SS-fi Hntt Q|oU»ge of J^gncultute At (Darnell InwerBitH 3tl(ata. N. 1. ICibratg Cornell University Library TX 551.B69 ■3""'i924 00'3 541 665 The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924003541665 FATTY FOODS THEIR PRACTICAL EXAMINATION A HANDBOOK FOB THE USE OF ANALYTICAL AND TECHNICAL CHEMISTS BY B. RICHABp§ BOLTON, F.C.S. CONSULTING ANALYST ANp TECHNICAL CHEMIST AND CECIL RBVIS CHIEI' CHEMIST, MESSRS. WELFOKD AND SONS , LIMITED LONDON J . & A, CHURCHILL 7, GREAT MARLBOROUGH STREET PHINTED BY ADLAKD AND SON, LONDON AND DORKING. PREFACE ■ ■ Tn' no department of analytical chemistry is greater difRcuity experienced than in that which deals with the examination of natural food-products. Of these food- products perhaps none present so much diversity in natural composition as do the oils and fats. This diversity of oharacterj combined with a marked similarity of constitution, and, in consequence, bearing with it the greatest difficulty of resolution, has ma,de the analysis of oils and fats both prolific of method and uncertain in result. Unfortunately the wide varia- tions which pure oils and fats show in rare instances are often made so much of that a feeling of great un- certainty is raised in the mind, of the more casual investi- gator as to the reliance he may put on the analytical indications to be obtained—a feeling which would not be shared by the expert in the field. We have, therefore, endeavoured to present certain methods of examination of edible oils and fats and such technical details as are necessary for their interpretation in a more concise and definite manner than that in which they are usually treated, passing over as much as pos- sible those rare exceptions with which the investigator may be never faced. Out of the vast host of analytical methods for arriving iv Fatty Foods at, not only various ends, but also the same end, we liave made a selection which experience has shown us are those which will arrive at satisfactory solutions in the great majority of cases. We have, as a rule, given only one method of achieving a certain result, and if the reader, whether critic or expert, should not therefore find some method, which he consequently thinks has been omitted through ignorance, it is suflScient to say that omission of a method may often be better evidence of the knowledge of it than its presentation. Nor does there seem the slightest use in preserving methods and ideas which, however efficient they have been in their time, have become superseded or unnecessary. The progress of analytical chemistry is not served either by a foolish clinging to obsolete methods or by the elabora- tion of new ones which are supererogatory. To some no arrangement or advice or scheme would be satis- factory or free from objection, and with them this book has no concern. To the more humble investigator in this difficult field we desire to present the results of many years' experience, with the hope that methods and ideas which have brought success to the authors may enable him also to elucidate the problems with which he may have to grapple, though we cannot lay too much stress on the fact that scientific and skilled adulteration has sounded the death-knell of rough and ready methods of examination. We desire to express our indebtedness to Mr. Francis H. Loder for procuring for us various commercial specimens of nuts and seeds, and of the oils and fats Preface v obtained from them, and also for some helpful sugges- tions. We are indebted to Dr. R. Vincent, Senior Physician to, and Director of, the Bacteriological Laboratory at the Infants' Hospital, Vincent Square, S.W., for the photographs of bacteria, etc., in the section on Milk. Our thanks are due to Messrs. Baird & Tatlock (London), Ltd., for making the special extraction apparatus described on p. 11, and for the use of the block for the purpose of illustration. The Director of the Royal Botanic Gardens, Kew, ■ has kindly allowed us to sketch some of the fruits and seeds which were not in our possession, and Mr. J. Masters Hillier has courteously facilitated this work. We also desire to thank Mr. T. Macara for his kindness in allowing us to include his hitherto un- published and most valuable work on the use of the " levigation " process in the analysis of cocoa, etc. Our greatest thanks, however, are due to Miss Enid M. Jesson, not only for the original drawings of various fruits and oil-bearing seeds, etc., but for all the labour entailed in the preparation of the MSS. for publication. We feel that without her aid this book in all probability would have fa-llen far short of that usefulness which we trust it will attain'. E. RICHARDS BOLTON. CECIL REVIS. 16j Flanohfobd Eoad, London, W. CONTENTS PAGE CHAPTEE I. G-ENEEAL Introduction ... 1 Objects of Analysis of Oils, Fats, and Fatty Foods. CHAPTEE II. General Analytical Methods . . 7 Preparation of Sample for Analysis — Extraction of Oil or Fat — Physical Tests — Cliemieal Tests — Colour Tests — Microscopical Examination of Beef and Lard Fats — Special Tests — Glycerol -^Impurities — Oils and Pats commonly employed in various Manufactures. CHAPTEE III. Analytical Procedtjee and its Interpretation . 74 Inferences from Tests — Typical Cases — Resolution of PaLty Acids. CHAPTEE IV. Beef Fat, Laed, Butter and Maegaeine, Ghee . 94 CHAPTER V. Vegetable Oils and Fats . . . 141 Commercial Products — AnalyticSil Data — Special Tests- Source— Characteristics — Fish and Marine Animal Oils. CHAPTEE VI. Eancidity . . . . 284 viii Fatty Foods TA.GT. CHAPTER VII. Cocoa, Chocolate, and Milk Chocolate . 289 Introductory — Methods of Analysis— Calculation of Com- position, CHAPTEE VIII. Pebding Stuffs 319 Introductory — Methods of Analysis — Calculation of Rations. CHAPTEE IX. Milk . . 345 Analysis and Examination for Use in Margarine — Pas- teurisation — Cleanliness and Bacterioloi^ical Condition — Cream and Separated Milk. Appendix A 281 Appendix B -ISi Appendix C '282 Appendix D 362 Botanical Index 364 Index 365 LIST OF ILLUSTRATIONS 1. Fat Extraction Apparatus . .10 2. ., „ . 11 3. Butyro-Refractometer, Zeiss . . 19 4. Special Funnel . ... . .46 Chart showing Ratio of Saponification Value, Iodine Yalue, and Refractive Index for cei-tain Oils and Fats 76 5. Reichert-Meissl Polenske-Kirschner Apparatus . . 118 6. „ ., , Value Curves 130 7. „ „ „ ., . 131 8. Attalea Cohune. Fruit and Seed . . 157 9. Shorea robusta. Fruit . . 166 10. Stillingia sebifera. Fruit . 169 11. Vateria indica. Fruit . 179 12. G-arcinia indica. Fruit . . ^ . 182 13. Butyrospermum Parkii. Fruit and Seed . 184 14. Bassia butyracea. Seed 188 15. Mimnsops Djave. Seed . 190 16. Bassia latifolia. Seed . . 192 17. Bassia longifolia. Seed ... . 194 18. Olea eiu'opaea. Fruit and Seed . . . 199 19. Arachis hypogsea. Fruit and Seed . 203 20. Briodendron anfractuosum. Fi-uit and Seed . 220 21. Glycine hispida. Fruiting Branch and Seed 227 22. ZeaMays. Fruit . . 231 23. Helianthus annuus. Fruit ■ . 238 / / X Fatty Foods PIG. PAGE 24. Papaver somniferum. Pruit and Seed 242 25. Aleurites triloba. Fruit and Seed 249 26. Aleurites Fordii. Fruit . 250 27. Hevea brasiliensis. Fruit and Seed 253 28. Ricinus communis. Fruit and Seed . 262 29. Taraktogenos Kurzii. Seed . 268 29. Hydnocarpus "Wightiana. Friiit . . 268 29. Hydnocarpus venenata. Seed . . 268 30. Scyphoceptalium Ochocoa. Seed . 270 31. Buchner Funnel, with Glass Cylinder . 297 32. Curve for Nitrogen in Cocoa Nib and Shell 305 33. Curves for Nitrogen, Fibre and " Levigation " in Cocoa Nib and Shell . . 306 34. Kjeldahl Apparatus . . 332 35. Fibre Estimation Apparatus 337 36. Gottlieb Apparatus 348 PLATES FOLLOWING Frontispiece. African Oil Palm. (Elxis.gmtiieensis) I. Beef and Lard Crystals. (5 photomicrographs) . 56 II. Cocoa and Adulterated Cocoa. (5 photomicrographs) 302 III. Starches, ifi photomicrographs) . ' 302 IV. Starches. (6 photomicrographs) . . 302 V. Lactic Acid Forming Organisms. {Photomicrograph) . 356 Cellular Elements in Milk Sediment. {Photom,icro- graph) . . . ■ ■ 356 VI. Curd in Enteritidis sporogenes Test . * 360 B. enteritidis sporogenes from Milk Culture. (Photo- micrograph) ..... 360 ERRATA Page 15, 3rd figure in 1st column of table, " + 6 " should read " — 6," Page 31, 1st line, " 3 grm." should read " 1 grm." Page 177, title of page, " Malabar " should read " Mafura." FATTY FOODS CHAPTER I. GElSrEEAL INTEODUCTION. In the following chapters the examination of fatty foods is regarded purely from the edible standpoint. It must be clearly understood that only the properties of the various oils, etc., are consideredj which come within that purview. As a consequence, the treatment is somewhat one-sided, but the edible oil industry has undergone such enormous development of late years, that it has bpoome, for certain purposes, especially that of margarine manufacture, quite possible to regard oils and fats from the point of view of that industry only, and to supply all the necessary information to those engaged in such work, without treating of the various other purposes to which the same oils may be put. This method of treatment has been adhered to in every particulai-, with one or two slight exceptions in which the divergence seemed justified. The analytical methods ■ here given for the practical - examination of fatty foods presuppose a sufficient knowledge of the chemistry of fats and oils, and this 1 2 Fatty Foods -,__ knowledge is therefore tacitly assumed, but, if informa- tion be required in this direction, the reader is referred to such a standard text-boolt as Oils, Fats und Waxes, by Dr. J. Lewkowitsch. No attempt has been made to cover the whole of the field which to-day must be included under the title, but simply to place in the hands of those not intimately acquainted with the subject, such analytical methods and hints as to their interpretation, as will enable the casual investigator in the field to arrive at satis- factory conclusions regarding the more ordinary fatty foods which may come before him. There is no pretence of covering the whole field of analytical methods which have been, and are employed in such investigations; but those methods, which from long experience the authors consider in the vast majority of cases both reliable and sufficient, are given in such a detailed form that the performance of them, under strict adherence to the description laid down, will produce similar analytical figures to those which they themselves obtain. In the examination of fatty foods it is necessary to consider — (1) The object of the analysis. (2) The analytical methods to be employed; and then to pass to — (3) The interpretation of the results obtained ; for which purpose it is necessary to have — (4) Typical analytical data for the substances dealt with, and descriptions of their source and. origin. Object of Analysis. ' I. Oils and Fats (perse). The examination of these is undertaken in order to determine — (1) Ihe-ir purity, i. e. their freedom from other General Considerations 3 fats, oils, waxes and substances not natural to them. In tlie case of simple oils and fats, tHis may in many cases present no great difficulty, but when complex mixtures of two or more different fatty substances are the subject of the investigation, it is not too much to say that the interpretation of results can only be characterised as careful and systematic scientific guesswork, in which the experience of the analyst plays a part equally as great as that of his analytical skill. In many cases, on the other hand, particularly those of direct substitution, the deductions may be made with as great certainty as with any other type of analysis. Under the heading of purity will come also determina- tions of vfiuiiiture, non-futty impurities, metallic con- taminations and arlificioil colouring matters. (2) Their suitability fur technical purposes. To give an answer to such a question as " suitability" for a certain purpose, will generally require technical knowledge and experience, but to deal with this is outside the scope of this woi-k, and it is only possible to indicate the kind _of question which may be asked. When, however, a definite question is put, it is often the case, that the exercise of common-sense and reference to the properties of the substance under consideration will help, even the inexperienced, to give a satisfactory reply. The problems which may arise are : (a) The matching pf one oil or fat, or a mixture of such substances, by some other oil or fat or mixture, which in general properties and behaviour will enable it to be used as a substitute. Such cases would arise with biscuit and cake fats, cocoa-butter substitutes, etc. It is needless to say that the substitute must be cheaper than the real article. (6) The comparison of deliveries with samples sup- 4 Fatty Foods plied. This is often an easy task, but in many cases will require very careful examination to avoid sub- stitutes. (c) The investigation of competitive articles, such as, the precise nature of the mixtures used in similar processes by different manufacturers. {d) The possibility of cheapening a manufacturing process or article by the use of new oils and fats. In giving an opinion on these points it will be necessary to gather all the information possible as to (]) The keeping properties of the aih and fats which it is proposed to use, under the conditions to which they are to be exposed. For instance, a fat which would be possible for a certain purpose if used alone, might, from its likelihood of becoming rancid, or of being easily attacked by mould, be quite unsuitable if mixed with protein materials. (2) The consistency of mixtures of oils and fats, and their consequent behaviour at temperatures to which they will be exposed. Consistency, however, is not a simple function of the melting and solidifying points. It is rather a question of physical characteristics, and the result of the formation of eutectic compounds. This must be carefully remembered, for the addition of a high melting-point fat to a low melting-point fat does not always result in an intermediate mixture, as the melting-point of the mixture may be below that of the more fusible component until a considerable addition of the other has been made. It is necessary also to distinguish between brittle fats and plastic fats. Brittleness may be overcome sometimes by small addi- tions of. other fats (which may have to be looked for very carefully), or it may be got rid of by mechanical treatment only. (3) Digestibility : To this there is no adequate means of reply. It is usually taken that ease of saponification and digestibility are synonymous terms, General Considerations 5 but it is very doubtful if tliey are really so. (4) The presence of poisonous substances : These may arise either from (a) the fat itself — such as in castor/croton, chaulmugra oils etc.; or from (h) the process of manu- facture, such as contaminations from copper, arsenic and other injurious metals, as well as from organic substances. Some vegetable oils and fats in their crude state may, on account of the presence of cyanogenetic glucosides in the original seed, etc., contain appreciable quantities of prussic acid, which is, however, never likely to be found in the refined oil. (5) Taste and smell : While some oils or fats are valued for the presence of a characteristic taste or smell, others are judged by the absence of these. Information may often be required as to whether it would be possible to permanently I'ender an oil entirely tasteless, or odourless, or whether, on the other hand, a desired taste or smell may be enhanced, or is of such an intensity as should be reasonably expected. This, however, i-cquires a technical knowledge of the processes used in oil refining. (6) Colour : It is often necessary to determine whether the colour is natural or artificial, and in those cases in which the colour is objectionable, it may be necessary to say whether it can be removed or improved. In Chapter V, dealing with the various fats and oils, the above properties will be given as far as possible. II. Fatty Foods. In the examination of these substances in which the oil or fat only appears as a constituent, it will be necessary to determine : (a) The content of fatty matter, and in many cases whether it be natural to the food, and, if not natural. 6 Fatty Foods whether suitable or prejudicial to the natui'e and use of the food. (6) The content of non-fat ly constituents; in which case it may be necessary only to determine these quantitatively under certain arbitrary headings, or in certain special cases their actual nature may be required. (c) In the case of faulty products, whether the fault lies with the fat or the non-fatty constituents. General Analytical Methods CHAPTER II. Genebal Analytical Methods. It is unfortunately too often assumed that the analysis of fatty materials is of a rough and ready character, but it cannot be too clearly understood that the utmost possible precision, not only of the analytical procedure, but ^Iso in the preparation and standardisation of the reagents and instruments employed, is absolutely neces- saryin such analysis. For this reason all burettes, pipettes and measuring vessels of any kind should be carefully checked by the analyst under the conditions of his owil work, though on account of the difficulty of checking instruments of large ca,pacity, it is advisable to obtain these certified, either at the National Physical Laboratory or at Gharlottenburg, and to bring all other apparatus into line with these by actual weighing. (1 c.c. = 0-99807 grm. weighed with brass weights in air at 60" F.) , Similarly in the case of the staridardisation of solutions, such standardisation should be carried out against solutions of the same normality. For instance N/10 acid should be standardised against N/10 alkali directly made up, and not against a solution made by diluting the normal strength ten times, and so on. The necessity of these precautions will be understood when we consider the high molecular weights involved in fat analysis. In these days of yearly publication of revised atomic weights, care must be taken iio employ 8 Fatty Foods the same figures for all purposes. For instance, in cal- culating nitrogen, analysts invariably employ 14 as the atomic weight of nitrogen, but will at the same time calculate N/HgSO^ as containing 49 grm. per litre (instead of 49-04). Preparation of Sample for Analysis. Case 1 : Fat oe Oil per se. — The sample must (if necessary) be brought into the liquid condition at the lowest possible temperature, and if then clear and bright, the analysis of it may be proceeded with forthwith. Turbidity may be due to suspended matters or moisture, and in either case filtration through thick soft filter-paper should be resorted to until a clear and bright filtrate is obtained. The filtration should be carried out at a temperature sufficiently high to keep the whole of the fat in a melted condition, any obvious water being allowed to settle out before pouring on the filter. While any good thick soft filter-paper will serve the purpose, we strongly recommend C.S. and S. No. 604. Case 2 : Oleaginoos Materials. — In this case the fat will have to be extracted after suitable preparation of the material. Great care must be exercised in adopting a method of disintegration suitable to the nature of the sample under examination. (a) Materials of comparatively low fat content, if of a friable ilature and not obviously greasy, may be ground in a coffee mill, or even with a pestle and mortar. Precaution must be taken that the whole of the material passes through, as in the case of certain substances portions which are of a less friable nature may be left in the mill. In all cases a sufficiently Extraction and Estimation of the Oil or Fat 9 large portion should be ground to produce a fair average sample. The ground sample should be immediately transferred to a stoppered bottle. {h) Materials of high fat content have to be carefully- handled, as a small amount of pressure may cause portions of the fat to be exuded. In many of these cases a shredding machine, such as is used for vege- tables, should be employed, or it may even be necessary to commence extraction without any preliminary dis- integration. It is undoubtedly due to faulty methods of shredding that figures are published as represent- ing the average fat content of such substances as copra, which to any oil miller are obviously absurd. In mill operations, a slight error in the fat content of a sample, caused either by faulty disintegration or extraction, may cause apparent loss or gaiii of many tons of fat in a short period. ^ Extraction and Estimation of the Oil or Fat. The only satisfactory method (with certain rare excep- tions) is a continuous extraction by means of a volatile solvent. For this purpose the authors prefer light petroleum spirit boiling below 40° C, as this solvent extracts a pure fat where in many cases methylated ether would extract non-fatty substances as well as the fat. The ordinary Soxhlet extractor may be used, but the form shown in the sketch (Fig. 1) is more efficient, eco- nomical of solvent, and rapid, and has been employed for many years by the authors for a very large number of extractions carried out in a routine manner, but requir- ing at the same time a high degi-ee of accuracy. The fitting up of the extractor will be seen from the sketch. The outer tube is indented about two inches above the 10 Fatty Foods neck in order to carry the extractor proper, which is lightly plugged at its lower end with cotton-wool, and has a disc of filter-paper resting on the wool, to prevent the material to be extracted passing into the wool. The lower end of the exteactor just enters a folded filter placed in the neck of the outer tube. The quantity of solvent used should not be sufficient to fill the extraction tube when working. The flask should be sunk to the shoulder in a water-bath, which may be safelj' and efficiently heated by means of a 16 c.p. electric lamp, partially submerged in tlie water, and as rapid a rate of flow is maintained as is consistent with safety. The authors have devised a special form of extractor (Fig. 2) enabling the analyst to deal with varying quanti- ties of material in the same apparatus and without undue loss of solvent. The details are easily seen from the sketch, thi'ee sizes of inner extraction tube being sup- The general method of Soivtrtt Fia. 1. plied with the apparatus, working it, is as described above. Extraction and Estimation of the Oil or Fat 11 flask is detached distilled off with inclined position splashing of the As a general rule, from half to two hours is sufficient for the first extraction, after which the extractor is removed and the solvent allowed to spon- taneously evaporate in a moderately warm place (too rapid heating may cause ejection of the contents). When the solvent has evaporated, the tube is placed in a water oven for about an hour, the contents having been first distributed over the walls of the tube by gentle tapping. After this the contents of the tube are transferred to a mortar, together with 1 or 2 grm. of clean dry sand (passing tiO mesh), and ground as finely as possible, the contents of the mortar being |i^ carefully brushed back into ^K extractor, and the mortar washed out with the solvent. Extraction is then continued for an hour to an hour and a half, after which the and the solvent the flask in an (so as to prevent fat on to the delivery tube), the flask being then placed in a water oven and dried to constant weight. The flask may with advantage be inclined side- ways, and turned from time to time. If heating be not unduly prolonged, there is little danger' of appreciable increase of weight due to oxidation, and in the-opihion of the authors the use of a vacuum drying apparatus is unnecessary, but on no account should a temperature of 105 C. be exceeded for drying. Fia. 12 Fatty Foods To obtain sufficient quantities of the fat for analytical examination an extractor of larger type may be employed, and the grinding with sand, as a rule, omitted. PHYSICAL TESTS. (a) Melting Point. Capillary Method. — The fat is melted and introduced into a tube about 0"02o in. bore, with the walls not more than 0003 in. thick. The fat in the tube must then be thoroughly cooled, either by being kept on ice for half an hour, or allowed to stand at room temperature for at least twenty-four hours. Many fats may, however, be introduced into the tube by simply pressing it into the fatty mass, but in certain cases there is a danger of pressing out som^^ the softer constituents. If the latter method be jJPerted to, and the fat has been solid for twenty-four hours previously, it is obvious that no subsequent cooling is necessary. The tube is attached to a thermometer by means of rubber bands and the thermometer cMtped in position in a beaker containing not less than^TOG c.c. of water, the top of the tube being submerged. The temperature of the water is now raised by means of a small flame, at the rate of about 0'5° C. per minute, the water being kept^t^^iyjy mechanical means, preferably by blowing a st^^i^^BSr through the water. The fat is observed by meaBigifjfe lens, and as soon as it is sufficiently melted to form a clearly defined meniscus on the top, the temperature is read and recorded as the point of incipient fusion. The heating is then continued until the fat is quite free from unmelted particles, the latter temperatui-e being recorded as the point of complete fusion, which latter is usually returned as the melting point. Solidifying Point : Titer Test 13 The difference between the points of incipient and complete fusion affords valuable information to the analyst as to the range of melting-point of the products of which the fat may be composed, and to the experienced eye may give a clue to the presence of high melting- point fats or waxes which may have been added. (6) Solidifying Point. Twenty-five to 30 c.c. of the melted fat are introduced into a tube about l^^in. wide and 5 in. long, in which is supported a thermometer by means of a cork. As soon as the fat shows any tendency to solidify stirring is com- menced with the thermometer till the fat solidifies, the temperature being closely noted from time to time. In most cases after a steady fall the te mp erature will be seen t6rise,thehighestpoint then reaoj^B^ing the solidifying point. Many fats, however, do^^^Hiow this rise very definitely, but gradually solidify ^Bout further fall of temperature, the mercury remaining stationary during the process. In such cases the stationary point may be taken as the solidifying point. The solidifying point of oils, liquid at the ordinary temperature of the air, is determined by immersing the tube in a freezing mixture. With fats which give an indefinite solidifying point, it will be found that the solidifying point of the fatty acids may be obtained with much greater precision. This determination is called : The Titee Tkst. For this test the preparation of the fatty acids must be carefully carried out, for on their purity the success of the test very largely depends. A simple and reliable method is as follows : In a 300 c.c. flask are placed about 50 grm. of the fat, together with about 15 grm. of stick caustic soda, 14 Fatty Foods and 100 c.c. of strong (95 per cent.) alcohol. The flask is connected to a reflux condenser, and heated in a boiling water bath with constant shaking until saponification is complete, which may be taken to be the case when the whole of the fat has gone into solution. The alcohol is then rapidly boiled off, and the soap dissolved in not less than 500 c.c. of hot water. To this solution, while still warm, are added one or two drops nf methyl orange and somewhat more dilute hydrochloric acid (1 ; 3) than is necessary to render the solution permanently pink, the solution being then kept in boiling water until the layer of fatty acids becomes quite clear. The water is syphoned off as completely as possible, and the residual fatty acids filtered through dry, soft, thick filter-paper, the operation being repeated uafl||||a perfectly bright clear filtrate results. The soli^^Kig point of the fatty acids so obtained is deterrn^red in a similar manner to that of the fats themselves. 'J'he above method is satisfactory for all ordinary purposes, but as in certain cases, especially in con- nection with soap manufacture, the necessary fats are bought and sold on a basis of the Titer test, it is necessary to carry out the final operation in a more standard manner. The usual method is known as Dalican's method, and for this purpose the melted fatty acids, obtained as above described, are poured into the standard tube mentioned above (1^ in. X 5 in.) until it is half full. The tube is placed in a wide-mouthed bottle 10 cm. wide and 13 cm. high, and so fixed in the mouth that the whole of the fatty acids are inside the bottle. A . carefully standardised thermometer (specially made for the purpose) is inserted into the fatty acids so that the bulb occupies a central position, and immediately Titer Test 15 crystals form at the bottom of the tube the mass is stirred, with the thermometer, three times from right to left, and then three times in the contrary direction, and then continuously and rapidly with a circular motion, avoiding contact "with the sides, care being taken to incorporate the solidified fatty acids as they, separate, until the whole mass is uniformly turbid. The mercury is then carefully watched until it rises sharply, when the maximum temperature obtained is carefully noted. This point is the Titer point. Duplicate determinations should not vary by more than O'l C. In determining solidifying points, it is of the greatest necessity that the fatty acids employed should be per- fectly dry, otherwise erroneous results will certainly be obtained.' Comparison of the solidifying point of the fat with its Titer point often affords very useful infor- mation as to the nature of the fa^^der examination. Melting point Titer Diffe- of fat or oil. test. rence. Maize oil . - 12 + 18 30 Cotton oil . . + 3 + 33 30 Soja oil . . . + 6 + 22 28 Sesame oil . . - 5 + 24. 29 Sunflower oil . . - 18 + 18 36 Linseed oil . . - 16 + 18 34 Hemp oil - 20 + 16 36 Poppyseed oil . - 20 + 16 36 Ka^e oil ... . - 8 + 13 21 Arachis oil . . + 27 27 Olive oil . . . + 5 + 23 18 Castor oil - 10 + 3 13 Shea nut oil + 18 + 52 34 Coooa butter + 26 + 49 23 Borneo tallow ... + 28 + 50 22 Coconut oil ... + 22 + 23 1 Cohune oil . . . + 20 + 22 2 The above table will illustrate the relationship between the Titer of the fatty acids and the melting- 16 Fatty Foods point of the fat or oil. The figures given are only rough averages, and it should not be forgotten that the figures for different specimens of the fats and oils are likely to vary sometimes considerably on either side of the figures in the table. (c) Specific Gravity. The specific gravity of all oils and fats may be determined most expeditiously and accurately in the ordinary specific gravity bottle, with a capillary bored stopper. A 25 grm. bottle is the most convenient, and its capacity at 15° C. should be determined, once and for all, with the greatest accuracy. The bottle is filled by pouring in the liquid fat or oil at a temperature slightly below that at which the determination is to be made, and is then immersed in water maintained at the desired temperaturqpjlfntil all expansion has ceased; after which it is removed from the water, excess of fat or oil carefully removed, and the outside dried. The bottle is then weighed with the ordinary precautions. As a general rule, the specific gravity of all fats and oils which are liquid at 15° 0. are determined at that temperature ; in all other cases the determination is made at the boiling point of water. For this purpose the bottle is immersed well over the shoulder in a beaker of water which is kept rapidly boiling, the bottle being conveniently kept in position by a wire holder. When expansion is complete, the surplus fat is removed with the finger, the flame turned out, and the bottle removed from the water after a short interval. It is wiped dry and allowed to attain the temperature of the air before being weighed. When specific gravities at the boiling-point of water are to be compared with the specific gravity of water at 15° C. it is necessary to introduce a correction for Optical Activity: Refractive Index 1? the expansion of the bottle. As the co-efEcient of cubical expansion of glass is 0'000025, the correction for a difference of 85° (using a 25 grm. bottle) will be 0"0531j which weight in grammes is to be added to the iveight of water which the battle contains at 15° C. before comparing it with the weight of fat obtained. The greatest confusion has been introduced into the analysis of oils and fats by a senseless multiplicity of temperatures at which specific gravities are determined. The two temperatures of 15° and 100° (usually stated as 99°) are enough for all practical purposes. For the routine examination of a large number of oils a Westphal balance may be employed, but the accuracy is not as great as that of the specific gravity bottle. (d) Optical Activity. In certain cases the examination of oils for optical activity is of some value, as the possession of this characteristic, in any marked degree, is limited to about lialf-a-dozen oils, most of which are either decidedly jDoisonous OT possess marked physiological properties. The exhibition of optical activity must not be construed to be the result of the poisonous nature of the oil, as most essential oils, with which we are not here con- cerned, are optically active. If sufficiently clear, the oil itself is placed in the polarimeter tube, otherwise a chloroform solution is employed. (e) Refractive Index. This is most conveniently determined in the butyro- refractometer of Zeiss, which instrument reads on an arbitrary scale, the divisions of which are usually returned and are so given in this book, but which may be converted into true refractive indices (n)^, by means of a table supplied with the instrument. 2 18 Fatty Foods This value has been much complicated by the employ- ment of greatly divergent temperatures, often aimlessly selected. We propose to adhere to a uniform tempera- ture of 40° C. (at which practically all fats can be examined). In the case of one or two exceptions the reading is made at a higher temperature and calculated to 40° C. by adding 055 per °C. above 40°, to the reading. In the tables of figures for the various oils and fats in Chapters IV and Y, the Zeiss toutyro-refractometer number has been in all cases designated " refractive index at 40° C." CONSTEUCTION AND MODE OF ACTION OF WoLLNy's Butyeo-Refkactometee. The refractometer consists of a beatable Abbe double prism and a permanently attached telescope, the objec- tive of which is adjustable in a slide by means of a micrometer screw. A scale, graduated from — 5 to -f- 105, is placed in the focal plane of the telescope objective ; the upper lens of the ocular is adjustable in order to be able to focus the lines and figures of the scale clearly. Clear daylight is projected into the double prism by (j) and penetrates through the stratum of fluid between the prisms. The naturally coloured horder-line of total reflection, produced by this passage of rays in the focal plane of the telescope, hence also in the plane of the scale, is achromatised in the case of butter by virtue of the special construction of the glass prism turned towards the telescope. Thus a sharp colourless horder- line, which intersects the scale vertically, is seen in the ocular between a light and a dark section of the field of view. The accuracy of the measurement to be made then depends on the exact determination of the point in the scale through which the border-line passes. Refractive Index 19 Arrangement of thjj Eefractometur and of the Hea'iing Appliance. The apparatus is withdrawn from its case by taking Fig. 3. hold of the base plate or the telescope carrier — never the telescope itself — and placed in a convenient position for looking into the telescope. The illumination may be supplied either by the daylight coming in through a window or by the light of a lamp. The refractom^ter can be used in conjunction with 20 Fatty Foods any kind of heating appliance which affords a current of water of constant temperature and uniform speed. Connection with the refractometer should always be established so that the water enters at (u) and flows off at (e). Application op the Sample to the Peisms. The prism casing is opened out by revolving the screw head (e) (Fig. 3) clockwise, giving it a half turn until checked, when one half (b) of the prism casing can simply be hinged down. The surfaces of the prisms and of the metal parts must now be cleaned with scrupulous care, soft linen and a little alcohol or ether being best for the purpose. A small quantity of the sample to be tested is then melted if necessary, and if not quite bright poured upon a small filter. The first two or three drops of clear fat or oil percolating the filter are applied to the surface of the folding prism, in doing which it is expedient to tilt up the apparatus with the left hand, so as to bring the surface in question to an approximately horizontal posi- tion. The observer now presses the component (b) against (a) and turns the screw head (p) in the reverse direction to contact with another stop, whereby (b) is secured against dropping back and close adhesion of the two prism faces is effected as well. The apparatus is at the same time replaced on its base plate. The mirror (j) should be arranged in such a position that the border-line appears distinctly, which may necessitate slight shifting or turning of the entire apparatus. The draw at the ocular should also be adjusted so that the scale can be seen distinctly. The first thing necessary is to ascertain that the entire space between the prism surfaces is uniformly Saponification Value 21 packed with fat. For that purpose the small image of the prism surface situated about 1 cm. in front of the ocular is scanned with a magnifier held at the requisite distance from the ocular (or with the naked eye). In this way minute air-bubbles within the stratum of fat\ which would prejudice the sharpness. of the border-linej will be readily detected. If a current of water of constant temperature has previously circulated for a time through the prism body, the border-line will quickly — generally in about a minute — assume a fixed position and attain the. maximum sharpness. The. position of the border-line relatirely to the scale is noted^ and at the same time the tem- perature registered by the thermometer is read off. Integral divisions of the scale are read off immediately in the field of view, tenths of a division being deter- mined by the aid of the micrometer screw (g in Fig. 3) in the following manner ; The border-line is adjusted by means of the screw upon a division of the scale, when the micrometer drum will indicate the number of tenths to be added to the number of integral jjarts of the scale appertaining to the particular division. The use of sodium light illumination greatly sharpens the border-line. A standard fluid is provided for testing the adjust- ment of the ocular scale and directions are given for re-setting the scale when necessary. CHEMICAL TESTS. (a) Saponification Value. Definition. — The saponification value is defined " as the number of milligrammes of potassium hydroxide whicli are required to completely saponify 1 grm. of the substance." 