation process. Eventually, crude neutral products with an iodine value of 0.3 or less were obtained; in this INVESTIGATION OF THE COMPONENT GLYCERIDES OF BUTTER-FAT.-AS alread 88 HILDTICH AND JONES THE FATTY ACIDS AND COMPONENT condition however they still possessed a small acid value due to the retention of traces of the difficultly-removable acidic glyceride compounds (e.g. of the type C,H5(0.C0.R),(0.C0. [CH,],.COOH) where R represents a higher saturated fatty acid radicle). They were therefore boiled in water to which dilute sodium carbonate was added until the whole was definitely alkaline to phenolphthalein.The aqueous layer (containing some emulsified neutral glycerides) was separated from the main fatty portion which formed a clear upper layer; the latter was boiled several times with water until the washings remained neutral. By this means 80 to 90 per cent. of the crude neutral product was obtained in the form of material of negligible acid value whilst ether-extraction of the united alkaline and aqueous wash-liquors furnished the remainder (10-15 per cent. of the whole) as a substance which still possessed a definite though low acid value. This acidity has been allowed for by assuming that it is due to the presence of acidic compounds of the same order as those removed by the sodium carbonate from the crude product (the free acidic compounds from the extracted alkaline wash-liquors having been isolated and their acid value determined).The data thus obtained are sufficient to determine the proportion of fully-saturated glycerides present (to within one per cent.) and it then only remains to accumulate sufficient of these (a minimum of 150 grms. is desirable) for accurate determination of their fatty acid composition according to the scheme given on pp. 78-82. FULLY-SATURATED GLYCERIDES OF BUTTER-FAT “ A.”-The fat yielded as a result of complete oxidation 33-6 per cent. of crude neutral products; oxidations were conducted on six batches of 100 grms. each in order to provide sufficient material for detailed analysis.On boiling the crude neutral product with dilute sodium carbonate as described above there were obtained: (a) 163.8 grms. completely neutral fat sap. equiv. 229-3 (acid value 0.4); (b) 22.9 grms. fat extracted by ether sap. equiv. 234.1 (acid value 6-4); (c) 12.5 grms. acidic material sap. equiv. 167.9 (acid value 211.2). Assuming that the acidic matter present in (b) has the same acid value as (c), the proportion of fully-saturated glycerides in the original fat is The value 31 per cent. of fully-saturated glycerides in butter-fat “A” has been used in the subsequent calculations GLYCERIDES OF SOME NEW ZEALAND BUTTERS 89 COMPOSITION OF THE FATTY ACIDS PRESENT IN THE FULLY-SATURATED GLYCERIDES OF BUTTER-FAT “ A.” Analysis of the combined neutral products (a) and (6) gave results which are summed up in the following table : Acids Per cent.Volatile non-volatile (excludiqg acids. in steam. Total. unsaporu-Grms. Grms. Grms. fiable Acid. 14.77 157.37 172.14 matter). Butyric 7.50 - 7.60 4.4 Caproic 5.79 - 5-79 3.4 Caprylic 0.75 1.37 2.12 1.2 Capric 0.62 3.83 4.45 2.6 Lauric 0.11 6.26 6-37 3.7 Myristic - 31.76 31-70 18.6 Palmitic - 78.78 78.78 45.9 Stearic - 34-99 34-99 20.3 (Unsaponifiable) . . - 0.38 0-38 -FULLY-SATURATED GLYCERIDES OF BUTTER-FAT “ B.”-After complete oxidation the original fat yielded 33-5 per cent. of crude neutral products which on boiling with dilute sodium carbonate gave : (a) 134.6 grms. completely neutral fat sap. equiv. 232.5 (acid value 0.3); (b) 9.5 grms.fat extracted by ether sap. equiv. 232.2 (acid value 15.4); (c) 19.3 grms. acidic material sap. equiv. 193.6 (acid value 104.9). The proportion of fully-saturated glycerides is therefore 33.5 (134.6 + 163.4 104.9 *’05) = 29-2 per cent. The value 29 per cent. of fully-saturated glycerides in butter-fat “B” has been used in subsequent calculations. COMPOSITION OF THE FATTY ACIDS PRESENT IN THE FULLY-SATURATED GLYCERIDES OF BUTTER-FAT “ B.” The final results of this analysis are summarised in the next table. Acid. Butyric Caproic Capric Lauric Myristic Palmitic stearic Arachidic caprylic Acids Volatile non-volatile acids. in steam. Grms. Grms. 12.2 116.3 5.35 -3-35 -3.50* 0.77 - 3.04 - 5.54 22.44 58-25 - 24.93 - 0.33 Total.Grms. 127.5 6.35 3.35 4-27 3-04 6-64 2 2 4 5825 24-93 0.33 Per cent. (exclueg unsapom-fiable matter). 4.2 2.6 3-3 2.4 4.3 17.6 46.7 19.6 0.3 * This figure is almost certainly high at the expense of correspondingly low values for caproic and capric acids owing to an unfortunately incomplete separation at the close of the fractional distillation of the volatile acids in this analysis 90 HILDITCH AND JONES THE FATTY ACIDS AND COMPONENT DISTRIBUTION OF THE FATTY ACIDS IN THE GLYCERIDES OF BUTTER-FAT.