22 Fatty Foods When corrected for the acid value (see below) we obtain the ester value of the fat^ which is in fact the true saponification value of the neutral glycerides present. From this ester value the following deductions may be made : Since 56"1 grm. of KOH saponify one third of the molecular weight of the fat in grammes, it follows that the mean molecular weight of the fat under examination can be determined directly from the saponification value, and in a similar way the mean molecular weight of the fatty acids. It follows also that the ester value is a measure of the quantity of glycerine combined in the fat as glycerides, and the quantity of glycerine may be approximately arrived at by the following formula : Ester value x 0'05466 = percentage of glycerine, as 1 grm. of KOH is equivalent to 0'54661- grm. of glycerol,, Also the mean molecular weight of fat = , w^here W = weight of fat taken, and y = c.c. of N/2 KOH for saponification, assuming the fat to consist of triglycerides. The saponification equivalent is the number of grammes of fat saponified by 56'1 {i.e. one equivalent) of- potas- sium hydroxide. It is obvious that this is mei-ely another mode of expressing the same result, and the one figure can easily be calculated from the other. Kottstorfer's original mode of expression (saponification value) is now more generally used. Method of carrying o^d the test. — Solutions required — (1) N/2 hydrochloric acid, accurately prepared. (2) Alcoholic potassium hydroxide, approximately N/2 strength. (3) One per cent, alcoholic solution of phenol- phthalein. Saponification Value 23 The potassiam hydroxide solution is prepared by- dissolving 17 to 20 grm. of stick potassium hydroxide (purified by alcohol) in the smallest possible quantity of water, and then making up to 500 c.c. with alcohol of not less than 94 per cent, (by weight). The solution is allowed to stand overnight, and the clear liquid syphoned oii for use. If the alcohol be pure the solu- tion will be colourless, or nearly so (it is advisable to test the alcohol before making up the solution, and to reject any which gives more than a very pale colour when boiled with a strong solution of sodium hydroxide). The test is carried out as follows : About 2 grm. of tlie clear filtered fat are weighed into a 200 c.c. Jena flask, and an accurately measured quantity of the alcoholic potassium hydroxide solution run in from a 25 c.c. pipette. A like quantity of the same solution is run from the same pipette, in exactly the same way, into a clean 200 c.c. fiask. The flasks are connected to reflux condensers and heated in a water-bath so that the alcohol boils gently for thirty minutes. The flask con- taining the fat should be shaken occasionally, particu- larly at the commencement. The flasks are -then removed from the bath, 20 drops of phenolphthalein solution added to each, and the contents titrated while hot with N/2 acid. If F = grm. of fat taken, X = c.c. of acid required in the control experiment, Y = c.c. of acid required to neutralise the excess of alkali in the test, (X-Y)x 0-02805x10 00, then the sapomjication value b= = 56100 and the saponification equivalent — — 5 — . Precautions. — Great care should be taken with the solutions and during the test to avoid absorption of 24 Fatty Foods carbon dioxide. A blank titration should be made with every set of tests. The end point of the titration should be ascertained as exactly as passible. (b) Iodine Value. The iodine value is defined as " the percentage of iodine absorbed by an oil, fat or wax." It is in fact a measure of the unsaturated bonds present in the substance under examination, and is often taken to bo entirely due to the presence of olein, but while this is true for a number of fats, there are many which contain other unsaturated glycerides, such as occur in linseed, arachis, cotton oil, etc., in which other more unsaturated glycerides are present. The same applies to the hroinine value, so that one may be calculated from the other. Several methods are employed for determining this value, and though by far the greater number of figures published have been determined by tlie Hiibl method, the authors now entirely employ the method of Wijs, which is not only much easier of application, but more rapid and reliable, and moreover the necessary reagent is stable. Like most good methods it has suffered from modification, but the original method is perfectly satis- factory. To carry out the test the following solutions are required : (1) JJecinormal sodium thinsuhphale. — Twenty-five grm. of the pure salt are dissolved in water, and made up to 1000 c.c. The solution, which will be found to be of alAiost the right strength, may be standardised against potassium bichromate solution in the following manner : 3'863 grm. of pure potassium bichromate (free from the sodium salt) are dissolved in water, and made up to 1000 c.c. Each c.c. of tliis solution will liberate 0-01 grm. of iodine (A.W. 126-92). Iodine Value 25 To standardise the tliiosulphate solution 10 c.c. of a 10 per cent, solution of potassium iodide are placed in a flask together with 50 c.c. of water and 5 c.c. of con- centrated HCl ; 20 c.c. of the bichromate solution are accurately measured in^ and the solution titrated with the tliiosulphate using a solution of soluble starch as indicator. The final green tint of the solution must not be mistaken for pei'sistence of the starch iodide colour. Since the 20 c.c. of bichromate solution will have liberated exactly 0'2 grm. of iodine, the value of the tliiosulphate solution in terms of iodine can be cal- culated. It may be noted that the solution of bichromate keeps indefinitely. • (2) Wijs' solutinn of iodine moiiocJiloride. — Thirteen grm. of iodine are dissolved in a litre of acetic acid (Kahlbaum, 99 per cent.) by warming on the water-bath, protecting the solution from access of moisture. The solution is cooled, and about 100 c.c. poured into another vessel. Dry chlorine gas is passed into the remainder (which is kept in gentle motion), until the deep brown colour of the solution changes to a clear dark orange, when the passage of the gas is stopped. The iodine solution which was kept back is now added until the colour of the solution on mixing again becomes faintly brown. In this way a slight excess of iodine over that necessary to produce iodine monochloride is caused to be pi'esent in the solution, and so any danger of the presence of iodine trichloride avoided. This solution is quite stable under ordinary conditions, and will not alter in titre during some months, its keeping powers being still further enhanced by heating the solution in a water bath for fifteen minutes. Dry chlorine gas is most conveniently prepared from bleaching powder by the action of dilute sulphuric acid. 26 Fatty Foods The gas production is easily controlled if the bleaching powder be first made into cubes by mixing the dry powder with twice its weight of plaster-of-Paris and moistening with sufficient water to produce a paste, which, after being allowed to set, is broken up. A sufficient number of these cubes are placed in a large flask, fitted with a sulphuric acid " bubbler " to regulate the pace of evolution of the gas. In order to avoid the preparation of chlorine, iodine trichloride is now prepared in JO-grm. tubes, and the contents of one of these tubes may be dissolved in about 300 c.c. of acetic acid, and a solution of 12 grm. of iodine in about 500 c.c. of the same acid poured in until the colour of the iodine solution, on mixing, is seen to be no longer absorbed by the trichloride. A slight excess of the iodine solution is then added as before, and the whole made up to a litre with acetic acid, and heated for fifteen minutes in the water bath. (3) Soluble starch solutiun. — About 1 grm. is dissolved in 100 c.c. of water by boiling, and the solution is then cooled. This solution must be made up fresh each daj", unless preserved by the addition of a few drops of carbon tetrachloride well shaken with the solution. Ordinary starch may be used, but is not so sensitive as the soluble variety. (4) Oarhon tetrachloride. — The ordinary commercial re-distilled product may be employed, but it should be previously dried over CaCl,. Method of carrying out the test. — For ordinary work 10 c.c. of the Wijs' solution is employed by us, and as this is a constant quantity, and the iodine absorption of fats varies, a suitable amount of the material to be examined must be weighed out, so that only about half the halogen content of the solution is utilised. If the iodine value of the substance be Iodine Value 27 not known, a preliminary trial should be made. The necessary amount (which will be between 0"1 and ]"5 grm.) is accurately weighed out into a clean dry flask, of about 200 c.c. capacity, 5 c.c. of carbon tetrachloride added, and the fat brought into solution. Ten c.c. of the Wijs' solution accurately measured, are run in, a stopper inserted, and the whole shaken round and allowed to stand for tjiirty minutes. Five c.c. of a 10 per cent, solution of potassium iodide are then added, followed by 50 c.c. of water, and the contents of the flask titrated with the standard thiosulphate solution, adding about 10 drops of the starch solution towards the end of the titration, which is then continued till the violet colour does not return on well mixing the contents of the flask. No notice should be taken of any colour which re-appears on standing for some time, this being due to decomposition of the iodine compound of the fat by light. A blank experiment is conducted in an exactly similar manner, using all the materials with the exception of the fat. From the figures so obtained, the iodine value is calculated by the following formula : (a-b) X F X 100 Iodine value = ^ ; where— a = c.c. of thiosulphate used for blank. b = c.c. of thiosulphate used for the test. p = weight of .iodine equivalent to 1 c.c. of thiosulphate solution, as determined in the bichromate standardisation. If the thiosulphate is strictly N/10, this will be 0-012692. w = the weight of fat used. Precautions. — The iodine solution must be most accurately measured, and on account of the high co-effi- 28 Fatty Foods cient of expansion of acetic acid, measurements must always be made at the same temperature. Titration should be carried to the nearest half drop. In the case of high iodine values, or where very great accuracy is required, the quantities given above should be in- creased. The presence of moisture or alcohol vapour must be carefully excluded during the expei'iment. (c) Acetyl Value. The acetyl value is defined as " the number of milli- grammes of potassium hydroxide required to neutralise the acetic acid obtained when 1 grm. of an acetylated fat or oil is saponified." Since many fats and oils contain glycerides possessing one or more hydroxyl groups not in combination with glycerine, it is possible to obtain a measure of their presence by replacing the hydrogen of these hydroxyl groups by the acetyl radical. At the same time any free alcohols such as glycerol, cholesterol, etc., will be included if present. Method of carrying out the test. — The form hero given is that devised by Lewkowitsch. Ten grm. or any convenient quastity of the material are boiled with about twice their weight of acetic anhydride fcr two hours under a reflux condenser. The solution is then poured into a beaker of about 1000 c.c. capacitjr and mixed with 500 to 600 c.c. of boiling water, which is kept boiling for thirty minutes, during which time a fine stream of carbon dioxide is passed in to prevent bumping. The boiling is then stopped and separation allowed to take place, the water is syphoned off, and the heating with water again repeated three times until all trace of acetic acid is removed, but too pro- longed washing is to be avoided. The acetylated fat Acetyl Value 29 is then filtered through dry filter-paper until free from water. About 5 grm. of this acetylated fat are saponified in an exactly similar manner to that employed in carry- ing out the saponification value. The alcoholic potash used must be accurately measured. After saponifica- tion the alcohol is distilled oS, and the soap dissolved in water. To the solution is added N/2 H^SO^ in quantity exactly eqiiivalent to the alcoholic potash employed, and the mixture is then gently warmed till the fatty acids separate completely as an oily layer. The acids are filtered off, washed with boiling water until the washings are no longer acid, and the total filtrate titrated with N/10 KOH to phenolphthalein. _ Number of c.c. N/10 KOH x 5-61. Weight of acetylated fat taken. It is recommended to use a slight excess, say 1"5 c.c. of the N/2 HjSO^ in order to cause the fatty acids to separate easily from the solution, and this amount must be allowed for in titrating the filtrate with N/10 alkali. The distilled water used in this process must be carefully freed from COj by boiling. True Acetyl Value. In the case of fats containing soluble fatty acids a correction has to be introduced, as a proportion, or in some cases all, of these will be washed out of the layer of fatty acids together with the acetic acid. The correction is easily determined by treating 5 grm. of the unacetylated fat in an exactly similar manner, employ- ing exactly the same amount of water for the final washing as is employed to wash out the acetic acid. The total volume to be titrated with N/10 alkali is to be the same in each case. The connection of this 30 Fatty Foods determination with rancidity is discussed under that heading. The following are a few values selected from those given by Lewkowitsch. The figures are true acetyl values, that is to say they have been corrected for the presence of volatile and soluble acids as stated above. Oil. True acetyl value. Castor .... 150-05 Japan wax . 27-31-2 Croton 19-82 Rape 14-7 Safflower 161 Cotton -seed 14-8-35-65 Maize 5-81 Coconut 0-9-12-3 Lard 2-6 (d) Bellier's Test for Arachis Oil in Liquid Vegetable Oils. The test is used as a qualitative method of determin- ing the presence of arachis oil in admixture with other liquid vegetable oils. The following solutions are required : (1) An alcoholic solution of potassium hydroxide, made by dissolving 4-25 grm. of potassium hydroxide (stick, pure by alcohol, contains about 87 per cent, of KOH) in 70 per cent, alcohol and made up to 50 c.c. (2) A solution of acetic acid, of which 1-5 c.c. will neutralise exactly 5 c.c. of solution (1). About 14-o c.c. of glacial acetic acid diluted to 50 c.c. will give about the correct strength. Method of carrying out the test. — In a test-tube Lead-Salt-Ether Method 31 (6 in. by 1 in.) are placed 3 grm. of the oil whicli is saponified by heating gently with 5 c.c. of solu- tion (1), till quite clear, shaking well and avoiding evaporation of the alcohol ; for which reason the tube is best held with the fingers. As soon as saponifica- tion is complete, 1"5 c.c. of the acetic acid are added and the tube cooled in water at 18° C. for not lefis than 30 minutes, with occasional shaking. To the solution are then added 50 c.c. of 70 per cent, alcohol containing I c.c. of strong hydrochloric acid per 100 c.c. The whole is mixed and placed in water at 18° 0. for one hour. If arachis oil be in the original sample a distinct precipitate is formed, even when 5 per cent, only is present, the volume of the precipitate increasing with increasing proportions. Other liquid vegetable oils only give at the most an opalescence and usually remain quite clear. As olive oil sometimes gives a definite precipitate at first, it is advisable to warm up the solution and again place in water at 18°C. Under these circumstances olive oil does not re-precipitate, while if arachis oil be present, the precipitate forms as before. The test is quite reliable for liquid vegetable oils, but breaks down if solid vegetable or animal fats be present. Coconut and palm kernel oils are exceptions among the solid vegetable fats. (e) Separation and Determination of Liquid Fatty Acids. (Lkad-Salt-Ether Method). On account of the much greater solubility of the lead-salts of the liquid fatty acids (which may be taken to be practically oleic, linolic, and linolenic acids), they 32 Fatty Foods may be separated in an approximately quantitative manner by the following method : 5 grm. of the sub- stance are saponified by boiling in a 300 c.c. flask with 2 grm. of solid caustic soda (the usual stick caustic soda, pure by alcohol; contains 87 per cent. NaOH), and 50 c.c. of strong alcohol under a reflux con- denser until saponification is complete. If preferred 2 c.c. of the usual caustic soda solution used for the Reichert-Meissl process may be employed for the saponification. Saponification being effected, the solution is made slightly acid to litmus paper with acetic acid, and titrated with an alcoholic solution of caustic soda till neutral to phenolphthalein. The alcohol is then distilled off and the soaps dissolved in 100 c.c. of water. To this solution are added gradually and with constant shaking, 200 c c. of a filtered 2'5 per cent, solution of normal lead acetate, previously raised to the boil. The flask, if not already full, is filled completely with hot water, and put aside to cool. When the lead soaps have become hard and cold, the liquid is decanted off through a filter, and any of the lead-salts left in the filter returned to the flask. The contents of the flask are washed two or three times with boiling water. Before pouring off the water the mixture is carefully cooled each time by a rapid rotation of the flask under a stream of cold water. The flask is finally turned upside down and held over the filter by means of a clamp, till as much of the water as possible has drained off. Should there be any particles of lead salts on the filter they should be carefully picked off and returned to the flask. Two hundred c.c. of ether are now added and the flask closed and thoroughly shaken, after which the contents are boiled under a reflux condenser with repeated shaking for thirty minutes. By this means the lead salts of the liquid fatty acids Lead-Salt-Ether Method 33 are dissolved out, the termination of the process being indicated by the separation of the insoluble salts in a pulverulent condition. The contents of the flask are allowed to cool, avoiding undue exposure to the air, and filtered into a separating funnel, the filter being kept covered as much as possible with a watch-glass. The insoluble salts in the flask are brought on to the 'filter with three or four successive quantities of 30 c.c. of ether. To the solution of the lead-salts in the ■ separating funnel are added 100 c.c. of water, and an excess of hydrochloric acid. The separating funnel is then securely stoppered and vigorously shaken until lead chloride ceases to settle out. The aqueous liquid and the precipitated lead chloride are now run off, and the ether washed with successive S7i;iall .quantities of water until the wash water is free from hydrochloric acid. The water having been separ- ated as completely as possible, a few small pieces of granulated calcium chloride are added to the ether and the whole well shaken, and allowed to stand for thirty minutes.. The ethereal solution is poured through the mouth of the funnel on to a small filter, placed in a weighed flask (taking care that none of the calcium chloride solution finds its way on to the filter), the separating funnel being carefully rinsed with a few c.c. of fresh ether, which are then poured on to the filter. The ethereal . solution 1 is then distilled off, and the resi- dual, fatty acids either dried in the water oven^ or if preferred, by placing the flask in hot water and attach- ing to a vacuum pump, but the vacuum employed must not-be more tlian twenty inches of mercury or collapse of the flask may ensue. The .liquid fatty acids being thus, obtained, their iodine value may be determined. ;It m^st,bp borne in mind that these liquid fatty acids 3 34 Fatty Foods are susceptible to oxidation, and that all reasonable care should be taken with them to avoid such oxidation. When dealing with drying oils, such as may occur in foodstuffs (for instance linseed, rubber-seed, sunflower, and perhaps maize and rape oils), it will be necessary to distil off the final ethereal solution in a current of hydrogen or carbon dioxide. The following formulae and iodine values will be of use in resolving mixtures when the iodine value of a known weight of the fatty acids is estimated : Per cent. Oleic acid .C^8H34,O2(M.W. = 282)Iodinevalue=90 LinoUc acid .CigH32O2(M.W. = 280) „ =181 Linolenic acid . CigHgoOg (M.W. = 278) „ ( J" ^^^ Rieinoleic acid. G^s'R^fis {MM. = 298) „ 5 Two modifications of the lead-salt-ether method are npplied to the detection of arachis oil and rape oil respectively. They are carried out as follows — (1) Renahd's Test joe Arachis Oil (with Archbutt's modifications). — Ten grm. of the oil are saponified by heating in a dish with 2 grm. of caustic soda and 100 c.c. of strong alcohol, keeping a funnel over the dish for the first fifteen minutes, after which the solution is allowed to boil down to about 20 c.c. The contents are then rinsed into a separator with hot water, and while hot, the soaps are decomposed with hydrochloric acid and the fatty acids extracted with ether (after well cooling the mixture), and the ether washed free of hydro- chloric acid. The ether is then run into a flask and dis- tilled off and the fatty acids dried in the water-oven. The fatty acids are then dissolved in 50 c.c. of 90 per cent, alcohol and cooled to between 38°— 42° C, and while at this temperature, 5 c.c. of a 20 per cent, solution of lead Renard's Test 36 acetate are added. The mixture is then cooled to 16° 0. and allowed to stand for thirty minutes, when the alcohol is carefully decanted off through a filter and the lead soaps washed once on the filter with ether. The filter is then pierced and the lead soaps washed back into the flask with ether and the whole well shaken. The ether is then J)Oured off through a filter and the soaps after draining again washed back into the flask with ether (after piercing the filter as before) and shaken well and again filtered. This is repeated about three times more, and the soaps being then brought entirely on to the filter, they are washed with fresh ether till the washings only give a faint brown colour when shaken with sulphuretted hydrogen water. (This rather complicated washing is avoided by pouring the original alcoholic mixture, containing the lead soaps, into a large plaited filter (15 cm.), and after the alcohol has drained, placing this in a Soxhlet extractor and extracting with ether till no colour with HgS water is given.) The soaps on the filter are washed, while still moist, into a separator with a jet of ether (if necessary followed by 20 c.c. of dilute hydrochloric acid and again with ether), and then decomposed by adding about 20—25 c.c. of hydrochloric acid (sp. gr. 1.10), and shaking well till lead chloride ceases to separate. The lower layer is then run off and the ether washed with water till free from lead chloride. The ether is mixed with two or three pieces of granulated calcium chloride and allowed to stand for thirty minutes, after which the ether is poured into a flask and the separator washed out with a few c.c. of ether, taking care not to pour out any drops of water. The ether is then dis- tilled off and the fatty acids dried in the water oven. They are then dissolved in 50 c.c. of 90 per cent, alcohol and the solution cooled till crystals separate. 36 Fatty Foods The flask is then closed with a cork^ fitted with a thermometer^ the solution warmed till clear, and the temperature of crystallisation observed, when a very good idea of the amount of arachis oil in the original sample will be obtained, using for this purpose the foUowingtable of Tortelli and Ruggeri (except when solid vegetable or animal fats — with the exception of coconut products — -are present, when the results are fallacious). Temperature of crystallisation. Arachis oil per cent. 35-38° 0. . . . .100 31-33° C. . . .60 28-30° C. . 50 25-26° C. . . .40 22-24° C. ... 30 20-5-21-5°C 20 18-20° C. . . . 10 16-17° 0. ... 5 The acids must then be further purified in order to obtain the melting-point. For this purpose the liquid is allowed to stand for about one to three hours at the temperature of the room (15°— 20° C), and the crystals then filtered off on a small filter or preferably on a Gooch. crucible, the mass being washed two or three times with 90 per cent, alcohol and then with 70 per cent, alcohol till the washings give no turbidity on dilution with water. The crystals are then dissolved off the filter with hot ether, the ether evaporated and the residue dissolved in sufficient 90 per cent, alcohol and allowed to crystallise. The crystals are filtered off and again crystallised from 90 per cent, alcohol. A small portion is then placed on a fragment of porous plate, and when dry the melting-point is taken. The melting-point slrould be between 72° C. and 73"5° C, and the point of incipient fusion should not be below 71° C. The method is made quantitative as follows : Renard's Test 37 The original solution of the acids, as obtained from their lead salts, in 50 c.c. of 90 per cent, alcohol, is allowed to stand at laboratory temperature (if possible not above 20° 0.) for three hours. The crystals are then filtered off, collecting the filtrate in a measuring cylinder, washing out the flask on to the filter with this filtrate. The crystals are washed three times with 5—10 c.c. of fresh 90 per cent, alcohol, receiving the filtrate each time in a test-tube and pouring back over the crystals two to three times (in order to be sure that the alcohol is saturated with the acids), the alcohol used being at the room temperature. These filtrates are all poured finally into the measuring cylinder. The crystals are then well washed with 70 per cent, alcohol,^ and dissolved out into a tared flask with hot ether, dried and weighed. Their purity must be confirmed by the melting-point. If below 71° C, the treatment must be repeated. The weight obtained must be corrected for the quantity of 90 per cent, alcohol used in washing, as the acids are distinctly soluble in this alcohol. The following correction table is due to Tortelli and Ruggeri : Weight of mixed acids obtained (grm.J . Correction (grm.) to bemadded per 100 c.c. of 90 per cent, alcohol used for crystallisation and washing at— 15° 17-5<> ,20° 0-05 + O'OSl + 040 + 0-046 1 0-10 0036 0-045 0-052 0-20 0-048 0-056 0-062 0-30 0-055 0064 0-071 0-40 0-061 0-071 0-078 0-50 0-064. 0-076 0-084 0-60 0066 0-080 0-088 0-70 0-067 0-082 0-090 0-80 0069 0-083 0-092 ■ •0-90 - 0070 0-084 . 0-092 1-00 0-071 0-084 0-091 2-70 0073 0-082 0091 ' These washings are thrown away. 38 ' Fatty Foods Pure arachis oil usually gives about 4"8 per cent, of mixed arachidic and lignoceric acids by the above methodj so that the quantity obtained multiplied by 21 is taken to give the percentage of arachis oil in the original sample. As with Bellier's test, this method gives excellent results as long as vegetable liquid oils are being dealt with ; when solid vegetable or animal fats (with the exception of coconut products) are present, the qua.nti- tative value of the test fails, but by repeated re- crystallisations of the acids from 90 per cent, alcohol, it is possible to determine the presence of arachidic acid by melting-point determinations, if a fair propor- tion of arachis oil was present in the original mixture. Failure to obtain crystals melting at 71°C.must not under these circumstances be taken to exclude arachis oil. (2) TOETELLI AND FoilTINl's M0DIJ?1CATI0N AS APPLIED 'J O THE Detection op Eape Oil in Mixtuees. In carrying out this test the following details must be adhered to exactly. Twenty grm. of the oil are saponified with 6 grm. of potash dissolved in 50 c.c. of 90 per cent, alcohol by heating under a reflux condenser. The liquid is then neutralised to phenolphthalein with 10 per cent, acetic acid, and the solution then slowly poured into a boiling mixture of 200 c.c. of 10 per cent, lead acetate and 100 c.c. of water, shaking vigorously during the addition. The mixture is then cooled under the tap, maintaining a rotary motion until the soaps begin to stick to the sides (if they do not stick at first, they will do so during the first washing). The water is poured off and the soaps washed three times with 200 c.c. of warm (60 to 70° C.) water, the water being then drained off and the last drops removed with filter-paper. To the dried soaps are added 80 c.c. of Tortelli and Fortini's Method 39 methylated ether and the whole is well shaken for several minutes till the mass is broken up, when it is heated under a reflux condenser for thirty minutes, shaking at intervals. The flask is then closed and placed in water at exactly 15° G. for one hour, after which the contents of the flask are poured on to a filter, the funnel being placed in the mouth of a separator, and the filter closely covered till all the ether possible has filtered out. The filter and contents are dropped back into the flask and the ether treatment (boiling and cooling) repeated in exactly the same way, using 40 c.c. of ether, and the mass filtered as before, again tightly covering the filter and allowing to drain as completely as possible. The flask is then washed out with a further 40 c.c. of ether on to the filter, the contents of which are well stirred up with the ether, which is then allowed to drain off. The combined ethereal solutions of the lead salts so obtained are then decomposed in the separator by shaking twice with 150 c.c. of 10 per cent, hydrochloric acid, after which the ether is washed with two quantities of 100 c.c. of water, the ethereal solution being then run out into a dry flask and allowed to evaporate spontaneously or by the use of gentle warmth in a current of hydrogen. The liquid fatty acids so obtained are dissolved in 40 c.c. of strong alcohol (97 per cent.) and a saturated solution of sodium carbonate added until the liquid is saturated (Na^COg separates). The alcohol is then distilled off and the residue dried, first in the water oven, distributing it as much as possible over the sides of the flask, and finally in a vacuum desiccator for at least forty-eight hours. The dry sodium soaps are then boiled with 50 c.c. of absolute alcohol and filtered in a hot funnel, the insoluble residue being boiled with a further quantity of alcohol and the treatment repeated till nearly the whole has been 40 Fatty Foods dissolved. The mixed alcoholic filtrates are freed from alcohol by distillation and the sodium soaps dried as completely as possible in a vacuum desiccator over sulphuric acid. According to Tortelli and Fortini the test is concluded as follows : 05 grm. of the dry soaps are placed in a large test- tube and dissolved by heating in 20 c.c. of absolute alcohol. A thermometer is then placed in the mixture, which is allowed to cool and the turbidity temperature noted. The following table gives some results obtained by them : Oil. Turbidity temperature. Olive . . . 20-24° C. Rape . . . 45-50° C. 1 pt. Olive ) 35_^QO c. 1 pt. Rape ) 8 pts. Olive } on qeO n c 2. -D i ■ ■ 30-35 O. 2 pts. Rape ) 9 pts. Olive J 30-34° C. 1 pt. Rape 3 Cotton . . . 14-16° C. Sesame .... 18-20° C, Arachis .... 18-22° C. The authors find it more satisfactory to dissolve 0"75 grm. of the soaps in strong alcohol (97 to 98 per cent.) and to leave the solution to stand at a temperature of 20 C. Under these circumstances rape oil commences to precipitate in a granular form in fifteen to thirty minutes, and 5 to 10 per cent, of rape-oil in admixture with other fats produces a spongy gelatinous precipitate within two hours, while in the absence of rape oil no precipitate usually forms under fifteen to eighteen hours. Acid Value 41 As the results are, dependent on the degree of dryness of the soaps and the strength of the alcohol employed it is more satisfactory to carry through the test with some oil, such as cotton-seed oil, as a control. The test under these conditions is quite reliable. (/) Acid Value. The acid value of a fat or oil is a measure of the free fatty acidity, and is defined as " the number of milligrammes of potassium hydroxide required to neutralise the free fatty acids in one gramme of the fat or oil!" Method of carrying out the Test. — About 5 grm. of the fat or oil are weighed into a 250 c.c. flask, and 30 to 50 c.c. of 95 per cent, alcohol (which has been just previously boiled, and rendered faintly pink to phenolphfchalein by the addition of N/10 NaOH) added. ■ The mixture of fat and alcohol is brought to the boil (the oil does not usually dissolve, but the solution of the free fatty acids will have been effected) and titrated hot with N/10 NaOH till a faint pink colour, permanent on gently mixing the contents of the flask, is attained. The acid value is calculated by the following formula : Acid value = , W where N = number of c.c. of N/10 NaOH used. W = weight of fat taken. This formula gives the acid value in terms of the definition given above, but the acidity is for other pur- poses sometimes returned as " per cent, of oleic acid," in which case the above formula becomes — 42 Fatty Foods N X 2-82 Oleic acid per cent. = ^ < where 282 is the molecular weight of oleic acid. In the case of coconut oil (particularly for soap making) it is now becoming the custom to return the acidity as lauric acid, as this figure represents very nearly the exact weight of the free fatty acids present. In this case the formula becomes : Lauric acid per cent. = ^^, W the molecular weight of lauric acid being taken as 200. Various other methods of calculation are also employed ; for example, on the continent the Kottstorfer value is often given, this being simply the number of CO. of N. KOH required, using 100 grm. of fat. The relationship of these methods of returning the acidity may be found from the following proportion : Acid value. Oleic acid. Laurie acid. Kottstorfer value. 1 : 0-5027 : 0-357 : 1-782 (g) Hexabromide Test (Halphen's Modification). This test furnishes a method of distinguishing quali- tatively between the fish, marine animal and drying oils on the one hand, and the non-drying oUs on the other. It is useful for detecting the presence of the former in the latter. Method of carrying out the test. — In a small dry test-tube are placed 0-5 c.c. of the oil and to this are added 10 c.c. of a mixture of twenty-eight parts by volume of glacial acetic acid, one part of bi'omine and four parts of nitrobenzene. The tube is then closed, the contents shaken and the behaviour of the liquid noted. Such non-drying oils as cuconut ■products, ulive, castor, Hexabromide Test 43 arachis, cotton, sesame, poppy-seed, etc., give no precipi- tate or only a very slight turbidity, even after standing for one hour. The d/rying oils, such as linseed, ruhher- seed, hemp-seed, etc., arid marine animal and fish oils, such as whale oil, Japanese sardine oil, etc., give a pre- cipitate which settles more or less rapidly to the bottom of the tube, the precipitate from whale oil being much heavier than from fish oils and the drying oils. Rape oil is distinguished by forming a turbidity, and the liquid separates on standing into two distinct layers, but the mixture resolves into a clear homogeneous liquid on mixing with an equal volume of methylated ether. As little, as 5 per cent, of the drying and marine animal and fish oils can be detected by this means in admixture with such liquid vegetable oils as give no precipitate themselves. When examining rape oil for these; the test should be mixed with 10 c.c. of methylated ether, as noted above, when, if only rape oil be present, the liquid will clear, but if the drying or marine animal oils be present, a precipitate will remain and slowly settle down. Ten per cent, of rape oil in liquid vegetable oils can be detected by the production of an oily layer and the clearing of the mixture on the addition of ether. It is usually stated that such animal fats as beef and lard give no precipitate, but the authors have found that beef and lard fats and also shea butter all give precipitates with this test. Beef and lard fats give a flocculent precipitate which eventually rises to the top of the tube and is not dissolved by ether. Butter fat as a rule behaves in a similar manner, but the precipitate is usually dissolved by ether. Shea butter gives a precipitate which settles to the bottom of the tube and is insoluble in ether. 44 Fatty Foods (h) Phytosteryl Acetate Test and Unsaponifiable Matter. Animal and vegetable fats are both characterised by containing very small quantities of a complex alcohol. In the case of animal fats this is called cholesterol, Cg^H^OHj and apparently the same substance is found in all animal fats. Isomeric alcohols called phytosterols have been isolated -from vegetable fats and oils, though apparently these are not always identical, yet very similar in properties. These two alcohols differ in their melting-points (cholesterol 147° C, phytosterol 137° C.) and differ somewhat in their method of crystallisation from alcoholj but are most easily distinguished by means of their acetates, which can be easily prepared, and, on account of their difficult solubility in 95 per cent, alcohol, purified by recrystallisation. The melting-points of these acetates are sufficiently far apart to detect admixture of one with the other, and on account of the greater solubility of cholesteryl acetate in 95 per cent, alcohol the presence of small quantities of phytosteryl acetate can be detected in the presence of the other by the steady rise of the melting-point on repeated crystallisation of the mixture. This property provides a delicate method of detecting the presence of vegetable fats in animal fats when no other evidence is forthcoming. The acetates are readily prepared from the sample by the following method : Fifty grm. of the filtered fat are boiled with 75 c.c. of 95 per cent, alcohol under a reflux condenser. The flask is then cooled, the alcohol poured off into a weighed flask, and the fat again boiled up with another 75 c.c. of alcohol, after which the flask is cooled and the alcohol again poured off. The alcoholic extract of Phytosteryl Acetate Test 45 the fat so obtained will contain practically the whole of the ' unsaponifiable matter, and some fat. The alcohol is distilled off, and the flask re-weighed in order to get a rough idea of the weight of the fat present. To every 5 grm. of fat are added 2 grm. of NaOH and 50 c.c. of alcohol and the mass saponified. The flask is tlien washed out into a porcelain basin, and the whole evaporated, stirring occasionally. When practically all the alcohol has been evaporated, twice as much sodium bicarbonate as NaOH used are added and a couple of small spoonfuls of fine sand, and the mixture well stirred and carried to dryness. The dry residue is thoroughly ground, dried in the water oven, and then extracted for three to four hours in a flow-through extractor with petroleum ether free from residue, and the unsaponifiable matter obtained by distilling off the ether. The nature of the unsaponifiable matter may be roughly ascertained by its behaviour on treat- ment with acetic anhydride. For this purpose the residue from the ether is dissolved in the smallest possible quantity of hot absolute alcohol, and washed into a small (1 oz.) well-stoppered bottle, the flask being rinsed out once or twice with small quantities of hot absolute alcohol. The alcohol is evaporated off, and to the residue left in thie bottle are added 2 to 3 c.c. of pure acetic anhydride. The stopper is then carefully tied in and the bottle heated in boiling water for fifteen to thirty minutes. During this process the behaviour of the solution is carefully noted. If after fifteen minutes' heating the solution, when cool, remains clear, the presence of anything but traces of hydrocarbons is negatived, but if the solution shows the presence of an oily layer, it may be taken as almost certa,in that parafiin wax or some mineral oil is present ■ in the original substance. In this case it will be necessary to 46 Fatty Foods estimate the amount of these substances by the method given below. If, however, the acetic anhydride solution be clear, the bottle is heated for a further ten to fifteen minutes. The stopper is then carefully removed and the anhydride allowed to distil by heating in an oil bath at 140°— 150° C. The contents of the bottle are dissolved in a few c.c. of absolute alcohol, and if the solution be coloured, mixed with about 1 grm. of very finely ground recently ignited animal charcoal, the stopper again inserted and the bottle heated in boiling water with repeated shaking until the liquid is seen to be nearly decolourised. The con- tents of the bottle are then filtered into a small beaker, rinsing the char- coal and filter two or three times with small quantities of boiling absolute alcohol. (In those cases in which the first alcoholic solution of the acetate is practically colourless, the above procedure is unnecessai'y.) The contents of the beaker are evapor- ated to dryness, the residue dissolved in the least possible quantity of hot 95 per cent, alcohol, and the solution allowed to crystallise. The liquid is poured off into the small funnel shown in the illustration, protect- ing it from evaporation by a watch-glass, and the crystals washed with a few drops of cold 95 per cent, alcohol, which is then poured over the crystals in the funnel. The contents of the funnel are washed back Fio. 4. Phytosteryl Acetate Test 47 into the beaker with a few drops of hot 95 per cent, alcohol, and the crystals dissolved by heating, and the solution again allowed to crystallise. The liquid is filtered off as before, and a few of the crystals spread on a porous plate and their melting-point determined. The contents of the beaker and funnel are washed again with a few drops of 95 per cent, alcohol and the recrystallisa- tion and washings repeated three times more, a few crystals being removed each time, dried on porous porcelain, and the melting-point taken. If the melting- points of the foiarth and fifth crystallisations are approximately the same, no further crystallisation will be necessary. If the melting-point be below 116° C. (corrected)^ the presence of phytosterol may be negatived,^ but if the melting-point lies higher than this,: and tends to rise somewhat on further crystallisa- tion, phytosterol may be assumed to be present with certainty. Should the temperature be 120° 0. (corrected) or above there will be no doubt as to the presence of phytosterol. The melting-point of cholesteryl acetate is 113° 0. (corrected), and of phytosteryl acetate 125°---133° 0., dependent on the kind of vegetable fat used. The above process requires some care in carrying out, but if the above directions are carefully followed, no difficulty will be experienced. In those cases in which observation of the acetic anhy- dride solution has shown the presence of hydrocarbons, or wax, the percentage of these may be ascertained with sufficient accuracy in most cases by determining the unsaponifiable matter in the following manner : , : ' Stem Correction for Thermometer. Ti = T + n (T-t)a; where T = indicated temperature, t ^ room temperature, n = number of degrees exposed at t°, T'' = true temperature of bath, ; a.^ co-efEcient of expansion of mercury in glass (0-000154.) 2 See pp. 48-49. 48 Fatty Foods Determinatcon of Unsaponifiable Mattee. Five grm. of the fat are weighed into a large porcelain basin and heated with sufficient strong aqueous solution of caustic soda to introduce 2 grm. of NaOH and 100 c.c. of strong (95 per cent.) alcohol, the dish being covered with an inverted funnel. The dish is heated on a boiling water bath with occasional stirring for thirty minutes, after which the contents are evapo- rated to a pasty condition, when a teaspoonful of sand and 2 to 3 grm. of sodium bicarbonate are added and the whole carefully mixed and carried to dryness. The contents of the dish are then well ground, thoroughly dried and carefully transferred to a flow-through ex- tractor, and extracted for two to three hours with petroleum ether. The contents of the extractor are dried, removed to a mortar, again thoroughly ground, dried in the water bath and extracted as before for another hour. The contents of the flask are then evaporated, dried and weighed as unsaponifiable matter, the percentage, when anything more than traces of wax or hydrocarbons are present, being taken to consist of those substances. As small quantities (less than 0"01 per cent.) of paraffin , wax are sometimes added to lard in order to defeat the detection of vegetable oils by the phytosteryl acetate test, it is necessary to treat the unsaponifiable matter in such cases by a method due to Polenske in order to circumvent this form of sophistication. This method has been very carefully investigated by Lewkowitsch. He points out that if the paraffin wax be as much as 10 per cent, of the unsaponifiable matter, a minute drop of paraffin wax will be seen floating on top of the acetic anhydride during acetylation, and the acetylated mass on successive crystallisations will show an abnormal and Unsaponifiable Matter 49 probably decreasing melting-point. In a case cited by him^ the melting-point of the second crystallisation of a mixture of 90 per cent, cholesterol and ,10 per cent. paraffin wax was 90° to 99° 0., and of the fourth crys- tallisation 86° to 92° 0. A mixture, however, of choles- terol, with small quantities of phytosterol together with about 7 per cent, of their weight of paraffin wax, gives melting-points for the acetates which can scarcely be dis- tinguished from that of pure cholesteryl acetate. If, there- fore, such small quantities of paraffin wax are suspected, the unsaponifiable matter from at least 100 grm. of fat should be prepared as above described, and decolourised by dissolving in absolute alcohol, and heating under a reflux condenser with a small quantity of finely ground and recently ignited animal charcoal, filtering and washing the charcoal with a very little absolute alcohol, the clear filtrate being then evaporated to dryness. The unsaponifiable matter so obtained from 100 grm. of fat is shaken with 1 c.c. of light petroleum ether for twenty minutes at 15° to 16° 0. The mixture is then poured into a small funnel plugged with a small piece of glass wool, and the undissolved matter in the funnel washed five times with 0'5 c.c. of the ether. The residue on the filter is then acetylated as described above, and the melting-point taken after, the necessary recrystallisations. To determine the amount of wax the ethereal filtrate is evaporated to dryness and heated with 5 'C.c. of pure concentrated sulphuric acid at 105° 0. for at least an hour. The wax is then extracted by shaking the diluted acid with petroleum ether and weighed. According to Polenske it is preferable to repeat the treatment with sulphuric acid if the amount of paraffin wax, is very small... The method, in. spite of the apparent complications, is capable of great accuracy if i due care be exercised in carrying out the various details. 4 50 Fatty Foods Lewkowitsch has also shown that further valuable information may be obtained by acefcylating the unsapon- ifiable matter in the usual way, and saponifying the weighed mixture of acetates and paraffin wax with alcoholic potash, so as to obtain their saponification value. The saponification value of cholesteryl acetate being taken as 135'5, the amount of paraffin wax in the mixture which was saponified may be calculated by the deviation from this value, and the process allows of both the cholesterol and the paraffin- wax being recovered and dealt with, if desired, by Polenske's method as given above. The Separation op Cholesteeol and Phytosteeol. While the above process of recrystallising the mixed acetates is in nearly all cases sufficient to demonstrate the presence of even small quantities of phytosterol in admixture with cholesterol, the reverse is not true, and the necessity may arise when lard or beef fat is suspected in hard vegetable fats. In order to demonstrate the presence of a small quantity of cholesterol in admixture with phytosterol it is necessary to resort to a rather complicated process of separation, which is not likely to be successful unless the unsaponifiable matter is obtained from nearly 1000 grm. of the substance. The sample is treated as described above (p. 48), pre- ferably in lots of 200 grm., and the alcoholic extracts united. The final dry mixture of soaps, sand and un- saponifiable matter is extracted most simply by heating it in a flask under a reflux condenser with sufficient petro- leum ether to rather more than cover the solid matter. After boiling gently for one hour the ether is filtered o£E and the solid matter dried and re-ground and boiled twice Cholesterol and Phytosterol 51 more with petroleum ether. The combined ethereal extracts are then distilled to obtain the unsaponifiable matter. This is carefully shaken with a very small quantity of petroleum ether, increasing the amount until about one quarter of the total remains undissolved (choles- terol is more soluble than phytosterol). The ether solution is filtered and distilled off, and the residue is treated by the method of Windaus (Ohem. Zeit., 1906, xxx, p. 1011). To each gramme of the residue are added 10 c.c. of dry methylated ether and solution effected. To the solution are added 10 c.c. of a solution of 5 grm. of bromine (dry) in 100 c.c. of glacial acetic acid and the mixture cooled in ice for an hour. If only small quantities of cholesterol are present the. dibromide will not separate, in which case 50 per cent, acetic acid is added until a permanent turbidity appears. A further 2 to 3 c.c. of the 50 per cent, acetic acid are then added, and after a short time the precipitate filtered off on a very small hard filter and washed with a few drops of the 50 per cent, acetic acid. The filter-paper and contents are dropped into a small flask, 5 c.c. of glacial acetic acid and about 0"25 to 0"5 grm. of zinc dust added, and the whole boiled under a reflux condenser for two hours. The contents of the flask are then washed into a small separator with 20 c.c. of water and finally with 10—20 c.c. of methylated ether and the whole shaken. The aqueous layer is run off and the ether washed three times with a little water and filtered (if necessary) into a test-tube and the ether boiled off. The residue is dried in the water oven and then washed out with a very small quantity of absolute alcohol into a small evaporating basin and evaporated to dryness. The residue is covered with 1—2 c.c. of acetic anhydride, and the basin closed with a clock glass and heated on a water bath for thirty minutes. The acetic anhydride is 52 Fatty Foods then distilled off, and the residue, if much coloured, washed with absolute alcohol into a test-tube and boiled with a little recently ignited powdered animal charcoal. The mixture is filtered (washing the charcoal with a little absolute alcohol) into a small beaker, evaporated to dryness and recrystallised from the least possible quantity of 95 per cent, alcohol ; the crystals are separated, recrystallised from 95 per cent, alcohol, and the melting-point taken, and further recrystallisa- tions carried out if possible. The crystals, if cholesteryl acetate, should melt below 116° C. Twenty per cent, of cholesterol in the presence of phytosterol can be found by this means, and probably 10 per cent, may be found with care when a large quantity of material is employed to start with. Colour Tests for Oils. Of the many colour tests which have been described for the detection of certain oils the following are alone of any importance. Halphisn's Test foe Cotton-Seed Oil and Cotton- Seed " Steaeine." The test is carried out as follows : 2"5 c.c. of the liquefied sample are dissolved in an equal quantity of amyl alcohol in a strong test-tube, and 2"5 c.c. of a 1 per cent, solution of sulphur in carbon disulphide added. The tube is then securely corked with a sound cork, which is well tied down with a piece of linen. The tube and its contents are placed in a boiling water bath for exactly thirty minutes. In the presence of cotton-seed oil or cotton-seed " stearins' " a characteristic crimson colour is produced, and in this way as little as 1 per cent, of these oils may be detected. An approximate idea of the amount of cotton-seed oil Halphen's Test— Baudouin's Test 5B present may be obtained by comparing the colour developed, with that of standard tubes of mixtures containing known quantities of cotton-seed oil which have been treated in an exactly similar manner. The gradation of the colour is distinct up to 20 per cent, of cotton-seed oil, and if this limit be reached in the test the original sample should be diluted with some indifferent oil in order to bring the colour shade within the desired limits. It is quite satisfactory to keep a series of standai'd tubes for comparison, but under these circumstances the comparison should not be made until after twenty-four hours' standing, and both test and standard should be warmed to such a temperature that the contents are entirely fluid. It should be noted that all specimens of cotton-seed oil do not give the same intensity of colour, and some oils which have been highly refined by certain modern processes produce very much less colour than the original oil. For this reason the quantitative indication of this test must be accepted with reserve. If an indefinite reaction be obtained, the test may often be rendered more certain by applying it to the fatty acids. It must be remembered that hapok oil is the only known oil that gives a colouration with this test similar to that given by cotton-seed oil. In order to distinguish these two oils recourse must be had to Becchi's test as modified by Millau (p. 54). Batjdodin's Test foe Sesame Oil- The test is carried out as follows: Ten c.oof the melted fat are placed in a test-tube and two drops of a 2 per cent, alcoholic solution of furfural added, together with 10 o.c. of concentrated hydrochloric acid. The tube is then well shaken, when in the presence of sesame oil a crimson, 54 Fatty Foods coloii/r will be developed. The test is generally sensitive to 1 per cent, of sesame oil. It must not be forgotten that many of the artificial colouring matters used in butter and margarine give a similar colour when shaken with hydrochloric acid. It is necessary, therefore, when examining these fats for sesame oil to carry out a blank test, only omitting the furfural, any preponderance of colour in the test being attributed to sesame oil. The alcoholic solution of furfural keeps indefinitely in the dark. If furfural be not available, the oil should be shaken with 10 c.c. of concentrated hydrochloric acid which has been saturated, by shaking in the cold, with finely powdered cane-sugar. The above reaction is not given by any other known oil, and for this reason sesame oil has been made an obligatory ingredient of margarines intended for use in most Continental countries. Becchi's Test. This well-known test has been superseded by Halphen's test for cotton-seed oil. The modification of the test devised by Millau is, however, of use to dis- tinguish kapok oil from cotton-seed oil when applied in the form recommended by Durand and Baud. The test is carried out as follows : Fifteen c.c. of the oil are saponified with caustic soda and alcohol in the usual manner, 200 c.c. of boiling water are added and the whole boiled till the alcohol is evaporated. The fatty acids are then thrown out by the addition of N/IO H3SO4 in slight excess. The fatty acids are skimmed off, and shaken twice with 15 c.c. of cold distilled water, the water being then drained off and the fatty acids dried rapidly in an oven at 105° C. Five c.c. of these fatty acids are shaken with 5 c.c. of a 1 per cent, solution of silver nitrate in absolute alcohol. Rosin 65 Under these circumstances cotton-seed oil only pro- duces a barely perceptible brown colour, while kapok oil rapidly develops a deep coffee coloration. By means of this test it is possible to recognise 1 per cent, of kapok oil in other liquid oils. Rosin. This substance ought not to appear in edible products^ but as occasion may arise which may necessitate its detection and estimation, the methods are here de- scribed. (a) Liebermann-8torch Reaction. A few drops of the sample are boiled with 2 or 3 c.o. of acetic anhydride in a porcelain basin, and the un- dissolved fat separated by filtration after cooling. The filtrate is placed in a porcelain basin, and sulphuric acid (sp. gr. 1'53) is allowed to drop gently on to the surface of the liquid. If rosin be present a characteristic violet colour is produced, which fades to a brown. This test will detect very minute amounts of rosin, but it should be borne in mind that cholesterol gives a similar colouration, though, in general, the small amount of the latter present in fats will not be sufficient to vitiate the test. [b] Quantitative Determination of Rosin by Twitchell's Process. This process will be found to give the most reliable results. About 2 grm. of the sample are dissolved in 25—30 c.c. of absolute alcohol and dry hydrochloric acid gas is passed into the solution, which should be kept cooled in water. After the gas has been passed for about an hour, the completion of the operation will be 56 Fatty Foods shown by the escape of unabsorbed gas from the liquid. The gas is then stopped and the flask is allowed to stand for another hour. About 150 c.c. of water are now added, together with a piece of pumice, and the whole boiled till quite clear. The contents of the flask are cooled, and the oily layer, consisting of neutral fat, ethylic esters, and the rosin acids, is dissolved by vigorously shaking with ether in a separator, the flask being carefully washed out several times with ether. The ether is allowed to separate, the aqueous layer drawn off and the ether washed with successive portions of distilled water until free from hydrochloric acid. A volume of alcohol equal to that of the ether is poured into the separator, and the contents titrated therein with N/2 KOH to phenolphthalein. Then : „ . ^ c.c. of ]Sr/2 KOH X 34-6. Kosin per cent. = Weight of the substance x 2. This formula assumes 346 as the combining equivalent of rosin, which of course is not necessarily true for all forms of rosin, but the results obtained by the above method are sufficiently accurate for the practical purposes of this book. In cases where free fatty acids are present and the acid value has to be determined, the rosin acids will be titrated together with the free fatty acids in the usual determination of the acid value. The total acid value must therefore be corrected for the quantity of rosin acids found, if the true acid value of the fattv acid be required. Microscopical Examination for Beef and Lard Fats. If either of these fats be suspected, but no idea has been obtained as to the amount present in a sample, Key to Plate I. 1 2 Beef fat ,^^^^I^Kt.^ (polarised light). polarised light). X 150. "* ^"■ 3 Beef fat. X 150. 4 5 Lard. Lard. X 250. X 150- PLATE T. Beef akd Lard Crystals. To face page 5G. Microscopical Examination 57 two- separate mixtures of 2 and 5 grm. of the fat respectively with 10 c.c. of methylated ether (sp. gr. 0"720) are made in test-tubes, which are then closed with cotton-wool and allowed to stand at a tempera- ture of about (but not below) 20° C. for twenty- four hours, or less if much crystallisation occurs before the lapse of that time. A piece of glass tubing, closed at one end with the finger, is lowered into the mixture till it touches the crystalline mass, when, on removing the finger, some of the crystals will pass into the tube, and may be removed from the liquid. A very small portion is dropped into the centre of a drop of clove oil previously placed on a microscope slide, and quickly covered with a cover-glass. The crystals are then examined with a ^-in. and finally with a ^-in. objective. Beef fat crystallises in characteristic fan-like .tuffs, the ends of which are more or less pointed, but are never chisel-shaped, as in the case of lard. The examination of the ends should be carried out with the higher power. The examination must be carried out with great care, as it often happens that lard crystallises in bunches of plates, almost identical with the fan-like tufts of beef-fat, but the ends of the crystals will . be found to be distinctly chisel-shaped, while those of beef are pointed or round. If, however, lard crystals are recrystallised several times, the characteristic chisel-end formation may be lost and the crystals become indistinguishable from those of beef. Inferences, therefore, as to the presence of lard should only be drawn from the results of the first crystallisation. The appearance of the crystals is shown in the plate given. The authors find it very convenient to carry out the examination on the crystals deposited from the Halphen tube after boiling in the water bath as described on p. 52. For this purpose the tubes 58 Fatty Foods are stood aside at room temperature until crystals deposit. A rough idea is then obtained of the amount of beef fat present by comparing the volume of the deposit with that from standards made up in a similar way with known amounts of beef fat. The super- natant liquid is poured off from the crystals and these are dissolved by the addition of 10 c.c. of methylated ether, warming till solution takes place, the tubes being then plugged with cotton-wool, and allowed to stand at room temperature to crystallise, and the crystals examined as described above. If no crystals deposit on standing for twelve hours, it should not be inferred that beef fat is absent until the tubes have heen cooled to 0° G. for not less than thirty minutes. Under these circumstances very small quantities of beef fat will deposit, but the presence of beef fat must be established microscopically. (See p. 134.) This method gives very characteristic crystals. It is well to note in this connection that shea nut oil deposits crystals which, though not identical with, are suffi- ciently like those of beef fat to be mistaken for them. Usually if there is sufficient of this oil to give crystals, some indication of its presence will be arrived at from the somewhat high amount of unsaponifiable matter present in the sample. (See shea nut oil.) Bjorklund's Test. (As interpreted by Lewkowitsch for the examination of cocoa-butter). Three grm. of the sample are mixed with 8"5 c.c. of methylated ether at 18^ C. in a test-tube, which is then closed and shaken vigorously. Pure cocoa-butter will generally dissolve without heating, but if not, solution is effected using the least possible amount of heat. Determination of Stearic Acid 59 The tube is then set aside and the manner of crystallisa- tion noted. Pure cocoa-butter will often not separate unless the temperature is reduced to about 10° C, and in any case it separates in small ball-like tufts, first at the bottom and gradually up the side of the tube, leaving a bright solution. Five per cent, of beef fat, or less of beef "stearine," results in the separation of flocculent masses, and the solution remains turbid. In such cases the precipitate is examined microscopically. If the precipitate is due to beef fat, it will not go back into solution by shaking at 18°C., which will be the case with the deposit from genuine cocoa-butter. It should, however, be noted that some of the modern cocoa-butter substitutes known as "green but- ters " and Borneo tallow are not possible of detection in admixture with cocoa-butter by means of this test. Estimation of Stearic Acid by Hehner's Method. The following method becomes useful when a quantitative determination of the actual stearic acid present in a mixture of fatty acids is required. It is particularly of use in ascertaining whether small quan- tities of beef " stearine " are present with vegetable fats, with the exception of those which contain appreci- able quantities of stearic or higher acids in their glycerides (such as shea nut oil, cocoa-butter, etc.). Method of carrying out the test. — The fatty acids from 10 grm. of the sample are prepared as directed under the Titer test (p. 14), and from 0'5 to I'O grm. (if solid) to 5"0 grm. (if liquid) are weighed into a small, light, tared flask and dissolved in about 100 c.c. of alcohol which has been 60 Fatty Foods saturated at 0° C. with pure stearic acid. The flask is placed in an ice-bath over-night, after which the con- tents are shaken and again placed in the bath for a short time. If crystals are formed the liquid is then syphoned out (with the flask still in the ice-water) by means of a tube the end of which is covered with a small mat of fine linen. The flask and crystals are rinsed twice with about 10 c.c. of stearic acid alcohol solution (also at 0° C), the rinsings being syphoned out by the tube. The linen is then rinsed into the flask with a few drops of hot absolute alcohol and the flask and contents dried and weighed. A small correction has to be deducted for the traces of stearic acid left in the flask from the stearic acid alcohol rinsings. For exactitude this should be determined on the flask used, but the correction of 0"005 grm. as given by Hehner is suflScient for ordinary purposes. Hehner found beef " stearine " to contain about 50 per cent, of its fatty acids as stearic acid, arachis oil about 7 per cent ; while other liquid vegetable oils, and coconut oil gave no crystals at all, cotton - seed " stearine " gave about 3 per cent. The stearic acid solution is prepared as follows : About 3 grm. of pure stearic acid are dissolved in 1- litre of strong alcohol in a stoppered bottle which is then stood in an ice-bath overnight, submerged up to the neck. The saturated liquid is separated by means of a syphon while the flask is in the ice-water. A small piece of muslin should be tied over the mouth of the syphon which dips into the bottle in order to prevent any solid particles passing over. As syphonage is often slow, it is advisable to connect the receiver to a suction putnp and create a small reduction in pressure. Glycerol 61 Glycerol. As particular methods for the determination of glycerol have become ofEcial these methods only are here described, though, like all standardised methods, they are necessarily rather over-burdened with detail. Methods of Crude G-LYCEEmE Analysis recommended BY THE International Committees. {Abstract from their report.) Sampling. The most satisfactory method available for sampling crude glycerine liable to contain suspended matter, or which is liable to deposit salt on settling, is to have the glycerine sampled by a mutually approved sampler as soon as possible after it is filled into drums, but in any case before any separation of salts has taken place. In such cases he shall sample with a sectional sampler, then seal the drums, brand them with a number for identification, and keep a record of the brand number. The presence of any visible salt or other suspended matter is to be noted by the sampler, and a report of saine made in his certificate, together with the temperature of the glycerine. Bach drum must- be sampled. Glycerine which has deposited salt or other matters cannot be accurately sampled from the drums, but an approximate sample can be obtained by means of the sectional sampler, which will allow a complete vertical section of the glycerine to be taken, including any deposit. Analysis. .' (1) ^Determination of free caustic alkali. — Weigh 20 grm. of the sample into a 100 c.c. flask, dilute with 62 Fatty Foods approximately 50 c.c. of freshly boiled distilled water, add an excess of neutral barium chloride solution, 1 c.c. of phenolphthalein solution, make up to the mark and mix. Allow the precipitate to settle, draw off 50 c.c. of the clear liquid and titrate with normal acid. Galcu- ate to percentage of NagO existing as caustic alkali. (2) Determination of ash and total alkalinity. — Weigh 2 to 5 grm. of the sample in a platinum dish, burn off the glycerine over a luminous Argand burner or other source of heat giving a low flame temperature, the temperature being kept low to avoid volatilisation and the formation of sulphides. When the mass is charred to the point that water will not become coloured by soluble organic matter, lixiviate with hot distilled water, filter, wash and ignite the residue in the platinum dish. Return the filtrate and washings to the dish, evaporate, and carefully ignite without fusion. Weigh the ash. Dissolve the ash in distilled water and titrate total alkalinity, using as indicator methyl orange cold or litmus boiling. (3) Determination of alkali present as carbonate. — Take 10 grm. of the sample, dilute with 50 c.c. distilled water, add sufficient N/1 acid to neutralise the total alkali found at (2),, boil under a reflux con- denser for fifteen to twenty; minutes, wash down the condenser tube with distilled water, free from carbon dioxide, and titrate back with N/1 NaOH, using phenolphthalein as indicator. Calculate the percentage of NagO. Deduct the Na^O found in (1). The differ- ence is the percentage of Na^O existing as carbonate. (4) Alkali covibined with organic acids. — The sum of the percentages of Na^O found at (1) and (3) deducted from the percentage found at (2) is a measure of the NagO or other alkali combined with organic acids. Glycerol 63 (5) Determination of acidity. — Take 10 grm. of the sample, dilute with 50 c.c. of distilled water free from carbon dioxide, and titrate with N/1 NaOH and phenolphthalein. Express in terms of Na^O required to neutralise 100 grm. (6) Determination of total residue at 160° C. — For this determination the crude glycerine should be slightly alkaline with NagCOg not exceeding the equiva- lent of 0'2 per cent. NagO, in order to prevent loss of organic acids. To avoid formation of polyglycerols this alkalinity must not be exceeded. Preparation of glycerine. — Ten grm. of the sample is weighed into a 100 c.c. flask diluted with water and the calculated quantity of N/1 HOI or Na^COg added to give the required degree of alkalinity. The flask is filled to 100 c.c, the contents mixed, and 10 c.c. measured into a weighed Petrie or similar dish 2'5 in. diameter and 0*5 in. deep, which should have a flat bottom. In the case of crude glycerines abnormally high in organic residue a less quantity is to be evaporated, so that the weight of organic residue does not materially exceed 30 to 40 mgrm. Evaporation of the glycerine.