-The next tables show the general composition of 100 parts of the glycerides of butter-fats “ A ” and “B,” as indicated by the foregoing analyses. BUTTER-FAT “ A. ’’ Fully- Mixed saturated-Original saturated unsaturated glycerides fat.glycerides. (by difference). 100 31 69 Glycerol residue . . 5.1 1.7 3.4 Butyric acid Caproic Capric ,, Lauric J , Myristic ,, Palmitic ,J Stearic ,, Arachidic J J GpTlic J J 3.2 1.7 0.9 1.8 2.9 9.2 26.2 11.6 0.7 1.3 1.0 0.3 0.8 1.1 5.4 13.4 6.0 -1.9 0.7 0.6 1.0 1.8 3.8 12.8 5.6 0.7 (Molecular ratios). 22 6 3 6 9 17 50 20 2 Oleic , 32.6 - 32.6 116 Linoleic , 4.2 - 4.2 16 BUTTER-FAT ‘ r B.” Fully- Mixed saturated-Original saturated unsaturated glycerides fat. glycerides. (by difference). 100 29 71 ~~ ~~ Glycerol residue . . 5.1 1.6 3.5 Butyric acid Caproic ,, Caprylic J , capric >, Lauric ) J Myristic Palmitic ,, Stearic ), Arachidic ,) 2.9 1.8 0.8 1.9 3.7 10.0 26.6 8.1 1.0 1.1 0.7 0*9* 0.7 1.2 4.8 12.6; 5.4 0.1 1.8 1.1 1.2 2.5 5.2 14.1 2.7 0-9 (- 0.1) ~~ ~~ ~ ~ (Molecular ratios).20 9 7 13 23 66 9 3 -Oleic J 34.6 - 34.6 123 Linoleic J J 3.5 - 3.5 12 * This figure is almost certainly high cf. footnote to Table p. 89. On the whole the series of analyses for the two fats (which it will be re-membered differ mainly in that “A” has somewhat higher Reichert-Meissl and Kirschner values than “B”) are in fair accordance and show clearly: That the proportion of fully-saturated glycerides in these fats is of the order of 30 per cent.; (a GLYCERIDES OF SOME NEW ZEALAND BUTTERS 91 (b) That all the saturated acids are distributed more or less evenly throughout both the fully-saturated and the mixed saturated-unsaturated parts of the fat. Recent work on the Reichert-Meissl Kirschner iodine and other values of fractions of Irish butter-fat separated by chilling at various temperatures has led Arup (ANALYST 1928 53 641) to a similar conclusion. The general distribution of the saturated fatty acids is made clearer by com-paring the proportions of these acids present in the whole fat the fully-saturated part and the mixed saturated-unsaturated part. The tables which follow give these data for each fat both in the form of weight-percentages and of molecular percentages (the proportionate numbers of equivalmts of each acid present).RELATIVE COMPOSITION OF THE SATURATED FATTY ACIDS. Acid. Butyric Caproic Caprylic Capric Lauric Myristic Palmitic Stearic Arachidic (i) WEIGHT PERCENTAGES. Butter-fat “ A.” Whole fat. 5.5 2.9 1.5 3.1 5.1 . . 15.8 . . 45.0 . . 19.9 1.2 Fully-saturated part. 4.4 3.4 1.2 2.6 3.7 18.5 45.9 20.3 -Mixed saturated-unsaturated part. 6.7 2.5 1.7 3.6 6.4 13-1 44.3 19.4 2.3 Whole fat. 5.2 3.2 1.3 3.3 6.6 17.7 46.9 14.2 1.7 (ii) MOLECULAR PERCENTAGES. Butter-fat “ A.” Butter-fat “B.” --GGz Fully- saturated-saturated unsaturated part. Part-4.2 6.1 2.6 3.7 3*3* ,* 2.4 4.2 4.3 8 4 17.6 17.7 45.7 47.8 19.6 9.0 0.3 3.0 Butter-fat I‘ B.” Acid.Butyric Caproic Caprylic Capric Lauric Myristic Palmitic Stearic Arachidic Whole fat. 13.6 5.4 2-3 3.9 5.5 15.0 38.2 15.2 0.9 Fully-saturated part. 11.0 6.5 1.8 3.3 4.1 17.9 39.6 15.8 -Mixed saturated-unsaturated part. 16.2 4.6 2.5 4.5 6-8 12.3 36.9 14-6 1.6 Whole fat. 12.7 6.0 1.9 4.1 7.0 16.8 39.5 10.8 1-2 Fully-saturated part. 10.5 4.9 5*0* 3.1 4.7 17.0 39.3 15.2 0.3 Mixed saturated-unsaturated part. 14.6 6.7 5.2 9.0 16.4 39.4 6.7 2.0 ,* * Cf. footnotes to Tables on pp. 89 90. Bearing in mind the undesirability of attaching too great importance to the values for caproic-lauric acids (which are present in amounts too small for accurate determination by the methods employed) it nevertheless appears that there is a definite though slight tendency for the lower fatty acids to associate with the unsaturated fatty acids more than with the higher saturated acids.This is compensated for by a slight corresponding concentration of myristic palmiti 92 HILDITCH AND JONES THE FATTY ACIDS AND COMPONENT and stearic acids in the fully-saturated glycerides; but it is noteworthy that the relative proportions of palmitic acid vary less than those of the other acids. Palmitic acid indeed appears to stand somewhat apart in its general relationships from the other acids throughout the whole series of analyses-a feature which we believe is characteristic of this acid in many other fats of vegetable as well as animal origin.We have not yet attempted to separate the fully-saturated glycerides by selective crystallisation from an appropriate solvent but their free solubility in ether and acetone leads us to believe that they consist of a complex system of mixed glycerides and to share the view of other workers that simple triglycerides are not present; from the general characteristics and properties of the material it does not appear probable moreover that glycerides containing only palmitic and stearic acids are present in any notable proportion. COMPOSITION OF THE MIXED SATURATED AND UNSATURATED GLYCERIDES.