- — -The dish is placed on a water bath (the top of the 160° oven acts equally well) until most of the water has evaporated. Prom this point the evaporation is effected in the oven. Satisfactory results are obtained in an oven measuring 12 in. cube, having an iron plate | in. thick lying on the bottom to distribute the heat. Strips of asbestos millboard are placed on a shelf half-way up the oven. On these strips the dish containing the glycerine is placed. If the temperature of the oven has been adjusted to 160° 0. with the door closed, a temperature of 130° to 140° can be readily maintained with the door partially 64 Fatty Foods open, and the glycerine, or most of it, should be evaporated ofE at this temperature. When only a slight vapour is seen to come off, the dish is removed and allowed to cool. An addition of 0'5 to 1 c.c. of water is made, and. by a rotary motion the residue brought wholly or nearly into solution. The dish is then allowed to remain on a water bath or top of the oven until the excess water has evaporated and the residue is in such a condition that on returning to the oven at 160° C. it will not spit. The time taken up to this point cannot be given definitely, nor is it important. Usually two to three hours is required. From this point, however, the schedule of time must be strictly adhered to. The dish is allowed to remain in the oven, the temperature of which is carefully maintained at 160° C. for one hour, when it is removed, cooled, the residue treated with water, and the water evaporated as before. The residue is then subjected to a second baking of one hour, after which the dish is allowed to cool in a desiccator over sulphuric acid and weighed. The treatment with water, etc., is repeated until a constant loss of 1 to 1"5 mgrm. per hour is obtained. Corrections to be applied to the weight of the total residue. — In the case of acid glycerine a correction must be made for the alkali added. One c.c. N/1 alkali represents an addition of 0'022 gramme. In the case of alkaline crudes a correction should be made for the acid added. Deduct the increase in weight due to the conversion of the NaOH and NagCOg to NaCl. The corrected weight, multiplied by 100, gives the percentage of total residue ai 160 C. Preserve the total residue for the determination of the apparent glycerol content (p. 68). (7) Organic residue. — Subtract the ash from the Glycerol 65 total residue at 160° G. Report as organic residue at 160° C. (Note : It should be noted that alkaline salts of organic acids are converted to carbonates on ignition and that the COg radicle thus derived is not included in the organic residue.) (8) Moisture. — This test is based on the fact that glycerine can be completely freed from water by allowing it to stand in vacuo over sulphuric acid or phosphoric anhydride. Two to 3 grm. of very pure bulky asbestos freed from acid soluble matei-ial, which has been previously dried in a water oven, is placed in a small stoppered weighing bottle of about 15 c.c. capacity. The weighing bottle is kept in a vacuum desiccator furnished with a supply of concentrated sulphuric acid^ under a pressure equivalent to 1 to 2 mm. of mercury^ until constant in weight. From 1 to 1'5 grm. of the sample is then carefully dropped on the asbestos in such a way that it will be all absorbed. The weight is again taken and the bottle replaced in the desiccator under 1 to 2 mm. pressure until constant in weight. At 15° C. the weight is constant in about forty-eight hours. At lower temperatures the test is prolonged. The sulphuric acid in the desiccator must be frequently renewed. Bichromate Process foe Glycerol Determination. Reagents required. (a) Pure potassium bichromate powdered and dried in air free from dust or organic vapours at 110° to 120° 0. This is taken as the standard. (b) Dilute bichromate solution,. — 7'4564 grm. of the above bichromate (a) are dissolved in distilled water and the solution made up to one litre at 15*5°C, 5 66 Fatty Foods (c) Ferrous ammonium sulphate. — Dissolve 3"7282 grm. of potassium bichromate (a) in 50 c.c. of water. Add 50 c.c. of 50 per cent, (by volume) sulphuric acidj and to the cold undiluted solution add from a weighing bottle a moderate excess of the ferrous ammonium sulphate, and titrate back with the dilute bichromate (b). Calculate tlie value of the ferrous salt in terms of bichromate. (d) Silver carbonate. — This is prepared as required for each test frorii 140 c.c. of 0'5 per cent, silver sulphate solution by precipitation with about 4'9 c.c. N/1 sodium carbonate solution (a little less than the calculated quantity of N/1 sodium carbonate should be used ; any excess of alkali carbonate prevents rapid settling). Settle, decant and wash once by decanta- tion. (b) Suhacetate of lead. — Boil a pure 10 per cent, solution of lead acetate with an excess of litharge for one hour, keeping the volume constant, and filter while hot. Disregard any precipitate which subsequently forms. Preserve out of contact with carbon dioxide. (f) Potassium ferr'i cyanide. — A very dilute solution containing about O'l per cent. The Method. Weigh 20 grm. of the glycerine, dilute to 250 c.c. and take 25 c.c. Add the silver carbonate, allow to stand with occasional agitation fol' about ten minutes, and add a slight excess (about 5 c.c. in most cases) of the basic lead acetate (e), allow to stand a few minutes, dilute with distilled water to 100 c.c. and then add 0"15 c.c. to compensate for the volume of the precipi- tate, mix thoroughly, filter through an air dry filter into a, suitable narrow-mouthed vessel, rejecting the first Glycerol 67 10 c.c, and return filtrate if not clear and bi-ight. Test a portion of the filtrate with a little basic lead acetate, which should produce no further precipitate. (In the great majority of cases 5 c.c. is ample.) Occasionally a crude will be found requiring more, and in this case another aliquot of 25 c.c. of the dilute glycerine should be taken and purified with 6 c.c. of the basic acetate. Care must be taken to avoid a marked excess of basic acetate. Measure oS 25 c.c. of the clear filtrate into a glass flask or beaker (previously cleaned with potassium bichromate and sulphuric acid). Add 12 drops of sulphuric acid (1 : 4) to precipitate the small excess of lead as sulphate. Afld 3"7282 grm. of the powdered potassium bichromate (a). Rinse down the bichromate with 25 c.c. of water and stand with occasional shaking until all the bichromate is dissolved (no reduction will take place). Now add 50 c.c. of 50 per cent, sulphuric acid (by volume) iind immerse the vessel in boiling water for two hours and keep protected from dust and organic vapours, such as alcohol, till the titration is completed. Add from a weighing bottle a slight excess of the ferrous ammonium sulphate (c), making spot tests on a porcelain plate with the potassium ferricyanide (r). Titrate back with the dilute bichromate. From the amount of bichromate reduced calculate the percentage of glycerol. 1 grm. glycerol = 7"4564 grm. bichromate. 1 grm. bichromate = 0'13411 grm. glycerol. Notes. (1) It is important that the concentration of acid in the oxidation mixture and the time of oxidation should be strictly adhered to. 68 Fatty Foods (2) Before the bichromate is added to the glycerine solution it is essential that the slight excess of lead be precipitated with sulphuric acid as stipulated in the process. (3) For crudes practically free from chlorides the quantity of silver carbonate may be reduced to one fifth and the basic lead acetate to 0'5 c.c. (4) It is sometimes advisable to add a little potassium sulphate to ensure a clear filtrate. Instructions for calculating Actual Glycerol Content. (1) Determine the apparent percentage of glycerol in the sample by the above process as described. (2) Determine the total residue at 160° C. (3) Determine the bichromate value of the residue at (2) in terras of glycerol. (4) Deduct the result found at (3) from the per- centage obtained at (1) and report this corrected figure as glycerol. Examination of Oils and Fats for Various Impurities. (a) Moisture. In the case of neutral oils or fats it is usually quite sufficient to dry 3 to 5 grm. in a tared dish or large flat weighing-bottle at 105° C. to constant weight, weighing every half-hour until a minimum is reached. In dealing with oils and fats which are specially prone to oxidation, such as the drying oils, it is preferable to place the oil in a long test-tube, which is weighed together with a finely drawn-out tube reaching to the bottom and through which a current of dry CO^ can be Moisture 69 passed. About 10 grm. of the oil are weighed into the tubCj which is then heated in a slanting position in an oil or salt bath at 105° C.^ passing a slow current of COg both during the heating process and until the oil has become cool. The heating is continued until the tube ceases to lose weight. When the oil or fat contains any marked quantity of free fatty acids, which are of a volatile nature, the estimation of moisture is rather difficult. In ' this case a few grammes of the oil may be placed in a dish containing a known weight of sand and the mixture allowed to diy over sulphuric acid in a desiccator, or some of the oil may be filtered through two or three layers of dry thick filter-paper, and equal quantities of the filtered and unfiltered sample dried side by side at 105° C, the difference between the two losses of weight being taken to be that of the moisture present. It is assumed that the two quantities lose an equal weight of volatile acids, an assumption which is, however, not strictly true, the error increasing with the percentage of water present. (b) Non-fatty Hatter. Carelessly treated oils may contain non-fatty con- stituents, which, in animal fats, will be tissue derived from the rendering of the fat, or, in vegetable oils, matter which has passed through the presses or extractors and has not been properly removed after- wards, or traces of soaps arising from refining processes. The quantity present is easily estimated by dissolving the oil in petroleum ether and filtering through a tared filter, after which the filter is placed in an extractor for a short time and then weighed. Usually tlie result of some refining process, by which the free fatty acids have been removed. 70 Fatty Foods (c) Metallic Impurities. Traces of various metals or their compounds, usually as soaps, are also found in oils and fats. These may be : (1) Copper, arising from the action of the fatty acids on copper containei's, or purposely added as a colouring material, as in the case of olive oil. (2) Lead and zinc, usually from containing vessels lined with these metals. (3) Lime, Magnesia, Soda, Potash, (4) Iron as an impurity from tanks or as a natural constituent of some oils and fats. In order to separate these impurities from the oil or fat, two methods may be employed : (a) About 20 grm. of the sample, melted if neces- sary, are shaken in a separator with 50 c.c. of 10 per cent, nitric acid, and allowed to separate. The aqueous layer is drawn off and evaporated to a small bulk and examined by the usual qualitative tests. This method may also be made quantitative if desired. Iron, copper and lead may be estimated colorimetrically. (&) 1'en grm. of the oil are incinerated in a porce- lain dish by a method due to M. Delecoeuillerie. For this purpose the dish is placed on a sand-bath and a cone of ash-free filter-paper placed in the oil, point upwards. As soon as the fat has reached the volatilising point, a light is applied to the point of the filter-paper, when the burning of the fat is accomplished verj' simply, the last traces of carbon being burnt off over a Bunsen flame or in a muffle. (5) Sulphur. — As carbon bisulphide is used as a Hydrocyanic Acid 71 solvent in the extraction of oilsj traces of sulphur may be present from this, as well as from other sources. It may be detected by saponifying 5 to 10 grm. of the sample, and adding to the solution of the soaps a small quantity of an alkaline solution of a lead salt, when, in the presence of sulphur, a brownish colouration will be produced. (6) Arsenic. — A quantity of the oil is shaken out with nitric acid, as described above, the acid liquid being then concentrated to small bulk and Reinsch's test applied. If a positive reaction be not obtained, the examination should be repeated with the Marsh test as laid down by the Joint Committee (Journ. Soc. Ghem. Ind., 1902, vol. xxi, p. 93). (d) Hydrocyanic Acid. A few c.c. of the oil are gently heated in a test-tube, in the mouth of which has been placed a strip of filter- paper, which has been dipped, first in a weak alcoholic solution of guaiacum resin, and secondly, without drying, in a 1 per cent, aqueous solution of copper sulphate. If hydrocyanic acid be evolved from the oil the paper develops a violet stain. The same test may be applied to the detection of the acid in oil cakes, in which case some of the ground cake is moistened with water and allowed to stand for a short tinae before testing, as in these cases the production of the acid is usually due to the presence of a cyanogenetic glucoside. The occur- rence of these glucosides in feeding stuffs has been carefully investigated by Henry and Auld (Journ. Soc. Chem. Ind,, U'OS, vol. xxvii, p. 428). The following method for the determination of hydrocyanic acid as given by them may be used : A convenient quantity of the product is rapidly 72 Patty Foods ground and placed in an extractor and exhausted thoroughly with hot alcohol. The alcohol is then distilled from the extract and the residue mixed with 50 c.c. of water and 10 c.c. of 10 per cent, hydrochloric acid. This mixture is then steam distilled until hydro- cyanic acid ceases to pass over. In the distillate the acid is estimated by adding a slight excess of sodium bicarbonate and then standard iodine solution until the solution is just faintly yellow. Then since HON + 13 = Hr+ CNI, it follows that 1 c.c. of N/10 iodine = 0-00135 grm. of HON. The Kinds of Oils and Fats commonly employed in various Manufactures. It may possibly be of some assistance to indicate briefly some of the oils and fats which are most com- monly employed in the manufacture of certain fatty foods, or of materials in which an oil or fat forms an im- portant constituent. It must not, however, be supposed that in the following category are included all the materials which may be used, these being merely set out as a guide. Chocolate. — Cocoa butter, coconut and palm kernel, " stearines/' Borneo tallow, Japan wax, shea nut " stearine," and, in fact, any hard, solid vegetable fat. The solid animal fats were used in the past, but very little at the present day.^ Biscuits and calces. — Butter fat, coconut oil, palm kernel oil, and mixtures of cotton-seed oil and beef " stearine," lard, premier jus, oleo oil, liquid vegetable oils, and margarines which may contain mixtures of the above. Hard vegetable fats, particularly coconut " stearine," are commonly employed in " puff " biscuits. ' The so-called " cream " of chocolate creams does not contain any oil or fat, but is chiefly a mixture of glucose and starch. Commercial Uses of Oils and Fats 73 Ice, wafers. — The layer of so-called " icing " usually consists of a mixture of coconut oil and icing sugar intimately incorporated while cold. Marzipan. — This substance should consist of ground almonds and sugar only, but it will be necessary to look for other nut and kernel oils, as cheaper nuts and kernels are often substituted for the almond. Caramels. — The fat usually present is butter fat, part of which may arise from the condensed milk used. Sometimes coconut oil and " stearine " and other fats are used, but are less suitable. Toffee. — Chiefly butter fat or coconut oil, but some- times margarine fats. Tinned food packing oils. — Sardines and similarly preserved fish should be packed in olive oil, but cheaper substitutes, especially arachis and cotton-seed oil, are used in less expensive varieties. In examining such oils, after use, it is to be remembered that traces of fish oil will be almost certainly present (see Henseval and Deny, abstract. Analyst, 1904, vol. xxix, P-115). Salad oils. — Any refined liquid vegetable oil, which does not deposit " stearine " on standing. 74 Fatty Foods CHAPTER III. THE ANALYTICAL I'EOCEDUEB AND ITS 3NTEBPEETATI0N. The methods and tests given in the previous chapters will, in general, be found sufficient for the examination of edible oils and fats in ordinary every- day work. Stress is again laid on the necessity for the use of the greatest care in carrying out these methods, and for attention to details, as the methods themselves ai'e often of an empirical' nature, and so dependent on conditions that a deviation from the stated procedure will often lead to error. This is no doubt a common cause of the private condemnation of useful methods. The descriptions given are, in our opinion, sufficiently full to contain all the necessary detail, without being burdened with undue elaboration. In the sphere of analysis of natui-al products, such as oils and fats, it will certainly be the case that no analytical method is infallible in its results, but because a method or test breaks down in certain isolated cases, it is no reason why, in the vast majority, its indication should not be received with due weight. There is, unfortunately, a type of mind which sees the exception and allows it to bulk so large, that the vista of agreement is entirely obliterated. The analytical methods for oils and fats are some- Analytical Procedure J'S what of the nature of detective aids to common-sense, and only a just appreciation of their results will lead to success. It is when they are regarded as fetishes, or interpreted in an inflexible manner, without due regard to the underlying principles, that disaster follows. To give a general scheme for the examination of edible oils and fats would be, in the nature of things, impossible. New oils are always appearing, and well- known one?, on account of better methods of refine- ment, are brought into use for purposes for which, hitherto, they may have been considered impossible. Further, in our present state of knowledge, there are many oils and fats for which we have no distinctive reactions, or which possess no such characteristic features as would provide any sure guide to their presence, so that when a mixture of oils is the subject of examination, it is by no means always the case that its composition can be elucidated. We have, on page 79, taken three general cases which may commonly arise, and have pointed out with the greatest brevity and diffidence that which is, in our opinion, the simplest method of attack. To supple- ment these we have appended a number of examples which may indicate how analytical results are inter- preted. Before passing to these, however, we again set out, even at 'the risk of repetition, the general value and meaning of the methods and tests described in the last chapter. Under each oil and fat in the following chapters there are given, as far as possible, the special tests or characteristics by which any such may be recognised, and when any particular oil or fat is sus- pected, reference should be made to the table of values appended, for further guidance. (1) Saponification value. — This value gives a clear 76 Fatty Foods idea of the mean molecular weight of the glycerides present (p. 21). The greater number of oils and fats have saponifica- tion values between 190 and 200 and chiefly between 192 and 197; values higher than 200 are possessed by the coconut oil group and butter, and some few fats, such as Chinese vegetable tallow, porpoise jaw oil, the chalmugra group, mafura tallow, ochoco fat, Japan wax and dika fat. On the other hand, perilla and castor oils, shea butter and djave butter fall below 190 as a rule, and particularly the rape oil group, which falls below 1 80. (2) Iodine value. — This is a measure of the un- saturated glycerides or acids, and when applied to a known weight of the unsaturated acids themselves, furnishes very clear information as to the actual nature of those present (see p. 24). High values characterise the liquid vegetable fats. Perilla oil is the highest known (206), followed by the drying oils (120—200), and the semi-drying oils such as cotton, soja, etc. (100—140), then the rape oil group (95-120) and non-drying oils (80-110). The fish oils range from 100—195, while the solid vegetable fats lie in the neighbourhood of 40—65 (except the coconut group), followed by butter fat at about 30—35, and the coconut oil group from 5—25. (3) Refractometer value. — This physical test is largely of confirmatory value to the last, as in nearly all cases a high value corresponds to a high iodine value. Its indi- cations, however, do not vary between such wide limits as the iodine value for any particular oil or fat, so that its indicative value is perhaps somewhat greater (p. 17). The i-elation of these three values for a number of different oils and fats is given on the annexed chart. In the case of mixtures, each constituent will affect the total Analytical Procedure 11 value found for each of the above three constants pro- portionately to its amount present in the mixture. (4) Beichert-Meissl-Polenske-Kirschner value. — This is fully discussed under "Butter and Margarine," and, gives very close results for the amounts of coconut, palm kernel oils, and butter fat present in admixture with other fats or oils, with' the exception of cohune oil, which, analytically, behaves like coconut. (5) Acetyl value. — The true acetyl value is the measure of the hydroxylated glycerides present, though, as Lewkowitsch points out, it is not strictly a constant, but a variable. Some oils, such as safflower, rape, etc., and the fish oils, are characterised by high acetyl values, so that in these cases it provides a distinct help in confirming their presence in mixtures (p. 28). (6) Bellier's test, followed by Beiiard's test, gives exact information as to the presence and amount of arachis oil in a mixture of other liquid oils. Bellier's test, however, is useless in the presence of nearly all solid fats, both animal and vegetable, and the quanti- tative determination of arachis oil by Renard's test is, in such cases, practically impossible, and even the qualitative determination of the presence of the oil necessitates very great care (pp. 30 and 34). (7) Acid value.- — This is only a measure of the free fatty acids, and as such gives infornaation as to the condition of an oil. It has no bearing on the compo- sition of a mixture (p. 41). (8) The HexahromideAest. — This test, used in a quali- tative manner, is of value in determining the presence of fish oils and drying oils in admixture with other liquid vegetable oils, and to a certaili extent gives evidence as to the presence of rape oil (p. 42).' (9) The Lead-salt-ether method.' — This process is really a preliminary to further investigation. By its means 78 Fatty Foods the unsaturated fatty acids may be separated, almost quantitatively, from the saturated acids. As a con- sequence, the presence of unsaturated acids other than oleic may be determined by means of the iodine value. The modification of Tortelli and Fortini allows of the detection of lape oil in the presence of practically all other fats (p. 38). (10) Unsapunifiahle matter. — This determination gives evidence, when the quantity is above normal, of the presence of mineral oils, waxes, etc. Apart from these, when high, it points to the presence of shea butter, which may contain 5 to 10 per cent., and in certain cases it may be the only means of detecting this fat (p. 44). If the liquid after titration for the saponification value be diluted with water, the production of turbidity will indicate mineral oils or paraffin wax. Shea butter, on the other hand, usually gives but a slight turbidity, and the precipitate (if any) is sticky and remains at the bottom of the flask. (11) Phytosteryl acetate test. — -This toomuch neglected test is practically infallible in its indications of vege- table oils ar fats in admixture with animal fats, but the reverse process, namely, the detection of small quantities of animal fats in vegetable oils, is not possible except by the use of the dibromide method (p. 44). (12) Optical activity. — The importance of this test lies in its indication of certain poisonous oils, and its determination should never be neglected when examining unknown mixtures intended fc^ edible use (p. 17). (13) Colour reactions : (a) Halphen's reaction. — This indicates cotton-seed oil and kapok oil. In the case of the former it can generally be made fairly quantita- tive. The test seldom fails for cotton-seed oil, except when the oil has been subjected to certain rare treat- ments. When only a doubtful result has been obtained, Typical Cases 79 using the oil itself, more distinct indication will often be obtained on repeating the test with the fatty acids prepared from the oil (p. 52). (b) Baudouin's reaction. — This is practically typical of sesame oil. Certain rancid oils sometimes give a doubtful reaction, and in such cases the test should be applied to the fatty acids (p. 53). (c) Becchi's test. — The modification of this test, pro- posed by Millau, is probably the only means of detect- ing kapok oil in admixture with other oils. It serves also to distinguish this oil from cotton-sepd oil. ' (14) 'l^he microscopual examination of the deposit from an ethereal solution of fats. — By this means the presence of beef and lard fats may be determined in the presence of liquid oils and of some solid vegetable fats. It also serves to determine the presence of beef fat in lard and vice-versa. Such solid vegetable fats as shea butter are not distinguishable by this test from animal fats (p. 56). (16) Stearic acid (Hehner and JVlitchell's method). — By this method the actual separation of stearic acid is accomplished, and if the content in stearic acid of any fat be known, a good quantitative idea of the amount present in admixture with other oils or fats, not con- taining stearic acid, can be obtained (p. 59). Thbke Typical Cases. Case 1. — A sample of oil guaranteed pure. — It is required to say whether the description be true. Procedure. — Determine the saponification value, and on completion of the titration, dilute well with distilled water and carefully note if there be any turbidity which would indicate unsaponifiable matter, such as mineral oils or waxes. Then determine the iodine and refractometer values, and compare these with the values 80 Fatty Foods given for the oil. Tf not normal, take into considera- tion the cheapest possible adulterants of similar nature which could account for the abnormality (see current price lists in trade journals), and apply characteristic tests (if any) for such. If no characteristic tests are available, and the figures are so abnormal as to He out- side the known values of the oil in question, it is pro- bably safest, if no further information be available, to return the oil as abnormal in composition. If the figures be normal the sample may be pure, but in expert adulteration, additions will be chosen which will fall within the usual limits for as many of the values for the real article as possible, and the analyst must therefore consider what adulterants will fulfil these requirements, and will make such tests as will verify or disprove such suppositions. For instance, Bellier's test for jirachis oil in olive oil, etc., the hexabi'omide test for drying oils in non-drying oils and so on. Taste and odour are sometimes of value for suggesting possible adulterants, but present-day oil refining has largely discounted the value of these tests, and in any case they require experience. The odour is best obtained by rubbing the sample in the hands. In the case of oils extracted from manufactured products, or oils which have been used for preserving fish, etc., as in the case of sardines, it must be remembered that traces of some other oil may have been extracted with the chief product, so that, for instance, fish oil may quite possibly be found in olive oil which has been used for sardine packing. See also p. 73. Case 2. — An unhnown oil or fat, but hnown to be not a mixture. — If liquid, it may be presumed to be a vegetable oil (exceptions — lard oil and fish oils). Determine the iodine value, for as vegetable oils range from 79 to 206, this will best help to fix the probable Typical Cases 81 nature of the oil. Such oils as are thus indicated must then be followed up by all possible characterising tests. For fish oils see p. 273 ; limpidity and optical activity may also help. If soiid, the fat may be vegetable or animal, in which case determine the saponification value, as this will distinguish coconut and palm kernel oils, butter and allied fats. Then determine the iodine value, and, if necessary, apply the niodified Bjorklund's test, as this will distinguish beef and lard fats and shea butter (exa- mining deposit, if any, microscopically) from cocoa butter, certain illipe fats, etc. Shea butter can only be dis- tinguished, in the last issue, from lard and beef fats, by the percentage of unsaponifiable matter and by the phytosteryl acetate test. As regards the large number of vegetable tallows and solid fats, it is only possible to rely on the appearance of the sample and its analytical figures to arrive at a possible solution. Case 3. — A mixture of edible fats.- — Some of the principal cases of mixed edible fats will be dealt with under "Margarine" (p. 129). In attempting to give the composition of a mixture, the greatest weight should be given to those results which can be determined with the greatest accuracy. For instance, the saponification value is a most exact figure, and its indications should not be allowed to be set aside for any other considera- tion, i. e. a value appreciably over 200 indicates the presence of glycerides of the lower fatty acids, the source of which must be accounted for in spite of the fact that the interpretation tentatively derived from all other analytical figures is inconsistent with it. Again, the Reichert-Meissl-Polenske-Kirschner values are of great accuracy in regard to the quantities of coconut, palm kernel and cohune oils, butter and rarer fats which 6 SS Fatty Foods may be present in a mixture, and they must be given- due weight. The various colour tests and the hexa- bromide test are, in themselves, only indications of the presence of certain oils and fats, and no deduction as. to quantity (with the possible exception in most cases of cotton-seed oil) should be made, but when a certain oil or fat is thus known to be present, its average figures can be employed to determine how much can. possibly be present, in order to fit in with the general results obtained. The deposit (if any) from ether solution should be examined microscopically, as small quantities of beef " stearine " are used to stiffen softer vegetable fats. If small quantities of vegetable fats or oils are suspected in animal fats, and the figures obtained are not sufficient in themselves to certify to such addition, or if no really characteristic test for vegetable oils has been given (even cotton-seed oil may be so treated as to give such a slight reaction that small quantities might be passed over), the phytosteryl acetate test must be applied, always with care to ascertain the presence of paraffin wax added for the purpose of nullifying the test. For the purpose of illustrating the foregoing principles, the following examples of some typical cases which may arise are given. They consist partly of actual problems which have arisen in the course of wori, and partly of theoretical cases, so arranged as to illustrate the method of procedure and the deductions to be. made from the results obtained. In giving such examples, we are perfectly aware that the results which we arrive at are not always the only possible ones ; they are based on the methods which we have given onlj-, and it is quite probable that, if a much greater number of determinations were made, somewhat more exact ideas might be obtained. The deductions which we have Typical Cases 83 made are, we believe^ fully justified, from the nature and history of the samples examined and by the general tendency and bearing of industrial processes. By taking as a basis of calculation the extreme analytical limits, and by working out all the permutations and combinations of possible constituents which may arise therefrom, it would probably be quite possible to find half a dozen different solutions in some cases to the following problems. The ordinary investigator will, however, be quite safe, in the great majority of cases, in working out his problems using the typical figures for those oils and fats which he has found to be present, and it is for his guidance that these examples have been worked out. With the. ultra-critical side of the question we have no concern, as they are intended rather to exemplify the methods of interpretation than to provide a field for criticism. Case 1. — Olive Oil, guaranteed pure. Iodine value ..... 8i"0 Eefractometer number . . . 56"0 On cooling the oil became turbid at 2° C. Colour and appearance normal. Of the above figures, the iodine value is not abnormalj the refractometer number and turbidity temperature are quite normal, and relying on these figures alone, the oil would have been passed as pure^ but as no olive oil should be passed without applying Bellier's test, it was therefore tried, and a heavy per- manent crystalline precipitate, reappearing after heating and cooling the solution, was obtained, and by carrying through Renard's test about 50 per cent, of arachis oil was found to be present. 84 Fatty Foods Case 2. — Cocna-butter, guaranteed pure. Saponification value . . . 204 8 Iodine value .... 27'3 Refracfcometer nuQiber . . . 44 2 Bjoi-klunds test gave rapid crystallisation from the ether, and the precipitate was not soluble at 20 C. Microscopically the insoluble crystals has the appear- ance of beef fat. The sample was hard and brittle, and smelt strongly of cocoa-butter. ^^Inferences. — The high saponification value indicated the presence of coconut or palm kernel products. If, however, the oils had been present, the melting-point would have been too low and the fat soft in character. The presence of the " stearines " was therefore inferred. Now the saponification value of either coconut or palm kernel " stearine " is 250, and from this figure and the normal value for cocoa-butter (194) 20 per cent, of one of these " stearines " is calculated. As it is not possible to arrive, in any simple manner, at any certain conclusion as to which of the two " stearines " might be present, it was taken to be coconut " stea- rine." This has an iodine value of 5,- and taking 20 per cent, to be present in the mixture, it follows that the remaining 80 per cent, has an iodine value of 33, and, by similar calculation, a refractometer number of 46'o. These figures would pass for cocoa-butter, but as beef fat has been demonstrated, a small percentage is evidently present, but the actual amount could only be arrived at by making up standard mixtures and com- paring the volume of the precipitates deposited from ether solution or by the method of Hehner and Mitchell (p. 59). i Typical Cases 85 The compo'sition of the mixture was given as : Coconut '7 stearins " . .20 per cent. Beef ^ » ■ . 5 „ Cocoa-huttier . . .75 „ \ \ Case 3. — A Lard Substitute. Saponification value . . . 194"0 Iodine value\ .... 9o'5 Refractometei' number . . . 56'0 The sample was a soft plastic fat. Inferences. — Th0 saponification value showed that no glycerides of low molecular weight were present. The iodine value shows the presence of a liquid vegetable oil, conftrmed by the refractometer number. Halphen's test was strongly positive, and on diluting down with an indifferent oil (until the reaction was below that given by the 20 per cent, standard) it was evident that about 80 per cent, of cotton-seed oil was present in the original sample. This high percentage of liquid oil nei?essitated the presence of a stiffening fat. A solution of 2 grm. in 10 c.c. of ether gave a copious precipitate, which on microscopical examination was evidently from beef fat, and, as only a comparatively small quantity could possibly be present, it was necessarily beef " stearine." It was assumed therefore that only cotton-seed oil and beef " stearine " were present, and, calculating from the iodine values, the composition was returned as : Cotton-seed oil . . . 85 per cent. Beef stearine . . . 15 ,, 86 Fatty Foods Case 4. — A Sample of Coconut CDH. Saponification value . . 2Jo7'0 Iodine value Refractometer number Reicliert-Meissl value Polenske value Setting point Free fatty acids 25-2 59-5 '5-7 ( 12-6 ^ 76-0° F. I 3*5 per cent. No precipitate was obtained fromi 2 grm. of the fat in 10 c.c. of ether when kept in ice /for thirty minutes. Inferences. — The case presented! some difficulty in interpretation. \ The saponification value suggestem the presence of another fat^ but sufficiently cheap av/ailable solid fats were negatived by the ether test. / Liquid vegetable fats were negatiyed by the setting- point, which is two to three degrees *.above the normal for coconut oil. Coconut or palm kWrnel "stearines " might be present, but were negatived\ by the refracto- meter and iodine values. All the fighires are normal for coconut shavings oil, and the sample was tlierefore reported as such (see p. 151). i It may be useful to point out that licLiuid vegetable oils are excluded by the values obtain/id, though, at first sight, their presence might be i^iferred. To account for the saponification value foufid, a mixture of 70 per cent. , coconut oil (iodine valr^e 8"5) and 30 per cent, of liquid vegetable oil (iodine value 100) would be required, or else a mixture of ©0 per cent, palm kernel oil (iodine value 14) and 20 per cent, of liquid vegetable oil. The iodine value u^ these two cases would be 36 and 31 respectively and are therefore much too high. Simi- larly the refractometer numbers would be 42 and 41 and are also tpo high, and if either of these were Typical Cases 87 adjusted to fit the found yalue, the saponification value, which is the most reliable figure, would show consider- able discrepancy. Case 5. — Oil extracted from Biscuits. Saponification value . . . 