-The numerical data which we have obtained permit us to estimate the relative molecular proportions of saturated and unsaturated fatty acids combined in this section of the fat and therefrom to give certain limiting figures for the amounts of mono-oleo- and di-oleo- glycerides and of triolein which may be present (in discussing this aspect of the results the unsaturated matter is referred to for simplicity as though it were all made up of oleic acid).I t would be possible to assign definite values to each of these three groups if for example we were able to obtain an independent figure for the percentage of triolein present; but our attempts to devise a procedure to this end have hitherto been unsuccessful. We have two almost independent methods available for determining the molecular ratio of saturated and unsaturated acids combined in the mixed saturated-unsaturated glycerides of the fat : Given the percentage of fully-saturated glycerides present the mean equivalents of these and of the original fat and the proportion of saturated acids in the original fat the mean equivalent x of the saturated acids linked with unsaturated acids in mixed glycerides can be directly obtained and hence the molecular ratio of the saturated and unsaturated acids: (i) Acids in 100 grms.fat Mean equivalent of total fatty acids. Acids in fully-saturated part Mean equivalent of these acids. oleic linoleic saturated acids acid acid in mixed part 282 280 x - -+ - + - + This gave the following results when applied to the experimental data: Mean equivalent of Ratio of mols. saturated acids saturated acids linked with un- per 100 mols.saturated acids. unsaturated acids. Butter-fat “A” . . . . 207.9 106 100 , , “B” . . 21 1-8 103 10 GLYCERIDES OF SOME NEW ZEALAND BUTTERS 93 (ii) From the differential determinations of each individual acid recorded on p. 90 the molecular ratios of the saturated and unsaturated acids can be directly derived (the ratios for each acid have been inserted in the fifth column of the tables indicated) ; these give the following values: Butter-fat “ A ” . . . . . . . . ,) , ,,B” . . Ratio of mols. saturated acids per 100 mols. unsaturated acid. 104 100 104 100 The agreement between the respective estimates for each fat is satisfactoriIy close and the mean ratios 105 100 for butter-fat “A” and 103.5 100 for butter-fat “B” have been employed in the calculations which follow.It is readily possible knowing this ratio to deduce the general composition of the fats on the successive hypotheses that either dioleo-glycerides or triolein are completely absent {absence of mono-oleo-glycerides is inconsistent with the observed figures) ; the resulting data are as follows: Molecular percentages. Weight percentages. r \ c > glycerides. glycerides. “olein.” glycerides. glycerides. “olein.” Per Cent. Per Cent. Per Cent. Per Cent. Per Cent. Per Cent. Butter-fat I s A.” 64 46 - 61 49 -26 77 - 23 74 -A A Mono-“oleo”- Di-“oleo”- Tri- Mono-“oleo”- Di-“oleo”- Tri-Bzctter-fat “B.” 63 47 - 60 60 -27 76 - 24 73 -Therefore in each of the original butter-fats there cannot be less than about 36 per cent. of mono-oleo-disaturated glycerides nor can there be more than 18 per cent.of tri-olein (or 35 to 36 per cent. of di-oleo-monosaturated glyceride). The actual values lie somewhere between the limiting figures given; in the absence of any direct method for determining triolein this is as far as the analyticaldata take us. We venture to predict however that the actual values are not widely removed from those for a mixture of mono-oleo-disaturated and dioleo-monosaturated glycerides. Our belief that triolein is not likely to be present in any large pro-portion is based (i) on the argument that since it is now tolerably evident that simple triglycerides of the saturated acids are either absent from or present in only minute amounts in butter and most other solid fats there is no reason to suppose that oleic acid will tend to form triolein in large proportions and (ii) on the absence of any positive indication of its occurrence in any fat hitherto investigated which contains sufficient saturated acid to provide mixed glycerides with all the oleic acid present 94 HILDTICH AND JONES THE FATTY ACIDS AND COMPONENT SuMMARY.