229'0 Iodine value ..... 34'2 Befractometer number . . . 42'5 A solution of 2 grm. in 10 c.c. of ether gave a copious precipitate on cooling in ice, and the crystals, micro- scopically, were characteristically lard. As the high saponification value indicated coconut oil or butter fat, the following determinations were made : Reichert-Meissl value . . . 6"7 Polenske value ..... 7'0 Kirschner value . . . . \'Q These correspond to a mixture containing 50 per cent, coconut oil and about 2 per cent, of butter fat. The mean values for the saponification, iodine and refracto- meter numbers were taken, and from these it was found that a mixture of — Coconut oil . . . .50 per cent. Lard . ■ . . . . 48 „ Butter fat . . . . 2 „ fits in well with the values found for the original sample. The above mixture is therefore very probably correct. The values for coconut • or palm kernel " stearine " do not fit in at all well, and their presence may be therefore taken as negatived. Case 6. — Oil from Oil Cake. Saponification value . . • 171'5 Iodine value ..... 98"8 Befractometer number . . . 58'6 88 Fatty Foods The very low saponification value pointed to rape oil, as no unsaponifiable matter was noted on diluting the solution. The only possibility which has to be accounted for is that of a mixed cake, and this possibility seems very doubtful unless the admixture was very small. Halphen's test was negative. The hexabromide test was quite typical of rape, and the solution cleared on the addition of ether. The oil was therefore reported as being from rape cake. Case 7. — A Sample of UI.eo Oil. Saponification value . 198-0 Iodine value . 50-8 Eefractometer number . . 'J8"l The presence of lard in the sample was practically excluded by a microscopical examination of the crystals obtained from a solution of 5 grm. in ether, and as the amount of the precipitate was very small, the presence of beef " stearine " was practically negatived. Halphen's test was negative. Baudouin's reaction was negative. These results, taken together with the quite normal figures, practically establish the purity of the sample. In order to definitely exclude the presence of any vegetable oil, it would have been necessary to use the phytosteryl acetate test. Case 8. — A Margarine Fat. Saponification value .... 196 Iodine value ..... 76'0 Eefi'actometer number . . . 52'5 Inferences. — The saponification value excludes coconut Typical Cases 89 oil products and butter fat, except in traces, and the iodine value suggests liquid vegetable oils. The Halphen reaction was strongly positive, and the quantity of cotton-seed oil present was roughly indicated as between 30 and 60 per cent. As a solid fat must necessarily be present, 5 grm. were crystallised from ether and the precipitate on micro- scopical examination indicated the presence of lard. As no beef crystals were found premier jus and beef " stearine " were practically excluded, but not oleo oil. As the unsaponifiable matter was only 0"3 per cent, the presence of shea butter as against lard was nega- tived. As this examination was made at a time when lard was higher in price than beef fat, it seemed exceedingly probable that oleo oil might be present, but, on account of the wide limits which it was necessary to assign to the amount of cotton-seed oil present, the actual incor- poration of oleo oil can only be deduced inferentially. For instance, the composition might be worked out as follows : supposing 30 per cent, of cotton-seed oil to be present in the mixture, the iodine value of the other 70 per cent, would be 61, which value would be quite correct if it were all lard, but the quantity of crystals obtained from the ether was too small for so large a percentage. If 60 per cent, of cotton-seed oil be taken to be present, a residual iodine value of 25 is necessary, which is clearly impossible. If a mean value of 45 per cent, cotton-seed oil be taken, a residual iodine value of 48 is obtained, so that, by taking all these into considera- tion, it was quite justifiable to assame that both oleo oil and lard were present, and that the oleo was in excess of the lard. The composition was therefore returned as probably : 90 Fatty Foods Lard . . . . .20 per cent. Oleo oil . . . .40 1! Cotton-seed oil . . . 40 3J Case 9. — An Adulterated Ghee. Saponification value . 219 Iodine value . . 34 Eeichert-Meissl value 24-8 Polenske value 2-4 Kirschner value 17-0 Difference figure (Ave-Lallement) + 12 Inferences. — The figures point to an addition of some solid fat (vegetable or animal), as the iodine value was too low for a liquid vegetable oil. The Polenske value excluded coconut oil products, and the relation of the Kirschner value to the Polenske value confirmed this. The addition, therefore, was either lard, beef, or a solid vegetable fat. The microscopical examination of the deposit from ether breaks down completely in the presence of butter fat, and could not therefore be applied. , The presence of shea butter was excluded by a normal amount of unsaponifiable matter. The only solution of the problem lay in the use of the phytosteryl acetate test. The acetates at the fifth crystallisation melted at between 116° and 118° C, from which it was concluded that a solid vegetable fat, not improbably, from the nature of the product, one of the Bassia fats, was present. The percentage cannot be found except in a very approximate manner, but from 15 to 25 per cent, was inferred. Resolution of Fatty Acids 91 Examination of the Fatty Acids by Fractional Precipitation. In certain cases it may be necessary to examine more closely the fatty acids which can be obtained from the sample under examination. A full scheme cannot be given here^ as such an examination does not form an ordinary analytical opera- tion, and for such, reference must be made to larger text-books. The following brief outline will give an idea of some methods of attacking the problem. The fatty acids are prepared from a large quantity of the fat by saponification with alcoholic potash or soda (about 15 grm. of caustic alkali and 250 c.c. of 90 per cent, alcohol to 50 grm. of fat). The alcohol is partially distilled off and the solution of the soaps poured into a. large excess of water containing sufiicient hydrochloric acid to decompose them, and the fatty acids allowed to rise and solidify. They are then skimmed off and placed in a large separating funnel and dissolved in ether, and the solution washed till free from hydro- chloric acid. If any noticeable quantity of the acids refuse to dissolve in the ether, it is generally due to the presence of esters formed by incomplete saponifica- tion. The water is drained as much as possible, and the ethereal solution shaken with a little calcium chloride for a short time and filtered through a triple layer of thick filter-paper into a dry flask and the ether distilled off, the last traces being removed by blowing air into the flask, or, when unsaturated acids are present, by driving a current of hydrogen through the liquid. It is sometimes advisable to re-melt the fatty acids at as low a temperature as possible, and filter through 92 Fatty Foods dry filter-paper^ taking care that no partial solidification takes place during the process. From the fatty acids so obtained, the unsaturated acids (oleic, linolic, etc.) are separated by the lead-salt-ether method, the solid acids being recovered from their lead salts by decom- posing them in the usual way. The composition of the unsaturated acids is ascertained by determining the iodine value very carefully. Oleic acid absorbs two atoms of iodine, linolic four atoms, etc., so that from the weight of acids taken, and their iodine absorption, the presence of one or more unsaturated acids can be determined, and their proportion with fair accuracy. For the resolution of the fatty acids, undoubtedly the simplest method is fractional precipitation by magnesium acetate as recommended by Heintz (Journ. prak. Cheiii., 1855, Ixvi, p. 1.) For this purpose, 40 to 50 grm. of the fatty apids are dissolved in sufficient alcohol to keep them in solution when quite cold. To this solution, while hot, is added a boiling alcoholic solution of magnesium acetate (about 1—2 grm. at a time) and the solution allowed to cool and the precipitate filtered oif. 'J'o the filtrate a further addition of magnesium acetate is made and the precipitate filtered off as before. When no more precipitation takes place, the solution is made alkaline with ammonia and more acetate added, leaving the whole to stand twenty-four hours before filtration. Any acids not then precipitated may be thrown out by lead acetate, after rendering the solution acid with acetic acid. The various fractions are washed with alcohol, and decomposed under ether with hydrochloric acid and the fatty acids recovered. The mean molecular weight of the fractions is then determined by saponifying with N/2 alcoholic soda or potash (mere titration is not suffi- cient on account of the tendency to form anhydrides), Resolution of Fatty Acids 93 and by this means a very good idea of the various acids present may be obtained. When lauric acid and acids of lower molecular weight are present, they may be separated from higher acids by decomposing some of the original saponification mass (after distilling off the alcohol) with dilute sul- phuric acid, and steam distilling till 1000—2000 c.c. have passed over, the acids being recovered from the distillate by extraction with ether, or by neutralising it and concentrating previous to acidification and extract- tion. By converting these acids into barium salts, a considerable separation may be effected by boiling out with water, as saturated fatty acids up to capric acid are soluble. The fatty acids can then be liberated again and their mean molecular weights determined. Separations carried out intelligently on the above lines will in most cases enable the investigator to arrive at the approximate composition of a fat or to determine the predominant fatty acids present. 94 Fatty Foods CHAPTER IV. ANIMAL FATS. For practical purposes the analyst has to consider only two animal fats and their products, viz. beef fat and lard. Beef Fat. This is the product obtained from bovine animals. A very similar fat is obtained from the sheep, but as there is little difference analytically between this fat and true beef fat they will be treated as similar, as even to the expert their differentiation in presence of one another is practically a matter of guesswork.- Beef fat is more valued on account of its superior taste and smell. Immediately the animal has been slaughtered the fatty portions are usually removed, and treated as follows : For products intended for the manufacture of oleo-margarine, the fat of the heart, caul and kidneys is chilled quickly to remove animal heat and rendered at as low a temperature as possible (100° to 120° F.). The clear fat which runs out is termed " Premier jits." This, with or without various clarifying treatments, such as washing with a solution of common salt, is allowed to " grain," and is then "bagged" up (2 to 3 lb. per bag) and submitted to hydraulic pressure. The fat which is so expressed sets to a soft buttery consistency and is known as "oho" or " oleo oil" and is the product employed in the manufacture of margarine. Beef Fat 95 The residue in the bags is in the form of hard cakes known as " oleo [or hepf) steariyie." In cold weather^ when a softer oleo oil is required, the pressing is not carried out so completely, the result being that^both products are softer in character with a larger yield of " stearine." The process is reversed in warm weather. The various ways of, and reasons for, compounding these products will be found under the head of "Margarine." All the other fatty portions of the carcase not sold as beef are rendered in such a manner as to produce the maximum yield of fat. These form the various grades of tallow, the bulk of which go to the soap maker, but some of the finer grades are used for edible purposes ; the residual tissue, after the render- ing of the premier jus, is worked up in a similar manner. In the following paragraphs only the necessary practical tests for the examination and valuation of beef products are given in connection with their use for edible purposes. Such tests as would be necessary "for soapmakers' work are not dealt with here. The examination of oleo products for adulteration with other fats is not of great importance, as in our experience edible beef products are not as a rule sophisticated. If such examination be desired the methods given under lard may be employed. The analytical examination is rather directed to their valua- tion for, definite technical processes. The point on which all such technical value turns may be summed up almost entirely under the heading of "consistency," which cannot always (paradoxical as it may sound) be adjudicated upon by the melting-point alone as usually determined. The analytical determinations v^hich must be made are : The iodine value, substantiated by the refractive index which give indications in the same direction, and % Fatty Foods perhaps the saponification value, though the iodine value is practically all that is necessary as it furnishes a direct measure of the olein present, the content of which (as a result of the methods of preparation) is the distinguishing mark of these products. The free fatty acids must be determined in connection with observations on the taste and smell, as the presence of more than 0"5 per cent, (calculated as oleic acid) will produce a distinct burning sensation in the throat, and render them unpalatable. Consistency. — Attempts have often been made to measure this most important property by the determina- tion of an arbitrary figure obtained by some physical means, but this has not proved at all satisfactory in practice. It is much better to determine the points of incipient and complete fusion and to draw deductions from their relation. For instance, if they are widely separated and the fat divides during fusion into a distinct liquid and solid portion, it is probable that the product is a mixture which contains a proportion of high melting- point " stearine," and the fat, therefore, will have a structure consisting of hard granules embedded in a semi-liquid matrix. ■ On the other hand, a product which melts within close limits and fairly uniformly will have a more lard-like, plastic, and less granular structure. The above are, of course, extreme cases, but are intended to show how such observations may be employed to assist the manufacturer to match two different consign- ments when he is not able to settle the point by simple observation. It is, of course, to be clearly understood tha;t in the case of sharp melting-points, the actual temperature of fusion is to be taken into account, as both beef "oleine" and beef "stearine" have sharp melting-points, but will differ widely in the actual Beef Fat 97 tfemperatures of fusion. The determination of tlie solidifying point is of little value, as no very definite figure can be obtained, but the solidifying point of the fatty acids is fairly ■ definite and is determined as described under Titer test and then gives similar information to the results of the melting-point detei-- minations. Such results taken in conjunction with the iodine value furnish all possible information obtainable for comparative purposes. Oleo oil and premier jus are coarsely granular fata of attractive appearance. Beef " stearine " is usually in the form of broken press-cakes showing the marking of the bags, but if re-melted and solidified in moulds it has the appearance of stearine candles. The colour may vary from pale creamy white to yellow, bleaching with age. Oleo oil is usually slightly deeper in colour than beef " stearine." The following figures will be sufficient guide to the examination of these products : Meltiug-point ° C. . Beef tallow. Premier jus. Beef stea- rine (oleo- stearine). Oleo oil. 4.7-49-5 47-49 50-54 Variable Solidifying point ° C. Saponification value . 32-37 (In- definite) 193-199 32-35 (In- definite) 195-200 47-50 192-197 Indefinite 198-202 Eet'ractive index 47-49 47-48 46-5-47-5 47-5-48-7 (Zeiss butyro - re- fraotometer at 40° C.) Iodine value 38-44 38-45 18-25 40-50 Titer test °C. . . 43-45 42-5-44 48-51 — Baryta values* — (a) Total 272-1 270-0 274-5 , lb) Insoluble . — 263-8 269-4 261-2 (e) Soluble — 8.3 0-6 13-3 b- (200 + c) — + 55-5 + 68-8 + 47-9 * Ave Lallemeni. 98 Fatty Foods The free fatty acids vary, but when the fats are intended for edible use the amount should, not exceed 1 per cent., and is usually not more than from 0'2 to 0'4 per cent. Lard. At the present day, lard must be taken as describing the whole of the fat derived from pigs. The fat is i-endered in various ways, by which a variety of pro- ducts are obtained which are classified under various grades in America, as follows : (a) Neutral lard. — No. 1. This product is obtained from the " leaf " by rendering the chopped mass, after thorough cooling, at a temperature of 40° to 50 C, and is the variety almost entirely employed for the manu- facture of margarine. No. 2 is obtained in a similar manner from fat derived from the back of the animal, and is used more particularly in biscuit manufacture, and probably to a certain extent in margarine. (b) Leaf -lard. — This is obtained by autoclaving the residue left after rendering the " leaf or neutral lard," at steam heat. (c) Choice kettle-rendered lard. — This is obtained by heating the residues left from the preparation of a " neutral " lard in open steam-jacketed vessels. (d) Prime steam lard. — This term is applied to the renderings of almost all the other parts of the animal from which lard may be obtained — direct steam heat being used for the purpose. Any fatty residues which are left over are worked up under pressure, in order to produce low quality fats, usually termed " greases," which are not used for edible purposes. For a full description of these varieties, and of their composition and characteristics, see Lewkowitsch, ' Oils, Fats and! Waxes,' vol. li, 19U9. \ Lard 99 It may be necessary in practice either to examine lard for its purity or to detect it as a constituent of mixtures. Examination of Lakd foe ADDiiTBEATiON. — Unfortu- nately pure lard itself varies rather considerably in its compositiouj according to the part of the animal from which it is obtained and also with the manner in which the animal has been fed. It is therefore necessary to obtain actual evidence of adulteration and not to rely on sim.ply abnormal figures. Water. — This may be determined by drying a few grammes of the sample in a thin layer in the water oven to constant weight. The amount may also be determined fairly closely by observing the temperature at which the melted fat becomes turbid, as described by Polenske (Arb. a. d. Kaiserl. Gesundheitsamte, 1907, XXV, p. 505) . The following figures have been obtained by this method by Fischer and Schellens : Water per cent. Turbidity, °C. 0-15 ... . 41-2 0-20 1!i i; 53-2 0-25 . 64-6 0-30 . 75-8 0-35 . 85-0 0-40 . 90-8 0-45 . 95-2 Iodine value. — This test is probably of some value as giving an indication of the purity of the sample, as a high iodine value will suggest the presence of some vegetable oil, providing that such admixture has not been, compensated for by the addition of beef "stearine," but the iodine value of genuine lard seldom falls out- side the limits of 48 to 66 per cent. Lards of American origin, according to Voigtlander, fall nearly always 100 Fatty Foods between 61 and 66 per cent.,, tending usually to the higher value. Refractive index. — This figure is of some little use, but only in conjunction with other tests, on account of its close proximity to that of beef fat. Unsaponifiahle matter. — As lard is often adulterated with a judicious mixture of beef fat and vegetable oils, which give no definite reactions, the examination of the unsaponifiable matter for the presence of phytos- terol may be the only way in which the adulteration may be detected. Particular care must be taken that this test is not defeated by the addition of traces of paraffin wax, and precautions against this must con- sequently be taken (see p. 48). Particular adulterants : Beef fat. — It is ■ perhaps best to say at once that 5 to 10 per cent, of beef fat in lard cannot be detected with certainty, if at all. The presence of higher percentages may perhaps be arrived at by the microscopical examination of the crystals obtained from an ethereal solution of the fat (see p. 56) and by the Titer test. G-oske gives the solidifying point of a number of lards as lying between 24° and 29° C, with one or two excep- tions, while in the case of lards adulterated with beef fat, solidifying points from 29'5° C. up to nearly 37° C. , were obtained. Undoubtedly the method of Hehner and Mitchell, in which the iodine value and stearic acid content of the mixed fatty acids are determined, may lead to useful indications, as they point out that a high iodine value and a high stearic acid content would suggest the presence of beef "stearine," as in genuine samples of lard the iodine value rises as the percentage of stearic acid falls. The following figures are given by Mitchell for pure lards (Allen's Organic Analysis, fourth edition, p. 318). Lard 101 Iodine values. Stearic acid, per cent. 57-5 16 61-2 13 61-2 6 to 7 63-6 7-4 65-6 9-9 to 10-6 Cottonseed oil. — This may be detected by Halphen's test, the colour given being interpreted with care, as there is not the least doubt but that lard from pigs fed on cotton-seed meals sometimes gives a slight positive re- action. Positive reactions must therefore be considered carefully in connection with other figures. As henfed cotton-seed oils give no colour with Halphen's test it is advisable, when a negative reaction has been obtained, to apply the nitric, acid test, as recommended by Lewkowitsch : For this purpose about 5 c.c. of the sample are mixed with 5 c.c. of nitric acid (sp. gr. 1'375) in a stoppered tube and violently shaken. Under these circumstances cotton-seed oil produces a coffee-brown colour, the production of which is not affected by previous heating of the oil. The test, however, does not show less than 10 per cent. Unfortunately, rape oil gives an entirely similar colour, but may be differenti- ated by the method given on p. 38 (Tortelli and Fortini). Sesame oil. — ^This may be detected by the Baudouin reaction (p. 53). Arachis oil. — This may sometimes be detected by examining for arachidic acid, by Renard's test, after repeated recrystallisation, but if much " stearine " be present the results are doubtful. Maize oil. — This oil may possibly be present as an adulterant when the presence of vegetable oils not giving specific reactions is suspected. As it gives a phytosteryl acetate of unusually high melting-point, this 102 Fatty Foods test may be of distinct value in detecting the presence of this oil. It may also be useful to obtain the iodine value of the liquid fatty acids, as they are particularly high (136 to 144) in the case of maize oil, while lard seldom gives a figure above 105. Shea nut oil and its products (such as the " oleine") are now on the market to some extent as lard substitutes, and may therefore be expected to find their way into lard. The presence of these products, except in small quanti- ties, may be fairly easily detected by the high unsaponifi- able matter, a slightly lowered saponification value and iodine value, together with an increased refractive index. Coconut oil offers but little difficulty in detection. Any appreciable quantities will greatly increase the saponification value and lower the refractive index and iodine value. Small quantities may be detected by a rise in the saponification value of an alcoholic extract as compared with that obtainable from pure lard under similar conditions. This is due to the fact that coconut oil is very much more soluble in alcohol than lard. A determination of the Reichert-Meissl-Polenske value will confirm these results. The greater majority of samples of lard will be found to give analytical figures within the limits of the following table : Melting-point, °C. . Solidifying point, °C. Saponification value Refractive index (Zeiss butyro- refractometer at 40° C.) Iodine value . Specific gravity f| °C. Free fatty acids (as oleic) Unsaponifiable matter 35 to 46 25 JJ 30 193 ii 199 49 >) 52 57 )) 66 0-860 3> 0-862 0-1 % >7 0-8 % 0-2 % ij 0-4 % Lard 103 Genera I G(>nsiderati(iii». In. detecting adulteration in lard only small quantities of other oils are likely to give great diiSculty. It must be understood that the addition of liquid vegetable oils will in any skilful form of adulteration be masked by the addition of beef " stearine " in order to produce the necessary consistency. If, therefore, vegetable oils are suspected, the likelihood of this concurrence must be taken into account when considering the figures obtained. As in margarine, so in lard, almost any vegetable oil may be sought for at the present time. Attention is again drawn to the great value of the determination of the unsaponifi.able matter (shea nut products) followed by the phytosteryl acetate test, as indicative of the vege- table oils in general, and of maize oil in particular, and also to the methods which are 'employed to defeat this latter test (traces of paraffin wax) . Lard Suhstitvtes. A variety of mixtures of fats, so blended as to resemble lard both in consistency and other properties, may be found on the market. The most common of these is a mixture of about four parts of some liquid oil and one part of beef " stearine." Some of these products contain a small quantity of starch, added in order to mask the tendency of the more liquid portion to separate. This type of mixture probably constitutes quite 90 per cent, of those offered for sale at the moment, but tbere is no doubt that shea nut products are coming on the market in large quantities and will begin to displace such mixtures. "Compound" lards should be understood to be mix- tures of true lard with other fats, but we have found products described as "lard compounds" in which no lard was present at all. 104 Fatty Foods We feel that tlie analyst should be cautious in con- demning such mixtures as fraudulent, as in most cases no fraud is intended, nor would it be to the prejudice of the purchaser. Refined coconut oil is commonly sold as " pure vegetable lard/' and we see no reason to quarrel with such nomenclature. Butter and Margarine. It is scarcely too much to say that the examination of these two substances presents the most difficult problem which the oil analyst has to face. This is due to the extreme variation and complexity which may arise in the production of the latter, and to the wide natural variations of the former, coupled with the fact that adulteration usually consists in the admixture with it of certain quantities of the latter. While a knowledge of the products on the market, ruling prices, and indirect information often enable the investigator to solve the problem of margarine mixtures with reason- able certainty, there is, in a large number of cases, absolutely no method of arriving at proof that the constitution of any sample of butter-fat is due to sophistication and not to natural causes. This natural variation has brought the analyst into the uncomfort- able and unscientific position of having to fix limits of variation in dealing with butter-fat, and the result, as always in such cases, is most unsatisfactory; but until some talisman be found which will distinguish the natural from the artificial, this state of affairs will probably continue. The authors, however, desire to express their decided opinion that little is gained bj- condemning the natural as artificial when reasonable evidence is available that the composition, on which the decision is based, is the result of natural causes Butter and Margarine 105 bej'ond the control of the human agent, and not simply due to inefficient or improper feeding, or unneces- sary exposure of the cattle to adverse climabic conditions. There is no evidence whatever that an unusually small content of the glyoerides of volatile and water-soluble acids due to natural causes renders butter-fat any the less digestive or nutritive, nor is the purchaser prejudiced thereby. In the present state of our knowledge it is better for the analyst simply to fall back on the recognised standard when there is no evidence to show that the composition is natural, and not to attempt to give any weight to his adverse decision by adding remarks as to nutritive value, etc., which are of doubtful utility. As the examination of butter and margarine run to a certain extent on parallel lines, they will be considered together. Sampung. By far the gr.eater number of discrepancies found in the analyses of different observers with regard to water content are caused by inefficient sampling and after- care of the sample. In sampling large bulks of butter or margarine, a piece about 2 in. long by 1 in. square is cut by means of a thin wire from the centre of the mass, and the piece, so cut, lightly pushed with the finger into a dry 6-oz. wide-mouthed stoppered bottle, in such a manner that a trace of the sample is left on the ground inner surface of the neck^ A well-fitting stopper is inserted and carefully twisted round until it is well greased. This procedure is essential in order that none of the water from the sample when melted should pass between the stopper and the neck by capillary attraction. The bottles containing the samples are placed in water (or preferably in an incubator) at .38°to 40° C. until melted. 106 Fatty Foods Watee. Routine Method for a Number of Samples. The estimation is carried out in large weighing bottles 2 in. deep and 1^ in. wide, having no shoulders. The requisite number are dried in the water oven, together with a tare, and allowed to cool in the air or under a cover. They are then weighed in succession, and about 3 to 5 grm. of each sample rapidly poured in from the sample bottles, after the contents have been shaken vigorously to the consistency of cream. The series of bottles, having been re-weighed, are then placed in the water oven, the contents being thoroughly shaken round at intervals until visible water has dis- appeared. The bottles may then be tilted on one side, and turned round from time to time until the weight is constant. If the shaking be skilfully done the whole operation will only take from two to three hours. Patrick's Method for Single Samples. About 5 grm. of the sample are poured out into a platinum basin and gently heated, preferably about 6 in. above the flame of an argand or small bunsen burner. The contents of the dish are gently boiled, while shaking, until " hissing " ceases, when all the water will have been driven off. If properly carried out there should be no discoloration of the curd, and the results are quite as accurate as in the case of the foregoing method. • Fat. The residue left after driving off the water in either of the above methods is exhausted five times with 20 c.c. of ether, allowing the residue to stand under the ether for at least a quarter of an hour at each extraction. The loss in weight after this ti-eatment gives the fat present. Butter and Margarine 107 OURD. The extracted residue is well stirred up with warm water, and brought on to a small tared filter, and washed till the washings are free from chlorides. The filter-paper is then dried and weighed. The increase is taken as " curd." It does not as a rule exceed I'O per cent. Salt. This is estimated by evaporating the washings from the curd estimation^to- a convenient bulk, and titrating with N/10 silver ^ytrate and potassium chromate in the usual way. Total mineral matter, if required, may be estimated by difference, that is, by subtracting the total weight of the water, fat and curd from the weight of the sample taken. Boric acid does not interfere with the salt estimation. Quite recently attention has been drawn to the practice of some butter-makers of adding dried milk to butter. This is probably done in order to improve the flavour of butters which have " gone off " in cold storage. Without giving any opinion as to whether this practice is legitimate or not, we feel that in those cases in which the addition can be definitely proved it should not be passed over without comment, and it should be made clear to the consumer that an addition to the butter has been made, leaving it to him to decide whether such " renovated " butter is or is not sold to his prejudice as pure butter. The presence of added dried milk can be easily detected by determining the lactose present. This is done by washing out the butter as described below under boric acid, making the aqueous extract faintly acid with acetic acid, filtering off the precipitated cksein and titrating the filtrate against Fehling's solu- 108 Fatty Foods tion, taking 10 c.c. of Fehling's solution to require 0'07 grm. of lactose monohydrate to reduce it (p. 314). Pure butters seldom exceed 0-5 per cent, of lactose monohydrate, and any percentage much exceeding this limit points to the addition of dried milk to the butter. Peeseetativbs. Boron Compounds. — For rapid routine tests, a portion of the sample about the size of a pea is melted in a small crucible lid on the water bath, one drop of strong hydrochloric acid added and stirred in with a rod, and then five or six drops of a saturated alcoholic solution of turmeric. A rose red colour, immediately forming as the edges dry, indicates the presence of boron compounds, the colour being changed to a purple green on the addition of a drop of ammonia. The turmeric solution should be made by boiling about 50 grm. of turmeric powder with 250 c.c. of strong alcohol under a reflux condenser for at least an houi", and then filtering. This keeps well. The above method will detect 0"02 per cent, of boric acid with certainty. If greater delicacy be required, a larger quantity of the butter must be rapidly boiled down in a basin, the fat poured off, and the residue incinerated, acidified with HCl, and evaporated to dryness, with a turmeric paper placed down the side of the basin and dipping into the liquid. If, as the paper dries, it turns a rose red colour, and this colour is changed to hlue-hlach on moistening with ammonia, then boron compounds are indicated. Estimation.. — Twenty grm. of the sample ai-e weighed in a small beaker, and washed into a separating funnel with petroleum ether and water, using two quantities of about 10 c.c. of ether and three quantities of 10 c.c. Butter and Margarine 109 of water. After standing, tlie aqueous layer, up to the casein, is run off into a small flask and 25 c.c. mixed with 6—7 c.c. of 1 per cent, phenolphthalein solution in strong alcohol. The mixture is brought to the boil, and a distinct excess (2 to 3 c.c.) of N/10 NaOH added. The solu- tion is titrated back with N/10 acid till colourless, and then adjusted to the neutral point by alternate addi- tions of the acid and alkali, the solution being kept at the boiling point throughout the operation. Two grm. of mannitol are then added, the solution cooled and tiltrated with N/10 NaOH till just pink. If the water content of the butter be not known it may be taken as 15 per cent., when the amount of boric acid will be — 33 Boric acid (H3BO3) per cent. = X x — x 5 x 0'0062 where X = c.cs. of N/10 NaOH used. One c.c^ of N/10 NaOH = 0-0062 grm. of boric acid. The abpve is based on Richmond's method of esti- mating bijric acid. Pltjo^idbs (Hehner's method) . — One hundred grm. of the sai^ple are melted in a beaker and kept warm till the aqueous-^y4r settles. The contents of the beaker are poured into a separating funnel, the beaker being washed out with two lots of 25 c.c. of petroleum ether and then with two lots of 5 c.c. each of water. The contents of the! funnel are allowed to settle out and the aqueous layer run out iiato^ a platinum basin and rendered slightly alkaline with sodium carbonate, evaporated to drynesS;_and ignited till nearly white. The residue is taken up with water, and, without filtering, an excess of 20 per cent, solution of CaOlg added (this will be about 20 CO.) ; the mixture is brought to the boil and an excess of 110 Fatty Foods NagCO., (about 20 c.c. of a 20 per cent, solution) added. The liquid is filtered (the filtrate being tested to see if the precipitation be complete) and the filter and contents dried and ignited in a platinum basin. To the residue are added 30 c.c. of a 20 per cent, solution of acetic acid, and the whole allowed to boil on a sand-bath for fifteen minuteSj covered with an inverted funnel. The contents of the basin are then washed on to a small filter with hot acetic acid solution, and the filter dried and ignited in a platinum crucible and weighed. The residue is mixed with a few drops of pure concentrated HoSO^, and the crucible covered with a waxed watch-glass on which some design has been scratched with a pin. The watch- glass is filled with water and the crucible heated on a hot plate or the top of a water bath for thirty minutes. The glass is then removed and examined for etching. If fluorides are indicated the crucible is tljen ignited very cautiously in cm inclined position and re-weighed. The fluorides having been converted into sulphates by the operation, the amount present may be apnroximately calculated, since 78 parts of CaFg give 13C- parts of CaSO^. It is here assumed that all the .'esidue is calcium sulphate derived from calcium fluoride and in the case of butter this is practically correct. Jn very exact work a blank experiment with pure bi^tt^^ should be done and the small quantity of residil^ subtracted from the weight of CaSO^ found in the test. If boron compounds are known to be absent and an indication of the presence of fluorides or^ be required, the treatment with CaCla and NajCO^-iUay be omitted, as, in this case, the test will not be vitiated by the formation of volatile boro-fluorides. i Benzoates and Salicylates. — Fifty grm. are melted (less if only salicylates are being looked for) and shaken violently with 100 c.c. of water containing I'O per cent. Butter and Margarine 111 NaHCOg, in a separating funnel, and allowed to separate. The milky aqueous layer is run off, made slightly acid with hydrochloric acid, and neutralised to litmus paper with a solution of sodium hydroxide and cleared by adding 10 c.c. of Fehling copper sulphate solution, followed by 10 c.c. of a solution of potassium hydroxide (containing 31"15 grm. per litre) and the mixture filtered. The filtrate is rendered acid with sulphuric acid and shaken out with 50 c.c. of ether. The aqueous layer is run off and the ether washed two or three times with a few cubic centimetres of water. It is then mixed with about 10 c.c. of water and one drop of phenolphthalein solution, and a saturated solution of barium hydrate run in until, on violently shaking, the pink colour remains permanent. The aqueous layer is then filtered into a small porcelain dish and evaporated to 1 or 2 c.c, and again filtered through a small filter into a test-tube. Yery dikite acetic acid (1 : 100) is added drop by drop until the pink colour is discharged and then one or two drops of freshly prepared neutral solution of ferric chloride. In the presence of salicylates a violet colour will be produced, while in the case of benzoates a flesh- coloured turbidity or precipitate is formed. When both salicylic and benzoic acid are present together the solution, instead of being a clear violet, is turbid. Herde and Jacob (Zeit. Untersuch. Na.hr. Genussm., 1910, vol. xix, p. 137) state that salicylic acid can be destroyed in the presence of benzoic atid by means of potassium permanganate, but the authors have not been successful. with the process. Estimation of salicylic acid. — This will not often be required, and is rendered difficult on account of the solubility of the acid in fat. About 10 grm. of the sample are boiled with 20 c.c. of alcohol three times, pouring off the alcohol 112 Fatty Foods through a small filter after cooling sufficiently to solidify the fat. The alcoholic extract is diluted with 150 c.c. of water and neutralised with sodium hydroxide solution and 60 c.c. distilled off. To the residue are added 10 c.c. of N/sodium citrate (70 grm. of citric acid neutralised with NaOH and made up to 1000 c.c.) and then 10 c.c. basic lead acetate solution (p. 310), and then 10 c.c. of N/NaOH. Five c.c. of N/HCl are then added and 40 c.c. of a saturated solution of sodium chloride^ the whole made up to 250 c.c. and filtered. The salicylic acid is extracted from 200 c.c. of this filtrate by acidification with sulphuric acid and shaking out with ether. The ether is extracted by placing it in a separator with 20 c.c. of water and one drop of phenolphthalein solution and running in N/10 NaOH till the aqueous layer is permanently pink. This is run off and the ether washed with two more quantities of 20 c.c. of water and the water extract made up 100 c.c. In this solution the salicylic acid is easily and accurately determined colourimetrically in Nessler glasses against a standard solution of salicylic acid (O'Ol grm. per litre) using a solution of ammonia iron alum. The salicylic acid test solution should be diluted till the colour with the iron alum solution is easily compared with that of the standard. The iron alum solution is made by dissolving 1 grm. in 500 c.c. of water and boiling down to 100 c.c. and filtering before use. Two c.c. are added to 50 c.c. of the salicylic acid solutions in a Nessler glass. The above method is a modification of methods of Fellenberg, Harry, Mummery, Revis and Payne. CoLODEmct Matters — The following tests will usually be sufficient to detect such colours as are generally used in butter and margarine. If a more elaborate investigation be required, the method of Leeds (Analyst, 1887, vol. xxii, p. 150), or the still Tiiore elaborate Butter and Margarine 113 methods of Rota, should be employed (Ohem. Zeit., 1898, p. 437). Annatio. — A few grammes of the clear filtered fat are shaken with about 5 c.c. of warm dilute (10 per cent.) sodium hydroxide solution. The mixture is poured on to a wet filter and kept warm, till the greater part of the water has run through. The melted fat is then poured oif the filter-paper, which is gently washed with cold water. In the presence of annafcto, the paper will be stained a reddish-yellow of an intensity dependent on the quantity of annatto present. If the stain be very slight or indefinite, the paper should be dried and moistened with one drop of a 5 per cent, solution of citric acid, when in presence of traces of annatto a distinct pink colour will be produced, even when there is not sufficient iinnatto to produce a definite yellow stain on the paper. Azo-DTES (Doolittle's method). — About 2 grm. of the filtered fat are placed in each of two test-tubes together with 5 c.c. of petroleum ether to prevent solidification of the fat. To one»test-tube is added 1 c.c. of dilute (1 : 3) hydrochloric acid, and to the other 1 c.c. of dilute (10 per cent.) potassium hydroxide, the tubes being thoroughly shaken and allowed to stand. If an azo-dye be present, the lower layer in the hydrochloric acid tube will be pink, while that in the potash tube will remain colourless. In the case of annatto, the acid layer will remain colourless, while the alkaline layer (or, as more often happens, an intermediate layer of precipitated soap between the fat and alkaline layers) will be coloured a bright yellow. As a great number of mixtures of annatto and azo-dyes are sold for colour- ing purposes (and often as genuine annatto), both reactions will often be found to occur. Certain azo-dyes do not give satisfactory results with Doolittle's method 8 114 Fatty Foods in the presence of water, and in this case Low's method should be employed : For this purpose about 1 c.c. of the melted fat is mixed with 1 c.c. of a mixture of 1 part of concentrated sulphuric acid and four parts of glacial acetic acid, and the mixture heated nearly to boiling with constant shaking. In the presence of azo-dye the acid solution on settling out will be coloured pmfe or reddish, prac- tically no colour being produced in their absence. The colour often appears before heating. Examination of the Fat. A sufficient quantity of the sample is melted in a beaker at about 60° to 70° C, and kept warm until the bulk of. the water and curd have settled out. The liquid fat is then decanted through a plaited filter-paper, of a thick soft variety (C. S. and S., No. 604), prefer- ably in a hot-water funnel, as no partial solidification of the fat must be allowed during filtration. The fat is received into a clean dry bottle, and should be quite clear and bright, and, if (?onsidered necessary, a few pellets of dry filter-paper may be dropped into the fat to absorb any traces of water which may be present. Butter- Fat. The examination is usually made for the presence of margarine fats. Coconut oil is probably seldom added j)er se, but on account of the use of large quantities of this fat, as well as of palm kernel oil, in many modern margarines, these might easily occur. On account of the wide variations in the figures obtained for pure butter-fat, the greater number of tests are of little or no value by themselves, and have only a cumulative significance. It is necessary, therefore, to employ those Butter-Fat 116 tests winch will yield the most valuable information. Atter careful consideration the authors are of the opinion that the indications derived from the determi- nation of the Reichert-Meissl-Polenskft-Kirschner figures and of the baryta value by the method of Ave Lallement, supported by a few qualitative tests, will give all the information with regard to adulteration which analytical methods are capable of. It must, however, be remem- bered that adulteration with animal fats produces butters which are similar in their composition to those which are produced naturally by cows kept under abnormal conditions, or suffering from functional dis- orders of the udder. It is of importance, therefore, when coming to a decision with regard to butters which appear to be adulterated with small quantities of animal fats, to obtain, wherever possible, all the information available regarding the history of the sample, as the authors themselves do not hold the opinion that the abnormal should always be labelled as adulterated, or that the food value of a butter is dependent on its Reichert-Meissl value. If definite indications of the presence of coconut oil are found, it may be taken as practically certain that the sample is adulterated. The recent work of some investigators certainly shows that special types of feeding, particularly with coconut oil cake, may lead to albnormally high Polenske values, but it seems hardly probable that this practice is likely to afford much difficulty in ordinary work. Of the two main processes here given for the exami- nation of butter-fat, the Reichert-Meissl-Polenske- Kirschner process is based on the fact that butter-fat is unique in containing glycerides, into whose com- position butyric and caproic acids enter. It is these acids which account for nearly the^whole of the Reichertr Meissl value, while the Kirschner value is dependent 116 Fatty Foods almost exclusively on the butyric acid present, tlie caproic acid exerting but small effect, on account of the insolubility of silver caproate. The Polenske value is governed almost entirely by the presence of glycerides containing caprylic, capric and lauric acids, though, in practice, such acids as myristic and palmitic play a small part. It is for this reason that butter-fat, con- taining as it does but small quantities of glycerides having these water-insoluble-volatile acids, shows but a small Polenske value relative to the Reichert-Meissl value, while coconut products, which are largely made up of them, give a very high Polenske value. The addition of them to butter-fat, therefore, depresses the Reichert-Meissl value somewhat and decidedly raises the Polenske value. The process of Ave Lallement, depending as it does on the relative solubility of the barium salts of the fatty acids present, covers, indeed, much the same ground as the last process, but to the authors it has this advantage, viz. that whereas a considerable quantity of lard or margarine mixture (not containing coconut products), may be added to butters without transgressing the arbitrary limit of 24 for the Reichert-Meissl value, the transition from a negative to a positive value in the formula, b — (200 -|- c) of the Ave Lallement process takes place much more rapidly, and will often point to the addition of such mixtures when the evidence from the Reichert-Meissl value is not sufficient. Further, the inclusion of coconut products in the margarine mixture does not alter the evidential value of the formula b — (200-1- c), in sharp contra-distinction to the balancing effect which may be obtained for most of the values for butter when lard and coconut products .are added together. Butter-Fat 117 Beichert-Meissl-Folenske-Kirschner Method. Five grm. of the fat and 20 grm. of glycerine are weighed into a 300 c.c. flask, and 2 c.c. of a solution of sodium hydroxide (made by dissolving stick caustic soda in an equal weight of water) added. The flask is heated over a flame, with constant shaking, till it clears suddenly. The flask is allowed to cool somewhat and 100 c.c. of recently well-boiled, distilled water added, and solution of the soap effected. 0".l grm. of powdered pumice (which has been sifted through butter muslin, as the grade and quality are important) is added, and then 40 c.c. of sulphuric acid solution (60 grm. of con- centrated sulphuric acid diluted to 1000 c.c. and the solution adjusted so that 35 c.c. neutralise 2 c.c. of the above sodium hydroxide solution) . The flask is at once connected with the condenser and heated over a small flame till the insoluble acids are completely melted ; the flame is then increased and 110 c.c. distilled in nineteen to twenty-one minutes. Condenser water should be from 18—20'' 0. and dimensions of apparatus exactly as given by Polenske (see Fig. 5). When 110 c.c. have distilled, the flame is removed and a 25 c.c. cylinder placed under the condenser to catch any drops. The 110 c.c. flask and contents are stood in water at 10°— 15° C. for fifteen minutes. The contents of the 110 c.c. flask are then filtered and 100 c.c. titrated with N/10 baryta, using O'l c.c. of a 1 per cent, solution of phenolphthalein as indicator. This number of c.c. increased by one tenth, after subtraction of the blank (which must be determined in an exactly similar way, by using all the reagents except the fat)^ is the Beichert- ' In carrying out the blank test, the greatest care should be exercised not to overheat the glycerol soda mixture ; the solution after heating should be quite colourless. 118 Fatty Foods FlQ. 5. Butter-Fat 119 Meissl value. The condenser, cylinder and 110 c.c. receiver are washed with 18 c.o. of cold water, which are then poured over the filter used to filter the distil- late, and rejecteid. The condenser is washed out with four successive portions of 10 c.c. of neutral alcohol, which are received in the cylinder and poured over the filter into the 110 c.c. flask, the mixed alcohol solutions being then titrated with N/10 baryta or soda, using phenolphthalein as an indicator. A blank value is obtained in a similar way. The number of c.c. of N/10 alkali used, less the number used for the blank, is the Polenske value. Rather varying results have been obtained by various observers, but the following table will be found a very fair guide : Eeichert-Meissl values. 32 New butter values. 8-5 31 3-2 30 3-0 29 2-9 28 2-7 27 2-4 26 2-0 25 1-8 24 1-7 28 1-6 A "new butter value," exceeding by 0"5 c.c. the figure corresponding to the Reichert-Meissl value found, indicates the presence of coconut or palm kernel oil or their products. As it seems that the Kirschner value (see below) is practically a measure of the butyric acid present in the sample, it is evident that there will be greater evidence 120 Fatty Foods of adulteration with coconat oil if this figure be com- pared with the Polenske value. This point has not been sufficiently worked out yet^ but the authors venture to suggest the following values : Kirschner Polenske value. value. 20 1-6 22 2-1 24. 2-6 26 3-2 A variation of 1"0 must be allowed either way in the Polenske value corresponding to any particular Kirschner value, the addition of less than 5 per cent. of coconut oil causing the Polenske value to fall outside this limit. Kirschner s Extension of the above Method. To the 100 c.c. of the 110 c.c. distilled and titrated with baryta (care having been taken not to exceed the neutral point) is added 0"5 grm. of finely powdered silver sulphate, and the whole allowed to stand for an hour with occasional shaking. The liquid is then filtered, 100 c.c. measured off, 35 c.c. of water and 10 c.c. of sulphuric acid (as previously employed) added, together with a long piece of aluminium wire, and 110 c.c. again distilled^ff in the standard Reichert- Polenske apparatus in twenisy- rninu-fees ; of the distillate 100 c.c. are titrated with N/10 soda or baryta, and the number of c.c. used, corrected for the blank, is calcu- lated to the Kirschner value by the following formula : _ 121 (100 +' y ) ^ - "^ • 10000 where x = the corrected Kirschner titration, y = the number of c.c. of the baryta solution used to neutralise 100 c.c. of the original R'-M. distillate. Butter-Fat 121 The Barium Method of Ave-Iiallement. Five grm. of the filtered fat are saponified with 50 c.c. of approximately N/2 alcoholic sodium hydroxide (carefully standardised against N/2 HOI), boiling for thirty minutes (see saponification value). While the solution is still warm, it is titrated with N/2 hydro- chloric acid to phenolphthalein. The alcohol is then removed as completely as possible by boiling and blowing air into the flask. The soap is dissolved in hot, recently boiled, distilled water, and transferred to a 250 c.c. flask, brought to a temperature of 40° C, and made up to the mark at that temperature. One hundred c.c. are pipetted off into a flask, which is stood in a boiling water bath for five minutes, and then 50 c.c. of approximately N/5 barium chloride solution (about 25 grm. of crystallised BaOl2.2H30 in 1000 c.c.) are added. The mixture is allowed to remain for fifteen minutes in the water bath, to cause the insoluble, barium salts to coalesce. The contents of the flask are cooled and filtered into a 250 c.c. flask, washing the insoluble soaps well, and making up to the mark. Two hundred c.c. are pipetted into a beaker, acidified with 1 c.c. of concentrated HCl, heated on a sand- bath nearly to boiling, and 10 c.c. of approximately N/1 sulphuric acid then run in. The beaker is allowed to stand overnight, and the precipitate filtered on a Gooch crucible, and washed till free from chlorides, and finally with two quantities of 10 c.c. of warm alcohol. It is then dried to constant weight. The weight of barium sulphate found is increased by 25 per cent., and calcu- lated to BaO (BaSO^ x'0-657 = barium oxide). This is subtracted from the barium oxide value of the barium chloride solution (which must be standardised 122 Fatty Foods in an exactly similar way). The result is the. barium oxide value of the acids forming insoluble salts from 2 grm. of fat, and this, calculated to 1 grm. of the fat, = insoluble barium oxide value (&). The saponification value, also calculated to barium oxide (KOH x 1*367 = barium oxide) for 1 grm. of fat = total barium oxide value (a), and a — b = soluble barium oxide value (c), from which is calculated the value of fe — (200 + c). Butter-fat practically always gives a negative value for this formula, while both margarine and coconut products give a high positive value. The sufficient addition of these therefore, causes the value of the formula to become positive (see tables, pp. 126—127). In the case of the addition of coconut products, the insoluble barium oxide value is usually above 260. The method gives evidence of the presence of either coconut oil or lard compounds, the effect of the presence of these together being additive and not mutually destructive as in many other forms of investigation. The actual mean molecular weights are easily deter- mined by the use of the following formulae : b = Insoluble baryta value. c = Soluble baryta value. K = Saponification value. U = Unsaponifiable matter. M(, = Mean molecular weight of fatty acids forming insoluble barium salts. M| = Mean molecular weight of fatty acids forming soluble barium salts. S \ ftotal fatty acids. So \ weight of \ acids forming insoluble barium salts. Sj [acids forming soluble barium salts. ^ M,^^<_b. ^ ^Q^^ ■ 76-7 Butter-Fat 123 S = 1 - (K X 0-0002258 + U) Si = S — So Ml = ^Jii^ X 1000. c. Sq is directly estimated by drying and weighing the insoluble barium soaps and igniting. If the foregoing methods give indication of adultera- tion (see tables), it may then become desirable to make such, other determinations as the refractive index, iodine value, etc., in order to determine as closely as possible the amount of the added substances. It is also recommended to apply the following tests: The Valenta Test. The addition of lard and other animal fats causes the turbidity temperature of a solution of the gly- cerides in acetic acid to be raised, while that of coconut and palm kernel oils lowers the turbidity temperature. The turbidity temperature in the case of pure butter- fats unfortunately varies considerably, but the test, if intelligently performed and with due consideration of its limitations, is often of great help. The personal equation is a large factor in the results obtained, and determinations should be made with some pure butters by anyone using the test. The following is a simple and efficient method of carrying out the test : A long thin test-tube, preferably of Jena glass, about 0'5 in. in diameter, and sufficiently long to take in the scale of a thermometer up to 60° C, is marked accurately at 3 and 6 c.c. with lines all round the tube. A thermometer with very small bulb is fixed in the tube by means of a cork, so that the bulb is opposite the 3 c.c. line. The carefully dried butter-fat (filter- 124 Fatty Foods paper pellets should be. shaken up with it beforehand) is measured in at 27° to 29° 0. till the bottom of the meniscus coincides with the 3 c.c. line. Absolute acetic acid (Kahlbaum's is to be preferred) is then run in until the 6 c.c. line is reached (the acid should be measured at a definite temperature, say 15 to 16 C). The thermometer is inserted and the fat dissolved by shaking in water at about 50° 0. The tube is then withdrawn from the water and the contents allowed to cool in the air, shaking gently, and holding the tube in a good light. Immediately the faintest turbidity is noticed the temperature is read. The tub& is then slightly warmed and a fresh reading obtained. The end point is quite sharp, and consecutive readings should scarcely differ. The authors obtain figures usually lying between 30° and 42° 0. for pure butters, though they find low numbers prevailing during the early months of the year. Ten per cent, of lard produces a rise of about 7° C. The Halphen Test (see p. 52). As cotton-seed oil is never a constituent of pvire butter, a positive result may bo taken as absolute indi- cation of adulteration, in spite of the fact that cows fed on quantities of cotton-seed oil cake are said to produce butters giving positive results. At the present moment it may be taken that about 80 per cent, of the margarines in this country contain more or less cotton- seed oil. If present, a quantitative idea may be obtained by comparison. Baudouin's Test (see p. 54). As nearly all margarines made for sale on the con- tinent have to contain sesame oil, this test is of some Butter-Fat 125 significance. Care must be taken to do a bla.nk, using no furfuralj as if aniline colours be present a pink colour will be produced in the absence of sesame oil. If a colour be obtained in both blank and test^ the test will be a deeper colour if sesame oil be present. Phytosteryl Acetate Test (see p. 44). In exce.ptional cases in which the addition of a vegetable oil is suspected, but of which more definite evidence is required, this test becomes a final court of appeal. Microscopical Appearance of the Fat. The appearance of thin films of batter under pola- rised light affords some evidence of the addition of fats which have, in the course of their manufacture, been melted and re-set. When genuine butter is examined under a low power {\ in.) in a very thin film, made by pressing out a speck of the butter between a glass slide and cover slip, between crossed Nicols, the field appears uniformly dark. Light patches on the dark field point to the presence of margarine fat or renovated butter (*. e. butters which have been melted, washed with water and re-churned with milk) . The results of this test must be inter- preted with care as crystals of salt may be mistaken for crystals of fat. The examination is rendered easier if a cardboard tube be placed between the pre- paration and the microscope tube so as to enclose the objective. The Foam Test. Genuine butters when heated in a spoon over a small flame boil quietly with a gentle hissing, and with the 126 Fatty Foods production of considerable foam, while renovated butters crackle considerably with much spluttering and little foam. The results of this test may be conveniently noted when determining the water by Patrick's method. Interpretation of Results. The following table gives a number of results for various pure butters using the above mentioned tests . Butter. R-M. Pol. I K. Sajj. V. Total Ba. Insol. Ba. Sol. Ba. Difl. Danisli 30-01 2-15 ■ 23-i;7 228-3 312-0 253-4 58-6 - 5-2 30-58 2-52 23-0 228-7 312-6 249-0 63-6 -14-6 30-59 2-67 23-79 225-7 308-6 251-2 57-4 - 6-2 „ 29-71 2-68 24-lS 2311 315-9 252-9 62-9 -10-0 31-22 2-64 24-26 230-2 314-7 255-7 58-9 - 3-2 >• 30-28 209 21-31 227-9 311-5 252-9 58-7 - 5-8 New Zealand 3313 3-16 24-35 231-7 316-7 250-4 66-5 -161 ;, 33-51 2-35 27-57 229-3 313-4 244-7 68-7 -24-0 >. 30-83 2-15 25-63 228-0 311-6 253-9 57-7 - 3-8 „ 30-93 1-81 i 23-17 224-0 305-8 249-1 56-7 - 7-6 it 32-65 2-92 1 24-20 232-1 317-2 251-6 65-6 -14-0 English 29-79 2-44 i 21-94 228-7 312-7 253-0 59-7 - 6-7 30-49 2-59 23-71 228-7 312-6 253-8 58-8 - 5-0 „ 28-74 2-57 21-76 229-1 313-2 250-6 62-6 -12-0 „ 32-23 2-99 22-04 2320 317-1 254-5 62-6 - 8-1 French 29-88 2-65 21-59 228-0 311-6 252-4 59-3 - 6-9 jj 26-88 2-42 23-76 228-0 311-6 252-7 59-0 - 6-3 28-84 2-42 23-17 226-2 309-2 253-2 56-0 - 2-8 „ 31-49 3-08 22-01 231-5 316-4 251-7 64-7 -13-0 The methods of examination given above do not furnish indications of the actual amount of adulteration. The following figures will give some indication of the type of result obtained with mixtures of pure butter and coconut oil when using the Reiohert-Meissl-Polenske- Kirschner method and the method of Av6 Lallement. Butter-Fat 127 Butter A Butter A + io7o coconut oil Butter B Butter B + 10 7c coconut oil Butter C Butter C + io7o coconut oil Yal- enta. lod. No. R-M. val. Pol. val. 28-8 33-9 28-7 3-2 25-2 31-4 26-6 4-1 30-3' 36-5 28-1 2-5 26-4 33-7 25-8 3-9 27-0 35-7 305 3-5 240 32-8 280 4-3 Sap. val. Total Ba. (a) 228-4 I 312-2 231-1 315-9 227-8 230-3 2270 230-5 311-4 314-8 310-3 315-1 Insol. Ba. (J) Sol. Ba.(e) 6- (200+.;). 255-4 568 - 1-4 262-8 53-1 + 9-7 254-8 56-6 - 1-8 260-5 54-3 + 6-2 255-1 55-2 - 01 263-6 51-5 + 12-1 The mutual interdependence of the constants of butter-fat is well illustrated by the following table due to Thorpe (Journ. Chem. Soc, 1904, vol. Ixxiii, p. 254). No. of samples. Reiohert- Meissl number. Sp. gr. 37-9° 37-8° Saponifi- cation value. (1) Zeiss butyro- - refracto- meter niimbei- at 40°.(2; Soluble (S) acids per cent, on fat. Insoluble acids per cent, on fat. Mean molecular weight of insoluble acids. 7 17 15 27 37 51 78 56 41 18 10 22-5 23-5 24-5 25:5 26-5 27-5 28-8 29-5 30-5 31-3 32-6 0-9101 0-9104 0-9108 0-9110 0-9113 0-9114 0-9118 0-9120 0-9123 0-9125 0-9130 219-65 221-39 223-24 223-41 225-39 226-75 228-32 229-91 - 231-43 232-30 232-58 44-7 44-2 44-2 44-0 43-7 43-3 42-8 42-8 42-6 42-4 42-1 4-3 4-5 4-7 4-8 4-9 5-2 5-4 5-6 5-8 5-7 6-0 90-1 89-7 89-4 89-3 88-9 88-7 88-4 88-3 87-9 87-9 87-7 266-9 265-5 265-0 264-2 261-9 261-7 260-9 259-6 260-1 258-0 257-8 357 (') Calculated by Lewkowitscli from the saponification equivalents given by Thorpe. (2) Calculated from 45°C. by the authors. {^) Cal- culated as butyric acid. 1^8 Fatty Foods Reference to these figures will often be of help in detecting adulteration. Gross adulteration with lard products or margarine containing coconut or palm kernel oils with or without other vegetable or animal fats will be easily detected as already stated by the methods given. When, however, the addition is very small (say 10 per cent.) and is composed of oils and fats other than coconut or palm kernel products, the effective detection is decidedly difficult, unless some oil giving a characteristic qualita- tive test has been introduced. As already stated, in deciding the quantity of any such addition all possible values should be employed and the mean of their indica- tions returned. For public health control purposes in England a Reichert-Meissl value of 24 is taken to be the lowest limit for pure butter, but with the occurrence of butters obtained principally from Australia and New Zealand, having Reichert-Meissl values of 30 and over, the opportunity for the addition of animal fats is greats and will be passed over if reliance be placed on the Reichert- Meissl value alone. Such a case is exemplified by the following figures : Valenta Iodine value R.-M. value Pol. value Kirschner Sap. val. Total Ba. Insol. Ba Sol. Ba. (o) . b - (200 -I- c) {b) Butter. 30-2 38'7 per cent. 30-8 2-9 21-9 224-8 307-3 252-8 54-5 -1-7 Same butter + "1 10 per cent. lard, j" 35-2 40 6 per cent. , 27-7 2-4 19-5 221-8 303-2 254-6 48-6 + 6-0 Margarine Fat 129 Iodine Value. The iodine value may be used in certain cases for confirmatory purposes, but the variations in genuine butter are so wide that, in itself, it furnishes no evidence at all of adulteration. The authors have found the following to be fairly mean values ; Danish butters . 38 (Wijs) French „ . . . 37 „ Irish „ ... 42 „ The variations lie between 33 and 45. As the addi- tion of 10 per cent, coconut oil lowers the figure about 3, and 3 per cent, of lard only raises the figure about 9, little evidence is given by the iodine value. Margarine Fat. As the composition of margarine fat may be of a most complex nature, it will be necessary to make every determination ^which is likely to throw light on the composition (see p. 75 et seq.). The naponificatiun value (p. 21) is first determined as yielding the most useful information and guidance. If the value be below 193, the rape oil group or un- saponifiable matter are indicated and should be looked for (see " Unsaponifiable Matter," p. 48 ; " Hexabromide Test," p. 42, and "Tortelli and Fortini's method," p. 38). If the value be above 200, the presence of coconut, palm kernel, cohune oil or butter-fat may be tentatively inferred. Th« determination of the Reichert-Meissl-Polenslte- Kirschiier values* will show whether any such fats are actually present, and in particular, whether there is any butter-fat. The test having been carried out, reference * See Appendix A. 130 Fatty Foods / - / / / / / / / / 1 / / 1 1 t 4 / / yi 1^ / 1 / /* / f 6 / 1 1 / / / / ' 6 \ ^ / '} / / y y / / / V f! /> / / / 1 " i i 1 0= / / i f 1 t f / "?" ii*Mt ^ IW -- / / i /^ i / / / A '/ r-/ / ■7 / — H.rK M.«f ^ ler / / / tW-' ■^ EST — — / K^ s^ ■4S Uf -^ s _^ ■^ ' s s i % i s t i ■-'-I ? 'INaS Mill no IHKljajl KIW z^ 1 Margarine Fat 131 should be made to the following tables and curves, which have been compiled from determinations made s s = i 2 > 1 / // ^ — .-- 1 / 1 ^ / / X / ^ ^ t / / ;/ < 4.^ f / '. y ^^ h 4 <■• ^ /. ' K- / r ^ ^ / /. f # t " w' / / 4 ■5 '> / / / 4 ti OS / / / / / / / / / '/ . — // ' ^ ift» zr^ ■fX^V. - / / / ' / ;r" /^ ■ — ' ___ ' — / / ./ . __- Tti^ wo*'' ,/». 11- ' _^ — -' «hw * JM ^ •IN «i uad 110 1 INV030D s 1 6. by the authors using the standard process described on I. 117, and published in the Analyst, 1912, vol. xxxvii, K 183. P p. 183 132 Fatty Foods Polenske indicator value. Butter-fat. Coconut oil Oper cent. 2 per cent. 5 per cent. 10 per cent. per cent. 5 10 15 25 50 75 100 0-45 ■; 0-76 1 1-22 \ 1-75 1 2-91 i 710 1 1219 1 16-50 I K. E.-M. K. E.-M, K. E-M. K E.-M. K. E.-M. K. E.-M. K. E.-M. K. E.-M. 018 0-38 0-25 0-87 34 1-60 0-42 2-52 0-60 3-92 1-02 605 1-45 7-00 1-88 808 0-80 0-92 0-75 1-65 0-84 2-45 0-92 3-15 1-08 4-o7 1-50 6-88 1-92 8-35 1-49 1-70 1-55 2-15 ■ 1-60 3-42 1-68 4-35 1-82 5-55 2-20 7-72 2o5 9-20 2-70 3-25 2-82 4-10 2-90 4-90 2-96 555 3-08 6-55 3-38 8-95 3-70 10-50 Butler-fat. Palm kernel oil. Polenske indicator value. per 2 per 5 per 10 iwr cent. cent. cent. cent. per cent. 0-31 1 K. 0-18 0-80 1-49 2 70 E.-M. 0-38 0-92 1-70 3-25 5 0-59 i K. 0-30 0-80 1-51 2-70 E.-M. 0-o3 127 205 3 50 10 0-93 j K. E.-M. 0-35 1-00 0-85 1-62 154 250 2-75 4-00 15 1-30 1 K. 0-40 090 1-57 2-80 E.-M. 135 200 290 4-35 25 „ 1-67 I K. 0-48 0-97 1-65 2-S7 E.-M. 197 2 60 362 4-97 50 3 98 1 K. 0-72 1-16 1-!S7 3 12 E.-M. 3oO 40-5 4-92 6-22 75 6-70 1 K. 0-97 135 2-07 3-35 E.-M. 455 515 5-70 7-30 100 9-82 1 K. E.-M. 1-07 5-22 — — Margarine Fat 133 In using these tables and curves the Polenske value is first noted, as this acts as an "indicator," showing among which figures the R.-M. and Kirschner values are to be sought. The Kirschner value then determines the pres- ence (and probable percentage) or absence of butter- fat. The approximate agreement of the Reichert-Meissl figure with these indications determines the composition of the mixture. Of these three values (within the limits given) the Polenske value is practically independent of the presence of butter-fat. The values of the Reicliert- Meissl and Kirschner figures vary, however, with the composition of the butter-fat used, though the varia- tions are not wide. Por instance, 10 per cent of butter in a margarine gave a Kirschner value of 1 "5 for a butter of R.-M. value = 23"5, and a Kirschner value of . 2-7 for a butter of R.-M. value = 30-0. The R.-M. value varied similarly with the Kirschner value, the corre- sponding values in the above case being 2'2o and 3'25. The mean Kirschner and R.-M. values given in these tables and curves correspond to a butter giving an R.-M. of 30'0. In the presence of coconut and palm kernel oils there is a distinct difference in appearance of the insoluble acids which float on the surface of the R.