-Methods have been developed for the determination of the proportions of each of the fatty acids contained in butter fat and for a semi-quantitative determination of the manner in which the acids are combined to form the component glycerides of the natural fat.The investigation has been carried out in connection with three samples of New Zealand butter all of which gave results of a similar order. The composition of the mixed fatty acids has been obtained by: (i) Removing as much of the lower fatty acids as possible by prolonged distillation in steam the steam-volatile acids being recovered and fractionally distilled in the form of free acids; (ii) Separating the fatty acids non-volatile in steam into two groups by means of the lead salt and alcohol method followed by conversion of the acids from the soluble and insoluble lead salts into methyl esters which were quanti-tatively fractionated at low pressure in the usual manner.The approximate composition of the fatty acids was Butyric 3; caproic 2; caprylic 1; capric 2; lauric 4; myristic 11; palmitic 28; stearic 9; oleic 33-34; and linoleic 4-5 per cent. The values for the volatile fatty acids accord with recent determinations by other workers ; the values for myristic palmitic and stearic acids differ from many previously recorded whilst there is consistent evidence of the presence of a small percentage of acids less saturated than oleic.The procedure for the study of the component glycerides consisted in oxidising the butter-fat under conditions in which all unsaturated components were trans-formed into acidic products whilst glycerides containing only saturated fatty acids remained unaltered. The latter were recovered their proportion noted, and the composition of the mixed fatty acids contained therein determined as above. From the resulting data the following general conclusions were drawn : (i) The butter-fats examined contained about 30 per cent. of fully-saturated glycerides.The fatty acids therein were the same as those in the whole fat and in proportions not widely different from those of the latter; but a tendency was noted for the occurrence of somewhat less of the volatile fatty acids in the fully-saturated part coupled with a correspondingly slight concentration of the higher saturated fatty acids in this group of the glycerides. It is probable that all these glycerides are of the complex mixed type. (ii) The remainder of the fat (about 70 per cent.) consisted of mixed gly-cerides of saturated and unsaturated acids the molecular proportions of the acids being about 104 mols. of saturated to 100 mols. of unsaturated acid. The amount of mono-oleo-disaturated glycerides in the original fat is at least 36 per cent. and there cannot be more than 18 per cent.of triolein (or 36 per cent. of dioleo-mono-saturated glycerides) ; although no positive data are available it is quite probabl GLYCERIDES OF SOME NEW ZEALAND BUTTERS 95 that but little tri-olein is present and that the approximate composition (in round numbers) of the butter-fats examined is :-Per Cent. . . ca. 30 Mixed mono-oleo-disaturated glycerides . . . . ca. 36 Mixed di-oleo-monosaturated , ca. 34 Mixed fully-saturated glycerides . . It may be noted that all except the unsaturated acids are comparatively evenly distributed throughout the whole fat. These results of course do not necessarily hold in detail for butters produced from widely varying sources. The investigation is being extended to butters made from the milk of cows feeding on a variety of diets including (if possible) cases in which hard and soft oilcakes have been respectively employed in the diet.We desire to express our most cordial thanks to Messrs. Lever Bros. Ltd., who have assisted us by obtaining supplies of the butters which we have studied, by carrying out some of the determinations of the technical “constants” of the fats and especially by the provision of a Research Studentship in the Department of Industrial Chemistry of this University which has enabled one of us to pursue the research. THB UNIVERSITY, LIVERPOOL. DISCUSSION. The PRESIDENT said that in this remarkable and valuable paper (not the first which the Society had had from Professor Hilditch) the authors had dealt in a fundamental manner with a fat which many had to examine daily.He had a suggestion to make on one small point Professor Hilditch seemed to wonder whether the accuracy of some of his results was not due to coincidence. Would it not be possible to apply this method of fractionation to known mixtures of fatty acids so as to substantiate the percentage accuracy of his separation? Mr. E. R. BOLTON remarked that he was struck by the dramatic way in which the paper opened by pointing out that