-M. distillate _when cooled in water at 10° G. When palm kernel oil only is present the acids are white, hard and almost crystalline, while in the case of coconut oil they are nearly always partially oily. The values obtained in this process are practically independent of the constituents of a margarine mi.xture, other than coconut and palm kernel products and butter-fat. In- this lies the great value of the process, as it furnishes quantitative results which are a definite measure of a definite constituent ; whereas in the case of all other values, such as the saponification. I34 Fatty Foods iodine values, etc., the contributory effect of each con- stituent has to be considered. The values for the various mixtures given in the foregoing tables were obtained by the use of the whole oils. In practice, not only the whole oils, but the " stearines " either alone or in conjunction, may be employed. It therefore follows that figures will be sometimes obtained which do not fit exactly into the tables given. In such cases the percentage of these products is not so definite and has to be tentatively held and considered in conjunction with other analy- tical values. If the " stearines " are suspected the values of these, as given on pages 152 and 155, must also be taken into account. Cohune oil, which may soon be largely employed, behaves like coconut oil. Beef and lard fats should now be sought for by micro- scopical examination of the deposit from ethereal solution of the sample (pp. 56—57) (if necessary cooling to 0° C), bearing in mind that oleo oil, having a low " stearine" content, only gives deposits in concentrated solution, and with great difficulty in the presence of fats and oiU con- taining no " steariiie," as these exert a soloent influence. After this the refractometer nwniber should be deter- mined. In general, a reading over 50 will indicate the presence of a liquid vegetable oil. From the refractometer number (which in most margarines will lie between 35 and 60) an idea will be obtained of the weight of fat necessary to determine the ioiline value. For instance, if the refractometer number be 35, use 1"5 grm., if it be 60 use 0;1 grm., and for intermediate figures corresponding intermediate weights (the above quantities suppose that 10 c.c. of Wijs' solution is being employed). In the case of very low iodine values it is of course unnecessary to look Margarine Fat 135 for liquid oils. The iodine value having indicated liquid vegetable oils the presence of cotton-seed oil is sought by Halphe.n's reaction, I'lnd in the case of a positive result the percentage is ascertained by comparison with standards as described under that test. The presence of sesame oil is similarly sought by Baudouin's reaction, taking the same precaution against the presence of aniline colouring matters as noted under " Butter-fat/' and on page 54. A consideration of the above results will usually give considerable information as to the composition of the fat under examination^ for which purpose see the typical examples given on page 81 et seq. When definite indications of the presence of certain oils and fats has been obtained it may be fairly taken that they are actually present. There will, however^ always be a considerable number of vegetable oils which yield no distinctive individual characteristics^ but whose presence is evident from their general properties. To definitely state that such and such an oil is present in these cases is often beyond the power of the most experienced oil analyst. We can only refer the reader to the properties of the oils themselves and to careful attention to the behaviour of the fat under examination, as in many cases a clue can be obtained by little indications which are too many and too varied to enter into here. It may in certain cases be necessary to add to the above determinations that of the unsaponifiahle matter, this being, at the present time, practically the only analytical clue (not pi'oof) to the presence of solid vegetable fats which simulate lard and beef, such as shea nut oil, and certain fats of the " lUipe " group, some of which are distinguished by an abnormally high amount of unsaponifiahle -matter, to which reference 136 Fatty Foods will be made under their respective headings. The pliytosteryl acetate test also may be a final court of appeal when small quantities of vegetable fats are mixed with animal fats^ or vice-versa. In dealing with an oil for use in margarine manu- facture without any knowledge of its source, the deter- mination of the presence or absence of optical activity should never be omitted, as the possession of optical activity is a distinctive feature of the Chaulmugra group of fats, which are very poisonous. Fats used in Margarine Manufacture. It may be said, almost without contradiction, that at the present day there is scarcely an animal or vegetable fat (which can be procured in sufficient quantity) that has not been tried, or actually used in margarine, and even fish oils are being pressed into the service of the manufacturer. The consequence of this unfortunate idea that any oil or fat, if sufficiently refined, is suitable for the purpose has led on occasion to the use of con- signments of oils having poisonous properties, with disastrous results. It is, therefore, necessary when giving an opinion on any unknown oil to make quite sure that it does not belong to such a poisonous group. Modern processes of refinement and deodourisation have made it possible to render nearly every oil both odourless and tasteless, although certain oils are liable to regain some proportion of their original taste with too great a rapidity to allow of their genei'al use. As the variety of vegetable oils which are placed on the market is increasing daily, and in a most erratic manner, the examination of margarine mixtures is of a correspon- dingly variable nature. Consignments of fats are sold and used, and may perhaps never come again on the Margarine Fat 137 . market^ and what is more disconcerting, an entirely different fat may often be sold under the same name as that of a previous consignment. For instance the name "Illipe butter " covers a large variety of solid vegetable fats of totally different composition^ and the term is now practically void of any designative value. Considerations, both climatic and financial, largely govern the use of fats by particular makers at particular times, and therefore a study of market prices and the magnitude of consignments may often save the analyst the trouble of seeking for a fat which he is unlikely to find, and also prove a useful guide to the fats most likely to be present. The question of animal fats is particularly influenced by climatic conditions. For instance, margarine for Slimmer trade must have a higher melting-point than that for winter use, the, sale of margarine being quick on account of its liability to become rancid. In cold, weather the manufacturer will be using less premier- jus and more oleo oil, so that the former is inclined to fall in price, while paradoxically the price of oleo oil increases, because concurrently with its production oleo- stearine appears as a bye-product, for which the mar- garine manufacturer has, at such a time, less use than for premier-jus. A large content of oleo oil is generally restricted to high-class margarines. If, however, large quantities of suitable liquid vege- table oils are available, both oleo-stearine.and premier- jus will be used in the production of cheaper margarines, as they provide the necessary stiffness. Again, " price " almost entirely rules the question of neuti*al lard, which, although one of the. commonest con- stituents of margarine, usually disappears when the price advances beyond a certain limit. At the present time there is a growing tendency to 138 Fatty Foods use vegetable substitutes for lard, and although a variety of these have been on the market for sometime their use is now being reinforced by I'efined shea nut oil, and more particularly by shea nut " oleine." As vegetable lard substitutes, and, in fact, vegetable oils as a whole, are usually in a higher state of refinement than animal fats (particularly at level prices), this steady increase in the use of such fats is probably advantageous to the consumer. Though cheaper than beef-fat as a rule, mutton-fat is seldom used in margarine on account of its peculiar flavour, and finds an outlet instead among cake and biscuit makers. The great multiplicity of manufactured products obtained from oils, such as the commercial "oleines" and " stearines," and more recently the production of solid fats from liquid oils by " hydrogenization " pro- cesses,* render the interpretation of analytical figures very difficult, and unless very considerable every-dayand up-to-date experience has been had, it is not advisable for the analyst to return definite percentages of any constituent of a margarine, or to go a step beyond that for which there are absolutely definite analytical grounds. It is better to give fairly wide limits for each constituent, and where the presence of vegetable oils is inferred, but for which no definite confirmatory test exists, to return all such simply as " liquid vegetable oils " without attempting a more elaborate difEerentiation. Qhee (Qhi). In eastern countries on account of the high tempera- tures which prevail, this stable form of butter-fat largely takes the place of ordinary butter, more * See Appendix C. Ghee 139 especially for cooking purposes, etc. It is prepared by inoculating freshly boiled milk (usually of the buffalo) with sour milk, and after curdling has taken place, churning in a somewhat primitive manner. The fat so obtained is skimmed off, allowed to become somewhat rancid, and then heated in earthen pots until the water has evaporated. The fat clarifies on standing, and when sufficiently clear is run off from the sediment and con- stitutes " ghee" (see Analyst, 1910, vol. xxxv, p. 343, and 1911, vol. xxxvi, p. 392). The analytical figures are in the case of pure samples practically those of ordinary butter-fat, but the Reichert-Meissl values may be very high on account of the use of buffalo milk for its preparation. Ghee, how- ever, is seldom pure, and is much adulterated with both vegetable and animal fats. The animal fat is usually mutton fat, while the principal vegetable adulterants are phulwara butter and other Bassia fats, and less fre- quently liquid vegetable oils, such as arachis, cotton- seed, poppy-seed and sesame oils, as well as coconut oil (which is liquid in such countries as produce ghee). The examination of ghee follows the same lines as those for butter and margarine, and by means of the Kirschner value it is possible to absolutely certify the absence of butter-fat in " ghee substitutes." The following table ot figures gives a selection from analys'es made by the authors (for. reference see above) on pure and adulterated samples sold on the Eastern markets. It must be remembered that buffalo milk is chiefly used in the preparation of ghee, and that the fat of this milk seldom has a Reichert-Meissl value under 80, and may even approach 40. 140 Fatty Foods ^ 1 00 ■* ■* w O 05 in 00 m rH O CD rH S'l 1 S o CO CO 00 i-( "T) in O W lO CO cq 1 ;-( 1 P4 to ia CD 00 6 IN r-1 CO (M :| o CO CO lO w OS in CO cq cq cq -t- oq 1 ■* 00 cq ■^ Ai i) I I I cq CO to iH ^ in in CQ cq + I I s 3 3 3 5=" ^ i-l CJ) 05 CO CM cq cq in i> X —t TO 00 -^ i) 1-1 ^ CO rH CO cq I PM lO ' CO "^ "^^ "? V- O o « in OS in ■^ CO i-H cq CO CD CO cq cq + J! 13 g CO 1 CO 6 1-1 * (M U3 CO 05 05 U3 CO M oq Tq + r^ IS 6 cq th T-i 05 t~t us CO I CO M I I I 3 « O 1 ■^:S-|| 35 1 \t ai\ r- rt-Meissl ke value fication v bometer i (Zeiss so >1 1 -a 8 g ces-^ 3 o O g .S ,M Q P5 0HiBt5« 1 ^^ -+3 d-it ^ ■ ■ ■ 'J ■= s o ^11 §^+.^35 tacit Vegetable Oils and Fats 141 CHAPTER V. VEGETABLE OILS AND I'ATS. The vegetable oils and fats^ derived from oil-bearing seeds and fruits, etc., are obtained, in nearly all cases, by the employment of pressure, usually hydraulic. For this purpose the material is coarsely ground and pressed cold, by which process the finest oil is obtained and the product is commercially known as " cold drawn." The press cakes are then removed from the press, broken up and ground into a finer condition, heated in steam-jacketed kettles and again subjected to pressure while still warm. This second pressing produces an oil of slightly inferior quality, though, in some cases, there is no appreciable difference iu quality between the products of the first and second pressings. The high efficiency of- some modern presses has, however, enabled the complete expression of all the available oil to be carried out in one operation, but when using them the material requires nearly always to be warmed before pressing. Recently the Anderson screw press has come into operation as a substitute for hydraulic machinery. The press cakes derived from these processes usually contain from 5-10 per cent, of oil, according to the extent of pressure. Obviously the hard vegetable fats cannot be obtained by cold pressing, and must be heated both for first aiid second operations. 142 Fatty Foods Anothei- process, which, though by no means modern, has only of recent years become of practical importance, is that of extraction of the oleaginous material with volatile solvents — such as benzine, carbon bisulphide, carbon tetrachloride and trichlorethylene, etc. The earlier methods were not very efficient, chiefly on account of the impurity and unsuitability of the solvents employed, and also on account of the inefficiency of construction in the plant and of the difficulty of removing the solvent from both oil and meal. Modern extraction plants, however, have reached a high degree of efficiency in all these particulars, and considerable impetus has been given to the method by the introduc- tion of trichlorethijlene of great purity. This last point is one of great importance, as the insufficiently pure trichlorethylene, which was used at first, liberated small quantities of hydrochloric acid, which exerted a very deleterious action on the metal parts of the plant. Extracted oils and meals are now found containing such slight traces of solvent that its presence cannot be detected in them and the extracted meals are suitable for feeding purposes. It should be understood that the word " extracted," when used in conjunction with an oil, is commercially applied to those obtained by extrac- tion with solvent, no distinctive term being employed when " expressed " oils are referred to. An " ex- pressed" oil should not therefore be spoken of as " exti-acted " by pressure. Sometimes the two pro- cesses are combined, the meal is first pressed and the press-cake broken up and extracted. In many parts q| the world, especially those localities in which oil-bearing plants grow and in which modern machinery has not been introduced, the native produc- tion of oils and fats is still carried out in a very primitive fashion, such as by the use of wedge presses Vegetable Oils and Fats 143 or more often by boiling and pounding the material with water, the oil being afterwards removed by skim- ming. Oils so prepared are usually characterised by a high percentage of free fatty acids, and possess a bad colour, strong taste and smell, and contain a consider- able quantity of vegetable debris. Many oils, when first expressed, are more or less deeply coloured with matter derived from the meal. This is particularly the case when " extraction " pro- cesses are used, and is very pronounced when trichlor- ethylene is the solvent. The colour can, however, be removed or greatly diminished, practically in every case, by one or other of the numerous bleaching processes now employed, filtration through animal charcoal being one of the most common. Alkali refining,- though principally employed for the removal of free fatty acids, also effects considerable reduction in the colour. The above details are, of course, common knowledge, and are all that the analyst needs to know. The details of modern oil refining are naturally trade secrets of the various manufacturers and the authors therefore cannot be expected to disclose any such, particularly as they do not come within the scope of this work. It may not be commonly known that in oil mills, in which a variety of oils are produced, it would be much too costly a proceeding either to clean out the presses and tanks, etc., after each different batch of oil has been dealt with, or to have a separate plant for each particular variety of oleaginous material. It follows therefore necessarily that a certain admixture of a new batch with some of the preceding one will often take place, and this will account for the fact that traces of other oils are sometimes met with in samples which are commercially pure. Prom a commercial point of view 144 Fatty Foods no objection can be taken to this so long as the real product is not depreciated thereby. PeODDCTS OBrAINED FEOM VEGETABLE FaTS AND OiLS. " Oleines " and " stearines." — Some of the liquid vegetable oils, on prolonged standing, deposit solid fat, more particularly in cold weather or when artificially cooled. This deposit is termed " stearine," though it need not, by any means, contain any glyceride of stearic acid. The clear oil is syphoned off and the " stearine " obtained by filter-pressing the residue. A particular example of this type of product is cotton-seed " stearine," which is largely used in soap manufacture and to a limited extent for margarine, and, in fact, this " stearine " is practically the only one of any consider- able commercial importance obtained from liquid vege- table oils. Oils from which the " stearine " has been separated during cold weather are commercially known as "winter oils," and, as they remain clear at all times, find a ready use as salad oils. The term " summer oil " is applied to those from which the " stearine " has been but incompletely separated. In the case of solid vegetable fats, the " stearine " cannot be satisfactorily separated by any process of sedimentation at temperatures at which the oil is partially liquid. In such cases the fats are melted and allowed to " grain " {i. e. crystallise) by natural cooling. The solid mass is placed in bags and packed in layers between plates in a hydraulic press. The clear oil, which runs out on pressure, is known as the " oleine," while the hard residue left in the bags is, as in the case of animal fats, called the " stearine." While some of the softer fats may be pressed at ordinary temperatures, Vegetable Oils and Fats 14^ some of the harder varieties require a temperature which will maintain them in a sufficiently plastic con- dition to enable "oleine" to be expressed. Analytically, the tendency in the case of " stearines " is towards a lower iodine value than that of the original oil, but the other values follow no definite course, except, as would be expected, the melting-point rises. Foots. — This term is applied to any oleaginous sedimentary bye-product which is not wholly fatty. They may consist of soaps derived from alkali treatment, in which case they will include the alkali salts of the free fatty acids of the oil, together with neutral oil mechanically included, colouring matter, vegetable debris, free alkali and water, but the term is also applied to the vegetable debris obtained by sedimenta- tion of the oil during the process of clarification. In the following pages an attempt has been made to present the various vegetable oils and fats on a syste- matic plan. This has led to a somewhat stereotyped result, which, though satisfactory in the majority of cases, has been the cause of insufficient and unsuitable treatment in some instances, but it seemed that the advantages of tabulated presentation, both for ease of reference and for comparative purposes, outweighed these disadvantages. Under the analytical figures given will be found only those which we consider to be of real use in practical work, and for the more elaborate experimental investigations which may be carried out, reference must be made to other works. The column " Usual limits " is compiled from figures obtained by ourselves and by other investigators, omitting 10 146 Fatty Foods always any striking abnormalities, but giving sufficiently wide limits to include the greater number of samples. The column "Typical specimen" is in the greater majority of cases based on our own figures. Under the heading "Uses of non-fatty portion" we have given such analyses of meals and feeding-cakes, etc., as may be of interest. Except when otherwise stated, the figures are the results of our own analyses. The classification here adopted is that devised by C. Ainsworth Mitchell in Allen^s Organic Analysis, fourth edition, vol. ii. It is hardly necessary to say that only a selection from each group of oils is given here. Coconut Oil Group l^'' COCONUT OIL GROUP. 1. CocoNDT Oil* 2. Palm Kicenel Oil. 3. GoHUNK Oil. ■1-. Japan Tallow. . As some important manufacturers have recently decided to adopt the word " coconut j' (in preference to " coeoanut ") to designate the product obtained fi'om Oocos nucifera, on account of the confusion which sometimes arises with the product of Theoiroma Cacao (cocoa- butter), we have adopted this nomenclature. 148 Fatty Foods Coconut Oil. Trade Namks. — The refined oil, for edible purposes, is sold under various trade or proprietary names, of which the following are among the best known : Nucoline, CocolardOjVegetaline, Palmine, Nutrex, Nutto, vegetable lard, nut lard, etc. SoDECE. — The fruit of Coros nucifera (the ordinary coconut tree). The husk of the nut is removed by hand. The nut is split in two and dried in the sun, preferably on sand, which takes about three days, the shell coming off on the first day. The " meat " so dried and removed is termed "copra." Large quantities are, however, dried in kilns, but the product fetches a lower price as the colour is darker unle.'-s special precautions are taken. The copra is usually shipped and pressed in Europe to avoid the loss of oil in transit by leakages and to save the expense of packing, and on account of the more efficient methods of pressing employed. About one fourth of the weight of the original nut (without libre, but with shell and water) is "copra," which should not contain more than 5 per cent, of moisture. CoKTKNT OP Oil (in copra). — 63—70 per cent. On pressing 62—63 per cent, of oil and 37—38 per cent, of oil-cake are usually obtained from the average copra. Description of Oil.— Appearance. — "White toyellowish fat, according to the quality of the copra and the method of preparation; when melted it forms a colour- less to brownish-yellow oil. Consistency. — Solid, greasy, somewhat crystalline fat, with a certain degree of brittleness, especially at low tempei'atures, but becomes soft and nearly liquid in summer (see " Melting-point "). Coconut Oil 149 Taste and smell. — When unrefined, tliese may vary from the pleasant fresh taste and smell of coconut to strongly rancid and disagreeable. The best refined edible oils are absolutely odourless and tasteless. PossiBM Addlteeants. — Actual adulteration is un- common, palm kernel oil being the only likely adulterant, bat an occasional form of sophistication is the removal of part of the " stearine," when the remainder is sold as " whole oil " ; such treatment lowers the setting point and raises the iodine valve. L [ABILITY TO Rancidity. — The best refined edible oils keep well, but the unrefined oil rapidly turns rancid, especially in the presence of moulds. Special Tests. — Distinguished .by a high Reichert- Meissl value and very high Polenske value. Also by the highest saponification value of the commonly used fats (see palm kernel oil, manufactured products) . The iodine valae is low and the solubility in alcohol unusually great. Hinks (Analyst, 1907, vol. xxxii, p. 160) has taken advantage of this to separate a crystalline glyce- ride, which enables small quantities to be detected in butter by a microscopical examination of the alcoholic extract obtained under certain conditions. Uses of the Oil. — The refined " whole oil " is used to an enormous extent in every form of foodstufiE to which a solid fat may require to be added ; probably no other vegetable fat is so extensively used, or adapted to so many purposes, and its use is daily increasing, as a constituent of margarine, cakes, biscuits, sweetmeats, etc. Manueactuked Products feom the Oil : TheiuUses. — " Oleines " and '■ stearines ^' are obtained by hydraulic pressure by methods described on p. 144. The refined " stearines " are chiefly used as cocoa- butter substitutes in the manufacture of chocolate, from 150 Fatty Foods which, a portion of the true cocoa-butter has been removed^ in order to produce a cheaper, though perfectly wholesome and nutritious material, thus enabling the public to enjoy a sweetmeat at a price otherwise impossible, and of which enormous quantities are now made. The " stearines " are also used to some extent in the manufacture of biscuits,' notably "puff biscuits"; also for pharmaceutical purposes, and to a limited extent in margarine (the whole oil being generally employed for the latter purpose). The refined " oleine " is used for a very large variety of dietetic purposes, where a fat softer and of lower melting-point than whole coconut oil is required, but chiefly as a baking fat for biscuits, pastry and cakes. It figures largely in sweetmeats such as toffee, caramels and in the icing of biscuits. Uses of Non-Fatty Portion. —Of the non-fatty por- tions of the coconut, including the fibre, every portion is brought into use, with the exception of the milk, which is invariably wasted. The^6re is used for matting, etc. The hush as fuel for the kiln-drying of the copra. The residual cake (often called " poonack ") furnishes a most valuable cattle food, the importance of which, while readily appreciated on the continent, has not yet been recognised in " conservative " England. Attention has been drawn to the value of the cake by the Midland and Dairy College in a complete series of experiments, in which it is compared with other feeding stuffs (Analyst, 1911, xxxvi, p. 445). The manufacture of desiccated coconut is a very important industry carried out on scientific lines. This product contains more oil than copra itself owing to the fact that the rind, which contains less oil than the meat, has been removed, and also to the lower moisture content. The average percentage of oil is 68-72, and any lower quantity must be attributed to partial removal Coconut Shavings Oil 151 of the oil. This practice is by no means unknown and is carried out by various manipulations. When examining this product the presence of sugar, starch, etc., must be sought for as adulterants. Good desiccated coconut should be of a snow-white colour and have a fresh and pleasant taste of the nut, and contain not more than O'l per cent, of free fatty acids, calculated as oleic, on the oil present. Coconut Shavings Oil. — In the preparation of desic- cated coconut, the brown skin or rind next the shell is first removed. This rind is produced in sufficient quantity and has a sufficiently high fat content to make it worth while to express or extract the oil. This is often sold as coconut oil for want of a better description. Its analytical figures (see p. 86) show marked divergence from those of whole coconut oil. These differences have on occasion led to the condemnation of this product as coconut oil adulterated with other oils, and care should therefore be taken in condemning coconut oils with abnormal figures. Analysis of Coconut Oil Cake. Moisture . 12-22 Oil ... . . 7-oO Albuminoids* 19-37 Digestible carbohydrates . 42-33 Woody fibre . 12-10 Mineral matter! 6-48 100-00 - Including nitrogen, S'lO per cent, t Including sand, 0'92 per cent- 162 Fatty Foods Coconut Oil. Determination. Usual limits. Typical speci- men. Melting-point ° C, incipient fusion 20° to 22=' 22° „ „ complete fusion 23° to 26° 26° Solidifying point ° C. . 22° to 23i° 23-2° Saponification value .... 255 to 258 257-3 Refractive index at 40° C. (Zeiss butyro- refractometer) . 34-7 to, 35-5 35-2 Iodine value (Wijs) 7-9 to S-8 8-62 Specific gravity, {i° C. — frc. . 0-869 to 0-874 0-8698 Free fatty acids (as oleic) 3 to 7 per 009 per cent.* cent.t Unsaponifiable matter 0-15 to 0-3 per cent. 0-18 per cent Melting-point of fatty acids ° C. 22° to 26° 25° Special values : Beicliert-Meissl 6 to 8 7-5 Polenske 15 to 18 16-5 Kirschner ... 1-6 to 1-9 1-8 Av^-Lallement values : (a) Total baryta — 351-8 (b) Insol. baryta — 294-5 (c) Soluble baryta — 57-3 6 — (200 + c) . — -1- 37-2 Valenta No., ° C. 12 to 15 * Crude unrefined, f Refined. Determinalion. Melting-point (com- plete fusion) Solidifying point Saponification value Refractive index (Zeiss buty ro-r efracto- meter at 40° C.) Iodine value Specific gravity ||° C. Coconut " Stearine." From 32° C. do-ivn to that of whole oil. From 29° C. do-wn to that of whole oil. From 250 up to that of whole oil. From 34-6 up to that of whole oil. Coconut " Oleine." From 15° C. up to that of whole oil. From 13° C. up to that of whole oil. From 262 down that of whole oil. From 35-7 down that of whole oil. to to From 2-0 up to that From 14-0 down to of whole oil. that of whole oil. From 0-860 up to — that of whole oil. As these are manufaotvu-ed products made in a great variety of melting-points, according to the extent of pressure, only the very extreme figures are given, and practically all oommeroial samples yield figures well between the above limits. Palm Kernel Oil 153 Palm Kernel Oil. SonuCE. — Seeds of Elseis guineensis (the Africa;n oil palm, see Frontispiece). After the palm oil has been removed from the fleshy portion of the fruit (see "Palm Oil"), the remaining seeds are freed from their shells by hand, in a somewhat primitive manner, by women and children. The decorticated kernels are imported into this country, where' they are ground and crushed in a manner similar to that employed for copra and oleaginous seeds (p. 141). Weight of 100 kernels = 85 to 120 grm. Content of Oil. — The kernels contain 46-53 per cent, of oil, and on pressing, about 45 per cent, of oil and 55 per cent, of cake, containing about 8—10 per cent, of fat, are usually obtained. Dkscuiption of Ojl. — Appearance. — Like coconut oil, though perhaps somewhat darker in colour. Consistency. — Less brittle, and more greasy than coconut oi], though of higher melting-point. Taste and smell. — The oil has a strong, very charac- teristic and tenacious taste and smell. Possible Adulterants. — Not usually adulterated. Liability to Rancidity. — Dependent on state of refinement. The crude oil rapidly becomes rancid, while the well-refined product keeps well. Special Tksts. — It is characterised by the same peculiarities as coconut oil, but the actual values are different. Uses of Oil. — The " stearine,'' whole oil, and " oleine " are all used for the same purposes and in quite a similar manner to the allied products from 154 Fatty Foods coconut oil. It is quite common to find mixtures of the two oils or their products. Owing to their great similarity there is considerable difficulty in estimating and detecting these oils in the presence of one anothei-, although the experienced worker, who has by him sufficient data of these oils and their products, is able to arrive at very close approximations to their respec- tive quantities in many cases. " Stearine " from palm kernel oil is produced com- mercially of a somewhat higher melting-point than that of coconut "stearine," and the "oleine" may have a lower melting-point than that of coconut " oleine." There has been, in the past, an idea that the keeping properties of palm kernel products were not as good as those of coconut oil, but owing to improved modern methods of refining this distinction has passed away. Manufactured Phoddcts peom Oil. — " Oleines " and "stearines" quite similar to those of coconut oil are made commercially. It may be worthy of note that, while in the case of coconut oil the saponification value of the " stearine" is lower than that of the whole oil and decreases in proportion to the rise in the melting-point, that of palm; kernel " stearine," on the other hand, increases. For instance, coconut and palm kernel oils giving saponiQcation values of 257 and 247 respectively, will both yield "stearines" having values in the neighbourhood of 2oO. Uses op the Non-Fatty Portion. — A feeding cake or meal is obtained after pressing or extraction, very similar to that of coconut oil, but slightly inferior to that from the latter. For analysis see p. 156. Palm Kernel Oil 155 Palm Kernel Oil. Determination. Usual limits. Typical speci- men. M elting-point, ° C, incipient fusion „ „ „ complete fusion Solidifying point, ° C. Saponification value .... Eefractive index at 40° C. (Zeiss butyro- refractometer) . ... Iodine value (Wijs) Specific gravity, il" C. . . . ■^f C. . Free fatty acids (as oleic) Unsaponifiable matter . Melting-points of fatty acids, ° C. Crude oil. 21° to 24° 26° to 29° 24° to 26-5° 245 to 248 36 to 37-5 14 to 19 0-859 to 0-871 5 to 12 per cent. 26° to 29° Refined oil. 25-5° 28-5° 25-2° 246-2 36-8 14-65 0-8702 0-12 per cent. 0-22 per cent. 28-0° Special values : Eeichert-Meissl value Polenske value . Kirschner value Baryta value : (a) Total . (b) Insoluble (e) Soluble 6 - (2U0 + c) . . — 5-22 9-82 1-07 336-6 303-7 32-9 + 70-8 De/ermination. Melting - point, com- plete fusion Solidifying point Saponification value Refractive index at 40° C. (Zeiss butyro- refractometer) Iodine value Specific gravity -11° C. Palm kernel "'stearine" From 35° C. down to that of ihe whole oil. Prom 32° C. down to that of the whole oil. From 252 down to that of the whole oil. From 35-5 up to that of the whole oil. From 330 up to that of the whole oil. From 0-866 up to that of the whole oil. Palm kernel " oleine." From 23° C. up to that of the whole oil. From 21° C. up to that of the whole oil. From 244 up to that of the whole oil. From 37-5 down to that of the whole oil. From 300 down to that of the whole oil. As these are manufactured products made in a great variety of 156 Fatty Foods melting-points, according to the extent of pressure, only the very- extreme figures are given, and practically all the commercial samples yield figures well between the above limits. Analysis of Cake. Moisture 11-83 Feat . 6-57 Albuminoids* . 16-94 Carbohydrates . 47-14 Woody fibre . 13-80 Mineral mattert 3 7:i cent. 100-00 Including nitrogen, 2-71 per cent, t Including sand, 052 per Cohune Oil 157 Cohune Oil. SouECE. — Attalea Cohune (the Cohune Palm of Honduras^ Guatemala, Brazil and Guiana). The fruit (as shown in sketch) varies considerably in size and consists of (a) a hard woody shell, nearly \ in. thick, in texture like that of the coconut. This shell Fia. 8. — Fruit of Attalea Cohune : a — woody husk; b ■ (Natural size.) ■ seed. encloses the oval seed or "kernel" {b), which resembles the "meat" of the coconut in taste, smell and appear- ance, but contains more oil, which, as will be seen from the analytical figures, is very similar in composition to coconut oil. (The oil from Attalea funifera is so similar that it may be included under the head of "cohune oil," though it is sometimes distinguished under the name of " coquito " or " coquilla oil.") Weight of 100 seeds or "kernels" = 470 grm. 158 Fatty Foods The seeds or "kernels" represent 10 per cent, of the fruit. Content op Fat (in seed). — 65—72 per cent. Description op Fat. — Appearance. — White and solid. Consistency. — Rather less brittle than coconut oil. Taste. — Pleasant nutty flavour. Smell. — Recalling that of fresh coconuts. Special Tests. — High Reichert-Meissl and Polenske values, with other characteristics similar to those of coconut oilj but the slightly higher iodine value and lower solidifying point should be noted. Uses op the Fat. — As far as we are aware, the oil has not so far been put to any commercial use and has only been prepared on a more or less experimental scale. Several machines have been invented to remove the very solid and woody husk, but one recently devised appears to be specially suitable, and has the additional advantage of portability. As a result of this invention the immense quantity of cohune nuts available may be expected to be utilised, and large quantities of cohune oil may shortly appear on the market. The fat will find a ready use for margarine manufac- ture and will doubtless be used for baking and biscuit making, for all of which purposes the refined fat will be very suitable. Analysis of Kernels and Oil Cake. Moisture .... Kernels. 4-16 Oil cake (calculated) 12-4 Oil 68-53 6-5 Albuminoids 7-12 . 21-2 Digestible carbohydrates 13-09 . 38-8 Woody fibre . 5-42 . 16-1 Mineral matter 1-68 5-0 100-00 100-0 Cohune Oil 169 OoHDNE Oil. Determination. Usual limits. Typical speci- men. Melting-point ° C , incipient fusion „ „ „ complete fusion 22° to 24= — Solidifying point, ° C. . . 19° to 21-5° ]9-8° Saponification value 254 to 260 258-5 Refractive index at 40~ C. (Zeiss butyro- 3.5-2 to 36-2 35-7 refraotometer) Iodine value (Wii's) 9-0 to 12-5 11-92 Specific gravity, if° C — z Frie fatty acids (as oleic) Usually 0-25 3-5 per to 6-0 per cent. cent. Unsaponifiable matter .... — — Melting points of fatty acids, ° C. 25° to 29° — Special values : 1 Eeichert-Meissl value - — 8-4 Polenske value . — 15-87 Kirschner value — 1-57 Baryta values : (a) Total . ... — 848-3 . (6) Insoluble — 292-8 (c) Soluble — 55-5 b - (200 + c) . + 37-3 160 Fatty Foods Japan Tallow. Teade Name. — Japan wax. SouKCK. — The berries or fruits of Rhus vernicifera and Bhufi succedanea, largely grown in Japan as tlie source of lacquer, the tallow being merely a bye- product. It is prepared in a very rough and ready manner, so that the product which comes on this market is often found to be mixed with the non-fatty portion of the berry and with water. Weight of 100 berries, about 10 grm. Content of Fat. — 20 to 30 per cent. Desceii'tion of Fat. — Appearanc.(\ — Dark green or brownish ; a bleached variety is also sold. Consistency. — Very hard, brittle and wax-like. Taste and smell. — Strong and characteristic. Slightly suggestive of tallow. Liability to Kancidity. — Not great^ after being freed from non-fatty impurities. Possible Adui.tkeants. — Perilla oil, or other con- venient vegetable oils, which are added with the intention of assisting the extraction of the tallow by rendering it more fluid, rather than for actual adulteration. Special Tests. — High saponification value and very high specific gravity. Iodine. value low. It should be noted that though the substance is commercially termed Japan " wax," the name is incorrectly applied, as it consists of triglycerides. Its insolubility in ether (from which it crystallises in a characteristic form of needles) and somewhat unusual solubility in hot alcohol, may afford assistance in detecting it in mixtures. Usiis OP Fat. — According to Sachs (Chem. Rev. Japan Tallow 161 Pett-u-Harz. Ind., 1908^ xv, p. 30), 25 per cent, of Japan wax and 75 per cent, coconut "stearine " are used as a cocoa-butter substitute. It is probably not used other- wise for edible purposes. Japan Tallow. Determination. Usual limita. Typical speci- men. Melting-point, ° C, incipient fusion _ _ „ „ „ complete fusion 50° to 53° — Solidifying point, ° C 47° to 50° 47-5° Saponification value .... 209 to 220 219-2 Eefractive index at 40° O. (Zeiss btityro- refractometer) observed at higher temperature and calculated to 40^ C. . 47-6 to 49-7* 48-0 Iodine value (Wijs) 5 to 17 11-27 1 Specific gravity, \l° C. ... 0-975 to 0-999 — „ ifo. . . . 0-875 to 0-877 — ■Pree fatty acids (as oleic) . ■ Seldojn ex- ceeds 10 per cent. 3-12 per cent. tJnsaponifiable matter . 1 to 15 per cent. — Melting-point of fatty acids, ° C. . 56° to 57° Berg, Chem. Zeit., 1903, p. 755. 11 \ Cocoa Butter Group 163 COCOA BUTTER GROUP. 1. Cocoa Butter. 2. Borneo Tallow. 3. Chinese Vegetable Tallow. 4. Palm Oil. 5. Mafuea Oil and Tallow. 6. Malabar Tallow. 7. KoKDM Butter. 