the data availablewere in a chaotic condition. Professor Hilditch had gone forward step by step and had given us absolutely new information. One of the most striking points was the fact that the butyric acid as determined by the Kirschner method was rather in excess of that actually present; this was interesting as it had always been understood that the Kirschner method afforded a good measure of the butyric acid.Mr. Bolton then referred to the problem of the detection of carcase fat in butter and mentioned that the problem had become a very live one in India where certain castes were not allowed to eat any butter containing carcase fat. Government chemists therefore had to face this problem and had considerable difficulty in proving a butter to be free from this fat. If it were possible to bring the stearic acid of pure butter fat within narrower limits than the 0 to 22 per cent. recorded by Mr. Mitchell, to evolve a method for its accurate and rapid determination and to state that pure butter should not contain more than a certain amount of that acid it might afford a means of settling the problem of the Indian chemists.Mr. C. A. MITCHELL said the method devised by Professor Hilditch and his co-worker marked a great advance on the original oxidation method of Hazur 96 HILDITCH AND JONES FATTY ACIDS AND COMPONENT GLYCERIDES ETC. and Griissner which gave results of qualitative rather than quantitative value. With regard to Mr. Bolton’s suggestion of using the proportion of stearic acid as a means of differentiating between butter fat and animal body fat this was the god to which Hehner and he (the speaker) had worked for months but the idea had had to be abandoned owing to the very wide variations in the stearic acid content of butter fat. Most of the early experiments on the crystallisation of butter fatty acids from a solvent saturated with pure stearic acid gave deposits ranging from a mere trace up to about 6 per cent.but subsequently several undoubtedly pure samples gave deposits of 12 to 22 per cent. and subsequently this had been re-peatedly confirmed. It was also an interesting confirmation that the amount of stearic acid found by Holland and Buckley in American butter fat was over 20 per cent. Dr. H. E. Cox remarked on the fact that the Kirschner value was shown to give an apparent butyric acid content about 20 per cent. too high and asked for information as to exactly what acids were represented by the Reichert-Meissl, Polenske and Kirschner values respectively. There was sometimes observed in genuine butter an unusually large spread between the Reichert-Meissl and the Kirschner values amounting to as much as 25 per cent.or more of the Reichert-Meissl value which was a larger variation than appeared in Prof. Hilditch’s figures for the sum of the capric caproic and caprylic acids part of which with the butyric acid accounted for the Reichert-Meissl value. Professor HILDITCH replying said that with regard to the President’s question as to whether it were not-possible to test the accuracy of the results this had been done in the case of comparatively simple mixtures and the results could be relied upon but it had not been done with more complicated mixtures such as the particularly difficult mixture which resulted when Cr and Go acids were present. It might be a very good idea to try this. Professor Hilditch did not think he could add much on this point to what had been said in the reading of the paper. As far as his figures were concerned the butyric acid figure was more likely to be on the high side than on the low side and probably the Kirschner value included caproic acid. What he was really trying to get at was the constitution of natural fats as far as this method would take him and the examination of the composition of the glycerides contained in the fats; butter fats were one section of the work. With regard to the percentage of stearic acid in butter fat on the whole he agreed with Mr. Mitchell. He had found that there was considerable variation. Referring to the length of time taken by this test Professor Hilditch stated that at the present time he and his co-workers were trying to see if they could work out a modified test occupying not more than two days. They were also studying the method from the point of view of analytical application. Regarding Dr. Cox’s question with reference to the Reichert-Meissl and Polenske values he really hesitated to say but it was quite possible that the “spread” which was mentioned existed but he did not see that this could really be correlated with the results recorded in this paper. Mr. Bolton and Dr. Cox had both mentioned the Kirschner number