164 Fatty Foods Cocoa = Butter. Source. — Seeds of Theohroma Cacao, the well-known cocoa tree, which is extensively cultivated in the tropics, and more recently in Ceylon. Cocoa-butter is a bye-product of the cocoa industrj'^, in the manufacture of which the fat is expressed by hydraulic pressure so as to leave, as a rule, some 25 to 30 per cent, of fat in the pressed cake. Weight of 100 seeds = 100-110 grm., but from certain districts may weigh as much as 150 grm. Content of Fat : (a) In whole bean, 40—50 per cent. (6) In nib, 45—55 per cent, (c) In shell, 2—5 per cent. Description op Fat. — Appearance. — Buttery-yellow. Consistency. — Hard and very brittle. Taste and smell. — Like cocoa. Possible Adulterants. — Coconut and palm kernel " stearines," beef " stearine," " green butters," and various other hard vegetable fats. Liability to Rancidity. — It is often stated never to go rancid. This is not strictly true, but it is certainly less susceptible to this change than almost any other fat. Special Tests. — Bjorklund's test (p. 58). As the fat is never deodourised commercially, it retains its characteristic " cocoa " smell, by which it may be recognised, even when present in reasonably small quantities with other fats. Uses of Fat. — Chiefly in the manufacture of choco- late. For this purpose the mixture of cocoa matter Cocoa Butter 165 and sagar has to be incorporated with added cocoa- butter in order to allow of its being moulded, or when required for chocolate coating, to render it sufRciently fluid when warmed, for the purpose. Uses of Non-Fatty Poetion. — Oocoa or chocolate (see Chap. VII). Oocoa-Btjttee. Determination. Usual limit's. Typical speci- men. Melting-point, ° C, incipient fusion _ 30-5° „ „ „ complete fusion 32° to 34-.5° 33-5° Solidifying point, " C. ... 27° to 29-5° 29-4° Saponification value 192 to 198 193-5 Eefractive index at 40° C. (Zeiss butyro- 46 to 47-5 46-7 refractometer) Iodine value (Wijs) 35 to 40 ■ 37-25 Specific gravity, -^° C. — 0-968 „ WC. . — — Free fatty acids (as oleic) Seldom 0-65 per exceeds 3 per cent. cent. TJnsaponifiable matter . 0-8 to 1-2 1-07 per per cent. cent. Melting-point of fatty acids, ° C. 49 to 51 — Special values : Baryta values : (a) Total — 264-6 (6) Insoluble — 261-9 (c) Soluble . — 2-7 h - (200 + c) . — -1- 59-2 166 Fatty Foods Borneo Tallow. Native Name. — " Minjak Tangkawang." Source. — The seeds of various members of the family DipterocarpesB, of which Shorea ghysbertiana, Shorea PiGt. 9. — 1. Shorea roiusta, fruit. (Natural size.) 2. Portion of seed of Shorea stenopt^ra, as shipped. (Natural size.) aptera, Shorea stenoptera, Shorea rohusta, Isoplera horneensis and species of Hopea are among the chief representatives: they are natives of Borneo and the Straits Settlements. The seeds are usually sent into this country under the name of " Pontianak illipe nuts," being distin- guished by the prefix " large " or " small." This does not convey any botanical distinction, but is simply a Borneo Tallow 167 commercial differentiation of a purely dimensional character. The nature of the fruit will be seen from the drawing of S. robusta ; the other members of the family have, in many cases, a similar fruit varying only in size, but the persistent calyx, which forms the characteristic winged fruit, is removed before shipment. The seed itself usually splits up into two or more segments, and whole seeds are rarely seen in this country. The trees often grow in the vicinity of the sea, and the seeds, falling therein, are carried by tidal influences into various spots where they collect and are salved by the natives. Owing to the various botanical sources of these seeds, it follows naturally that samples of the fat vary somewhat in their analytical figures. The limits given by us are selected from figures obtained from a very large number of samples. Content of Fat. — In seed, 47-60 per cent. Desceiption of Fat. -^Appearance. — Yellow to green solid fat, easily bleaching. Consistency. — Hard and brittle, resembling cocoa- butter. Taste and smell. — The taste of the native-prepared fat is most unpleasant, and the product is very crude, but that produced by modern European processes often has only a verj^ slight taste, and a smell which is somewhat suggestive of cocoa-butter. Special Tests. — The analytical figures are very similar to those of cocoa-butter and afford no means of distinguishing one from the other. On applying Bjorklund's test this fat behaves in a very similar manner to cocoa-butter.' 1 Many methods of distinguishing these fats from true cocoa-butter have been proposed, but are not satisfactory. The authors hope shortly to publish some work on this subject. ■ 168 Fatty Foods Uses of Pat. — As a cocoa-butter substitute. Uses of Non-Fatty Portion. — The non-fatty portion is included among the various articles sold as " Ulipe cake/' and gives the follo-wing a.nalytical figures : Moisture 12-20 Fat 7-53 Albuminoids .... 11-32 Digestible carbohydrates . 55"35 Woody fibre . . . 7-64 Mineral matter . . . 5-96 100-00 Borneo' Tallow. Determination. Usual limits. Typical speci- men. Melting-point, ° C, incipient fusion „ „ „ complete fusion Solidifying point, " C. . . . Saponification value .... Kefractive index at 40° C. (Zeiss butyro- refractometer) Iodine value (Wi]'s) Specific gravity, i|° C. . Free fatty acids (as oleic) Unsaponifiable matter Melting-point of fatty acids, ° C. 34° to 36° 37-5° to 39° 28° to 32° 188 to 207* 45 to 47 29 to 38 Up to 25 per cent.t or even more 0-7 to 2-0 per cent. 49° to 51° 34-5° 38-2° 30-0° 194-3 461 34-25 0-8597 4-96 0-88 per cent 50° Special values : Baryta values : (a) Total . (b) Insoluble (c) Soluble . 6 - (200 + c) . 267-3 to 282-9 2591 to 267-8 8-2 to 15-1 + 50-9 to + 52-7 * Extreme limits. t The European prepared fat contains much smaller quantities. Chinese Vegetable Tallow 169 Chinese Vegetable Tallow. Native Name of Seeds — " Cay-soi." SotriiCE. — Stillingia sehifera [Groton sehiferum or Sapium sehiferum) Natural order Buphorbiacse ; indi- Fia. 10. — (1) Fruit of Stillingia sebifera, natural size. (2) Longi- tudinal section of single coccus x 3. m. Mesocarp. e. Endo- carp. s. Seed. (3) Single fruit. genous to China, but has been introduced into North India. The fruitj Fig. 10/ is a schizocarp splitting into three cocci. The brown outer husk (epicarp) falls off, leaving the tallowy layer exposed. Fig. 10 (2) shows a section of a single coccus or nutlet, which consists of : 170 Fatty Foods (m) The white tallowy material (mesocarp) which envelopes the seed and contains the tallow. (e) A hard brown shell {endocarp) enclosing — (s) A small soft yellow kernel (seed) which contains the liquid oil . Three commercial fatty products are obtained : — («) Chinese vegetable tallow (known locally as pi-yu) described below, obtained by steaming the seeds, in perforated vessels, which causes the tallow to run out. (fe) The liquid oil, known in this country as stillingia oil and in China as ting-yu, obtained by pressing the residue from the rendering of the tallow, , after crushing. (c) A mixture of the oil and tallow, produced from the whole berry in one operation, this mixture being locally known as mou-ieoit. Weight of 100 berries = 20 grm. The inner kernel represents 30 per cent, of the whole berry. CoNTKNT OF Fatty Matter. — 20'5 per cent, of tallow; 19 per cent, of liquid oil on whole berry. Description op Tallow. — Appearance. — The com- mercial tallow is a pale to dark green colour with a tendency to crack, while that extracted in the laboratory by means of petroleum ether from the clean berries, after separation of the kernels, is white and marble- like, a fact which rather suggests that the green colour of the commercial product is derived from leaves or stalks. Consistency. — Very hard and brittle. Taste and smell. — Tallow-like and characteristic. Optical Activity (of pure tallow). — None, but the liquid stillingia oil from the kernel has a rotary power of about 10 angular degrees in a 200 mm. tube. Chinese Vegetable Tallow 171 Special Tests. — Saponification value slightly higher than most fats. Iodine value low (see table) . It should be tested for freedom from stillingia oil, which is easily detected by its optical activity (stipra) as well as by its high iodine value and refractive index, as shown by the following figures obtained on a sample of pure stillingia oil extracted ' from kernels which the authors have separated by hand. Iodine value (Wijs) . . . 187-6 Refractive index (Zeiss butyro- refractometer at 40° C.) . 78-1 Uses of Tallow. — Although very large quantities of the tallow are imported for use in the manufacture of soap and candles, small amounts are occasionally used as a means of stiffening softer edible fats. The tallow is extensively used for edible purposes in China, but the liquid stillingia oil is alleged to be poisonous — a statement which is in need of confirmation, seeing that a certain proportion of the liquid oil is most probably extracted with the tallow when crude methods are employed, but until further information is obtained on this point one must exclude from edible uses any sample which is shown by analysis to contain stillingia oil. Description op the Mixed Oil and Tallow {mou- ieou). — This product usually consists of a hard white solid fat, but not so hard as the pure tallow and having a slightly greasy feeling which is not apparent in the latter. On keeping it rapidly develops a rancid smell and turns brown on the outside owing to the oxidation of the liquid oil. Uses of the Non-Fatty Portion. — A black dye-stuff for silk is made from the leaves. 172 Fatty Foods Many of the figures recorded by various observers have undoubtedly been obtained on samples which were not pure tallow, and were either admixed with small quantities of the liquid oil or were actually mou-ieou. Chinese Vegetable Tallow. Determination. Usual limits. Typical speci- men. Melting-point, ° C, incipient fusion 50° to 55° „ „ „ complete fusion : 55° to 65° — Solidifying point 39° to 42° 40-3° Saponification value 199 to 206 203-2 Refractive index at 40° C. (Zeiss butyro- 42 to 45 44-3 refractometer) Iodine value (Wijs) 18 to 29 18-26 Specific gravity, ^^" C. — — ., 4f°C. . — 0-8560 Free fatty acids (as oleic) From 2 per 342 per cent | cent, up- wards ; sel- dom exceeds 10 per cent. Unsaponifiable matter . 1 to 2 per 1-25 per cent cent. 1 Melting-points of fatty acids, ° C. . — — Optical rotation 0-0 — The two following analyses have been selected as typical of a number of samples, examined by the authors, of the commercial product containing some of the liquid oil. Containing a little liquid oil. Melting-point, ° C. 43-8° — Solidifying point, ° C. . 30-0° 27-7° Saponification value 206-3 200-2 Refractive index at 40° C. (Zeiss butyro-refractometer) 46-1 47-8 Iodine value (Wijs) 32-6 44-2 Specific gravity, ff° C. — 0-8623 Palm Oil 173 Palm Oil. SouECB. — Fruit of Elseii ijuineensis and other species^ as well as varieties of the former. The oil palm is indigenous to West Africa ; it occurs in great abundance in Sierra Leone, Senegal, Southern Nigeria, Gold Coast, French Congo, etc., never at any great distance from the sea. It is also found in Brazil, South America, West Indies, etc., and has been introduced into Borneo, Java, and the Philippine Islands. The fruits are borne tightly clustered together in large heads ; each fruit is about one and a half inches in length by one inch in diameter, and consists of an outer fleshy portion composed of soft pulp, usually of an orange or brownish-red colour ; this surrounds the palm " nut," which encloses, as a rule, one kernel, but sometimes more (Frontispiece). The colour and size of the fruits vary somewhat in the different varieties, and of these, the varieties yielding thin-shelled nuts are more profitable, in that they yield a larger proportion of palm kernels, and a higher per- centage of oil.^ The fruits, having been separated from the heads, are first soaked in water to soften them, after which they are roughly crushed and set aside to ferment. The oil is separated in a very crude manner by the natives, either by skimming off that which rises to the top, or that which is liberated by boiling. The oil liberated by boiling from the fresh fruit is called "chop oil" and is not imported, but used locally for food purposes (Lewkowitsch) . ' BuU. Imp. Inst., 1909, p. 4. 174 Fatty Foods In this way the seeds are set free and are separately treated for the production of palm kernel oil (see p. 153). Content op Oil. — In pulp, 55—70 per cent. Description of Oil. — Appearance. — Granular, and possessing a characteristic shade of orange red. Consistency . — On an average rather stifEer than tallow, but very rancid samples are much stiffer. Taste. — Usually very rancid. Smell. — The smell of the unrefined oil is very charac- teristic and not easily forgotten ; to a certain extent it recalls the odour of violets. Special Tests. — The Liebermann-Storch test gives in the heavier liquid a blue colour tinged with green. Malphen method. — 100 c.c. of the filtered fat (filtra- tion must be carried out at a temperature not above 70° and as quickly as possible) are immediately mixed with three times the volume of petroleum ether and shaken with 50 c.c. of 0'5 per cent. KHO. After settling out, the aqueous layer is removed and shaken with 10 c.c. of carbon tetrachloride. This is then separated, and to it, in a porcelain dish, are added 2 c.c. of a mixture of two parts carbon tetrachloride, con- taining 5 drops of hydrobromic acid (sp. gr. 1'19) and one part of crystallised phenol, and the whole gently mixed. An immediate hluish-green colour indicates palm oil. Owing to the presence of an enzyme in the fresh fruit which rapidly develops acidity if the fruit be allowed to stand before extraction, the oil arrives in this country in a very acid condition, the amount of free fatty acids being at least 15 per cent, but more usually nearer 50 per cent. This variation is, exemplified by the following figures : Accra addah . 51-2 per cent. Benin . . 37-a y.er cent. New Calabar . 31-3 „ Bed Sherbros 221 Palm Oil 175 Uses of Oil. — The amount of refined oil is very limited, one of its uses being to colour margarines, where artificial colours are not permitted. This limited use may be explained by the fact that the acidity cannot be profitably removed ; but there is no doubt that when, in time to come, fresh fruit is extracted and the present crude methods of the natives are superseded by more efficient ones, and by the use of proper machinery, a very much larger quantity of this oil will be used in margarine. It may be noted that the neutral oil has a very low solidifying point, and would in many cases be in a melted condition during the summer in this country. Uses of Non-Patty Portion. — None. Palm Oil. Determmation. Usual limits. Typical specimen. Melting-point, ° C, incipient fusion 20° to 40° _ „ „ „ complete fusion 25° to 50° 49° Solidifying point, ■* C. . . 35° to 40° — Saponification value 198 to 202 199-2 Eefractive index at 40° C. (Zeiss butyro- 41 to 45 42-0 refraotometer) 56-3 Iodine value (Wijs) 52 to 57 Specific gravity ii° C. . 99° P 0-921 to 0-924 — Tree fatty acids (as oleic) 15 to 80 per cent. 48-2 per cent. TJnsaponifiable matter . — — Melting-point of fatty acids, ° C. — — Titer . . . . 43° to 45° ~ 176 Fatty Foods Mafura (or Mafoureira) Oil and Tallow. iSouEOE. — Seeds of Mafourewa oleifera (Triehilia emetica) found on the East African Coast. The fruit (according to Daniel and McCrae, Analyst^ 1908, vol. xxxiii, p. 276) is about 1'8 cm. long and contains three loculi, in each of which there is from one to three seeds, each being about 1^ cm. long by 1 cm. broad and covered with a reddish-brown oily pulp. The appearance of the seed is characteristic : the shell is a bright yellowish-red, with a large black spot on one side ; it encloses a brown kernel, which readily splits into two parts. In old seeds, as a result of drying, the colour of the seed is sometimes uniformly brown. As the shell of the seed is thin and brittle the kernels are easily separated. The oil is obtained from the shell while tbe tallow is the product of the kernels. Weight of 100 seeds = 35 grm. Content of Oil and Tallow : (a) Shell, 35-40 per cent. (Mafura oil). (6) Kernel, 60—65 per cent. (Mafura tallow). (c) Whole seed, 55 per cent, (oil and tallow). Descbiption of Oil and Tallow : Appearance. — Oil: Very pale yellow. Tallow: Light brown. CoNSiSTKNCY. — Oil: Liquid at ordinary tempfi-a- tures, but solidifies at a few degrees above the freeziug- point. Tallow: Firm, but plastic. Malabar Oil and Tallow W7 O o TS CO QO r-H FH T T iS^ "p ,s!5°^ o oM Ofe " 1 03t-ffq 0) O to lOlM^ '*' O-^Irt '5s 05 05 '"' Sg (M cq P i Q ^ CQ E^ g o , "Po "",,&«., « ?S ^lo» 5l l|l i -p'.m" gi Fl cS M . 3 a i^.cd fi^ o O i3 a a3 ra i ^ ^lilii§ Sii?i CFl 'm 1 ' ■ * o *H >i 4J ^ o o '^ •i-t 'c! m CO M m S !S • ■ -fl CM ^ <3; i- o; S- f3 -P fi 53 01 •< -^^ 5 •S r-H Q o •?4 1 s _p O d 'O +3 ' %^ .pH r-j 1 fl &Sfl> >.-2 § -« I* fe 2 m " P rrf 01 a U 01 <""§ 2 a s n a c3 a-^ >. U.H;g Q CD -i-i a -So o-e " «= p-g 12 178 Fatty Foods Optical Activity. — The oil is inactive. Uses of Oil and Tallow. — The oil is edible, and is used locally for such purposes. Some doubt seems to attach to the tallow, which has been stated to be poisonous, and must be so regarded till proof to the contrary is forthcoming. Uses of Non-Fattt Portion. — The cake is unsuitable for use as a cattle-food, since it is very bitter and probably possesses emetic properties.' 1 Bull. Imp. Inst., 1908, vol. iv, p. 371. Malabar Tallow 179 Malabar Tallow (Piney Tallow). Source. — Seeds of Vateria indica (Dipterocarpese), a tree indigenous to fhe Bast Indies, and growing on the Malabar coast, etc. The seed (Fig. 11) is enclosed in a hard, brown husk, about ^—^ in. in thickness. Fia. 11. — Fruit of Vateria indica size.) Seed tallow tree), (f natural Content of Oil. — 49 per cent. (Lewkowitsch) . Description of Fat. — Appearance. — Greenish-yellow to white in colour ; is easily bleached by the action of light.: JOonsistency. — Rather stifEer than lard. Taste and smell. — Practically none, or very slight. Uses of Fat. — In India it is used as an edible fat, but in other countries, as yet, only for soap and candle manufacture. The tree also yields piney resin or white dammar, used in the manufacture of varnish. 180 Fatty Foods Mala-bae Tallow. Determination, Usual limits. Autliority. Melting-point, ° C, incipient fusion _ „ „ „ complete fusion 30° to 40° Solidifying point, ° C 25° to 35° Saponification value _ . . . . 189 to 192 Refractive index at 40 ° C. (Zeiss butyro- 47-5 (Crossleyand retractometer) Le Sueur) Iodine value 37-8 to 39-6 (Crossley and Le Sueur) Specific gravity \f C. . — Free fatty acids (as oleic) 5 to 15% (Crossleyand Le Sueur) Unsaponifiable matter . — — Melting-point of fatty acids, " C. 56-6° to 63-8° (Hbhnel and Wolfbauer) Kokum Butter 181 Kokum Butter. Source. — Seeds of Garcinia indica, Indiaj East Indies, etc. The small, black ovoid seeds are enclosed in a fleshy fruit about 1^—2 in. in diameter (Fig. 12). The fat is, at present, only prepared by the natives by the usual crude method of boiling and clarifying. Content op Fat. — In seed, 20—25 per cent. (Lewko- witsch) , Desceiption op Fat. — Appearance. — A solid, white fat. Consistency. — Hard and brittle. 'Taste and smell. — Not very pronounced. Special Tests. — None. Uses op Fat. — Much employed in India as an edible fat, where it is also used as one of the many adulterants of ghee. Recorded Constants. Melting point* . . 42-0° C. Solidifying point . . 37-6° C. (Heise) Saponification value* . 186'8 Refractive index at 40° C* . 46-0 (Zeiss butyro-refractometer) Iodine value* . . . 34'21 Melting-point of fatty acids . 60°-61°C. (Heise) * Orossley and Le Sueur, Journ. Soo. Chem. Ind., 1898, vol. xvii, p. 992. 182 SFatty Foods Fio. 12. — Fruits of Garcinia indica, Kokmu fruits. (Natural size.) The Ba^sia Group 183 THE BASSIA GROUP. Of this group we describe : (1) Bassia (Butybospeemdm) Paekii (Shea nut butter) . (2) Bassia. BUTYEACEA (Phulwara butter). (3) MiMusoPS Djave (Djave butter). (4) Bassia latifolia. (5) Bassia longifolia. Of these (1) and (3) are not as a rule confused with other members of the group, but in the case of the others there has been the greatest confusion. The two last named, owing to great similarity in appearance, are constantly shipped in a mixed condi- tion and are also mutually confounded. For this reason we propose that the names Illipe butter, Mowrah, Mahua, Mohwrah, Phulwara fats, Mee oil, etc., be abandoned, and that the names butyracea fat, longifolia fat and latifolia fat be substituted, as by this means confusion would be avoided and identification simplified. The drawings of these seeds, here given, are to scale, and show clearly the differences of shape. The analytical figures have been obtained on the oils extracted by us from the actual seeds. 184 Fatty Fpods Shea Nut Oil. Trade Names. — Shea butter, Karite butter. Other names : Bambuk butter, G-alam butter. Native Names. — Karit6, Ce, etc. Souncis. — Butyrospevmum [ov Bassia) Parkii. Natural Order Sapotacese, a large tree grown in West Africa, Fig. 13. — Butyrospermum Parkii. 1. Fruit (natui-al size). (Shea nut) (4 natural size). 3. Section of seed. Seed French Soudan, etc. In the drfed condition the fruit consists of a thin, brownish-red crinkled shell, loosely enclosing the dark-brown egg-shaped seed, which is about the size of a plum. The kernel, which is enclosed in a glazed fragile husk, is of a chocolate colour, this colour forming a distinction between this seed and many others of a similar shape and structure, which yield a fat closely related, but at the same time distinctly different, to Shea butter. The amount of shell varies enormously in different Shea Nut Oil 185 speoimensj according to their origin, but usually amounts to some 35—38 per cent, on the whole seed. The oil is prepared in a crude manner by the natives, who remove the pulp and bury it in the earth, when fermentation takes place. In this way the seeds are liberated aud are afterwards ground between stones and the fat extracted with boiJing water. Weight of 100 seeds = 344 grm. Content op Pat. — (a) Whole seed, 34-44 per cent.; (6) kernel, 50—60 per cent. Description of Fat. — Appearance. — The colour of the native prepared article varies from muddy brown to a greenish grey, according to the care with which it has been prepared. The refined, bleached or European product may, however, be almost white. Consistency. — A stiff, plastic fat, somewhat granular and under certain conditions, of a " stringy " nature. Taste and smell. — The native-prepared article has a strong taste and smell, while the refined European product varies from little to practi- cally no taste and smell. Special Tests. — Usually characterised by a high content of unsaponifiable matter, which is seldom less than 5 per cent, and often approaches 10 per cent. The crystals obtained from ether are somewhat similar in shape to those of beef fat. Saponification value, low. Refractive index unusually high for such a solid fat. Uses oe Fat.- — The refined fat, which can be made practically tasteless and odourless, finds an increasing use for edible products, although its keeping properties are not specially good. The "oleine" is used for margarine and baking purposes, and the " stearine " finds a use as a cocoa-butter substitute. 186 Fatty Foods Uses of Non-Fatty Poetion. — It has been suggested that the cake should be used for cattle-feedings but there seems to be some doubt as to its wholesomeness, and, as will be seen from the following analysis, the percentage of nitrogen is so small that its nutritive value cannot be very great. Analysis of: Moisture . Whole kernel. . 5-35 . ExUacted meal . 9-76 Fat . . 49-87 . . 2-93 Albuminoids^ . G-97 . 12-39 Carbohydrates Woody fibre Mineral matter 31-44 . . 2-99 . 3-38 . . 61-30 . 7-46 . 6-16 100-00 100-00 ' Containing : Nitrogen I'll per cent. 1-98 per cent. Shea Nut Oil Shea Nut Oil. 187 Determination. Melting-point, ° C, incipient fusion „ „ „ complete fusion Solidifying point, ° C. Saponification value Refractive index at 40° C. (Zeiss butyro refractometer) Iodine value (Wijs) Specific gravity, \^" C. J» 33 1 S ^' • Free fatty acids (as oleic) Unsaponifiable matter . Melting-point of fatty acids, ° C. Usual limits. Typical speci- men. 29° to 32° 37° to 42° 25° to 30° 180 to 190 55-5 to 56-5 57 to 63 From 2% I upwards ! 5 to 9 % I 41-2° 26-8° 186-9 56-3 58-93 8-29% 7-56 % Determination. Shea nut *' stearine." Shea nut ** oleine." Melting-point, ° C, incipient fusion 40-0° _ „ „ „ complete fusion 55-5° — Solidifying point, ° C. . 34-2° 24° Saponification value 179-7 181-6 Eefractive index at 40° C. (Zeiss butyro- 52-7 58-7 refractometer) Iodine value . 51-9 62-3 Free fatty acids 3-4 % 5-89% Unsaponifiable matter 6-25 % 7-72 % Keiohert-Meissl value — 2-60 Polenske value — 0-72 188 Fatty Foods Butyracea Fat. Native Name. — " Phalwa." Source. — -The seeds of Bassia butyracea, which occurs in the sub-Himalayan disfcricts^ from the Ganges to Bhutan, and is found growing up to an altitude of 15,000 ft. (Bull. Imp. Inst., 1911, vol. iii, p. 229.) The fruit consists of a thick, soft pericarp, which is blackish in colour and encloses one seed. This seed is made up of PiQ. 14. — Seeds of Bassia butyracea. From G-arhwal, N.W. Provinces, India. (Natural size.) a pale brown, highly polished, brittle shell which loosely surrounds a whitish almond-shaped kernel, smaller than that of B. latifolia or B. longifoUa (compare drawings). The seeds are known by the natives as " phulwara " and the fat as " phulwa." Weight of 100 seeds = 78 grm. ; weight of husk = 26 per cent. Content gji' Fat. — 66 per cent, on kernel, 49 per cent, on whole seed. Desceiption of Fat. — Appearaiice. — White and lard- like. Consistency. — Firm, and rather stifEer than lard. Taste and smell. — Pleasant and not very pronounced Butyracea Fat 189 when freshly prepared ; and if then freed from im- 'purities its keeping properties are exceptionally good — but the crudely prepared fat rapidly develops a very strong taste and smell. Special Tests. — There are no special tests. It may, however, be noted that (1) the saponification value is rather low, considering the comparatively small amount of unsaponifiable matter, and (2) there is a distinct Reichert-Meissl and Polenske value. Uses of Pat. — Largely used as an adulterant of ghee, for which purpose it specially lends itself, any taste and smell being completely disguised by that of the ghee. Uses oe Non-Fatty Portions. — None known. For analytical values see p. 195. 190 Fatty Foods Djave Butter. Nativk Name of Tree. — " Noumgou " in the Came- roons, and in G-aboon "Njave" or "Djave." The fruits are known as " Ourere " and " Abeku " (maho- gany nuts) in Gaboon and the G-old Coast Colony respectively. 1 Source. — The seeds of Mimusops Djave (or Bassia Pia. 15. — Seed of Mimasops Djave (Djave nut). (Natural size.) toxisperma) , a tree belonging to the natural order Sapotaceas and indigenous to West Africa (Gold Coast^ Nigeria^ etc.). The nuts are about 2|— 3 in. long by I5 in. broad ; they consist of a mahogany-coloured shell, highly polished, as in the case of the other members of this group, having a long oval hilum on one side, as seen in sketch (Fig. 15). This shell encloses a kernel which resembles a shea nut kernel in appearance, but is creamy white inside. Weight of 100 nuts = 1850 grm. ; weight of husk = 50 per cent. 1 Catalogue of exhibits. Gold Coast Colony. Franco-British Ex- hibition, 1908. Djave Butter 191 Content of Fat. — (a) Whole nut, about 35 per cent. ; {h) kernel, 65—70 per cent. Description of Fat. — Appearance. — Like shea butter, but much softer. Consistency. — Usually solid, but the more neutral samples have a very low melting-point, 60 —65 F. Taste and smell. — In no case strongly pronounced but vary according to mode of preparation. Uses op Fat. — Probably at the moment only used for soap-making, but may at any time be put to edible use, in which case it would be necessary to render it free from traces of hydrocyanic acid, which are derived from the enzymic decomposition of the non-fatty portion. Uses of Non-fatty Portion. — None, owing to the presence of a powerful cyanogenetic glucoside. The following figures have been obtained on a sample of the oil, extracted from the seeds with petroleum ether by the authors : Solidifying point °C. . . . 21-0 Saponification value . . . 184'2 Refractive index at 40 °C. (Zeiss butyro-refractometer) . 61'8 Iodine value (Wijs) . . 65'1 Specific gravity ff °C. . . 0-8578 Free fatty acids (as oleic) . . 9'27 per cent. • Unsaponifiable matter . . 3"86 per cent. Melting point of fatty acids °C. . 52-8 Solidifying point of fatty acids °C. 47-8 192 Fatty Foods Latifolia Fat. Native Names. — In' India Bassia latifolia is known under the names " mowrah/' " mahua," " mahwa/' etc., as well as " illupei " or " illupai," the former being Hindustani and the latter, Tamil names.' SouECHi. — Seeds of Bassia latifolia, which occurs mainly in Central India, from Western Bengal to Burma : it does not extend to Southern India. Fig. 16.- -Seed of Bassia latifolia (natural size). Central Provinces, India. From Seoni, The seeds are intermediate in size between those of B. longifolia and B. hutyracea, the average size being shown in the sketch. As in the other members of this group, the yellowish-brown polished shell of the seed encloses a brownish-red kernel, which is usually split into two parts. Weight of 100 seeds = 200 grm. ; weight of husk = 22 per cent. Content of Fat. — (a) On whole seed, about 45 per cent. ; (6) on kernel, 57—60 per cent. Description of Fat. — Appearance. — Buttery-yellow 1 Bull. Imp. Inst., 19H, vol. iii, p. 228. Latifolia Fat 193 when properly prepared, but bleaches readily on keep- ing. The native-made article may be very dirty and greenish in colour. Consistency. — Plastic, and rather softer and more greasy than butter. Taste and smell. — Pleasant when properly pre- pared, but the native-made article is often very disagreeable. Possible ^lDTJLteeants. — The fat is often mixed with other Bassia fats, not so much as an adulteration, as from the similarity of the seeds. Special Tests. — None. Uses of Pat. — It is used for edible purposes in India, particularly as a substitute for tallow and as an adulterant of ghee. In Europe, its principal use at the moment is for soap and candle manufacture. Uses of Non-Patty Portion. — The cake is bitter and poisonous, and is used as a manure. For analytical values see p. 195. 13 194 Fatty Foods Longifolia Fat. Native Names. — According to Sir , George Watt (The Commercial Products of India, 1908, p. 116) Bassia longifolia is known by the same native name as B. latifolia, but as Bafsia longifolia only occurs in the m Fig. 17. — Seed of Bassia longifolia. From Madras. (Natural size.) southern parts of India, it is usually known by its Tamil name of " illipi " or " illupei." SouECE. — The seeds of Bassia longifolia, occurring in Southern India only. The seeds are very similar to those of B. latifolia, but are longer and narrower. The fat is probably the original " Illip6 fat." Weight of 100 seeds = 157 grm.; weight of husk = 25 per cent. Content of Oil. — Usually about 55 per cent, on the kernel. Longifolia Fat 195 Description of Fat. — The fat is very similar to that of Bassia latifolia. Special Tests. — None. Uses of Fat. — As for latifolia fat. Uses op Non-Fattt Poetion. — The cake is poisonous, and according to the Biochemical Journal, 1910, p. 93, it contains a saponin-like glucoside. The taste is very bitter. The following figures have been obtained by the authors for fats extracted by them from the true seeds : Determination, Bassia lati- Bassia loTlgi- Bassia huty- folia. folia. raeea. Eetective index at 40° C. (Zeiss 47-7 49-3 47-8 butyro-refractometer Iodine value (Wijs) 59-4 62-6 42-6 Saponification value . 192-2 189-8 188-2 Specific gravity, ^° C. 0-8595 0-8624 — Free fatty acids (as oleic) 24-6 % 3-3% 8-74 % TJnsaponifiable matter — — 1-36 Baryta value : (a) Total 2630 258-2 257-3 (6) Insoluble . 252-0 252-8 255-7 (c) Soluble 11-0 5-4 1-6 I - (200 + c) + 41-0 + 47-4 + 54-1 Keichert-Meissl value . — — 1-31 Polenske value . 1 — — 0-65 1 Olive Oil Group 197 OLIVE OIL GROUP. 1. Olive Oil. 2. Aeachis Oil. 3. The Almomd Oil Geoup. 198 Fatty Foods Olive Oil. Source. — The fruits of different varieties of Olea europoea (one of which is shown in the sketch), the well- known olive tree, which is extensively cultivated in the countries bordering on the Mediterranean, especially Italy, Spain and France, the best oil coming from Lucca and its environs. The oval fruit is about the size of a small plum, and varies in colour from yellowish-green to purple-green. The oil is obtained from the fleshy portion which surrounds the seed. Olive oil varies in quantity and quality according to the degree of ripeness of the fruit from which it is obtained, the oil from the fruit which is fully ripe being inferior, more acid, and therefore liable to develop rancidity at an earlier stage. Content of Oil. — This is very variable, ranging from 30—65 per cent. Description of Oil. — Appearance. — The best quality oil is a greenish-yellow. Consistency. — A limpid oil, liquid at ordinary tem- peratures, and beginning to deposit " stearine " below 10° C. Taste and smell. — The oil is valued more particu- larly on its flavour. The finest edible oil has practically no smell and very little taste, but with inferior qualities there is a distinct odour and the taste becomes sharp and unpleasant, largely due to increasing proportions of free fatty acids. The best edible oils contain only 0-3 to 0"5 per cent, of free fatty acids, and any- thing much exceeding this renders it unpalatable Olive Oil 199 Fio. 18.— Fruiting branch (J natural size) and fruits (natural size) of Olea europsea (olive tree). 200 Fatty Foods and unsuitable for use as a salad oil. The acidity is sometimes removed by treatment with alkali; and oils treated in such a manner will be characterised by very low acidity and the absence of the absorption bands of chlorophyll which are often clearly given by the fresh oil. Special Tests. — The usual constants of the oil, though fairly characteristic, are not sufficiently re- moved from those of many other liquid oils to enable adulteration to be immediately detected. This will be emphasised by the example on p. 83, and for this reason the testing of olive oils becomes more of the nature of special separate tests for individual adul- terants than any sweeping search for sophistication in general. Halphen's, Baudouin's and Bellier's tests, followed, if necessary, by Renard's test, should be applied. Lard oil is sometimes used as an adulterant, and charac- teristic crystals may usually be obtained from ether at 0° C. A green colour is sometimes produced by the addi- tion of copper, which metal should therefore be tested for in green oils. Uses op Oil. — As a salad oil, as a packing media for sardines, and in the country of origin for frying purposes. Olive Oil. 201 Olive Oil. Determination. Usual limits. Typical specimen. Melting-point, ° C, incipient fusion ' _ _. „ '„ „ complete fusion — Solidifying point, " C. 0" to r 2° j Saponification value 190 to 195 190-8 Refractive index at 40' C. (Zeiss butyro- 55 to 56 55-3 refractometer) Iodine value (Wijs) Specific gravity H C 80 to 87 85 0-915 to 0-918 0-9165 000 p Free fatty acids (as oleic) 0-3 to 1-0 (edible) up to 200% in non-edible oil. 0-45 Unsaponifiable matter 0-5 to 2-0 % 0-8 % Melting-point of fatty acids, ° C. 22° to 27° 202 Fatty Foods Arachis Oil. Trade Names. — Nut oil, pea-nut oil, ground-nut oil, one of the best qualities being known as Rufisque oil (from Rufisque nuts). Source. — Arachis hypogoea. Natural order, Legum- inosese. The origin of the plant is somewhat doubt- ful, but in all probability it is indigenous to America (Brazil). It is now universally cultivated in tropical and sub-tropical climates ; according to Lewkowitsch, Senegal produces the greatest quantity of fruits, and only those which are imported in the un decorticated condition are suitable for the production of the best edible oil, as the seeds alone undergo decomposition in transit, and are only fi^t for making " soap oils." The bulk of the seeds and fruits are imported to Marseilles. The development of the fruit in this plant is some- what peculiar. After flowering the young pods bend down and are forced underground to ripen ; hence the common name of " ground " or " earth nut." The fruit itself consists of a crinkled pod, or legume (erro- neously called a " nut "), enclosing two or more seeds, which are covered with a membraneous brownish-red skin (Fig. 19). Weight of 100 seeds or " kernels " = 61 grm. Content of Oil. — In whole seed, 43—48 per cent. Description op Oil. — Appearance. — Golden-yellow ; the edible oil is usually bleached to a very pale colour. Oonsifitency. — Liquid at ordinary temperatures. A small deposit of " stearine " is formed at low temperatures. Arachis Oil 203 Seed FiQ. 19. — Plant {\ natural size), fruit (natural size), and seed of ArachA^ ht/pogoea. 204 Fatty Foods Taste and smell. — Pleasant and nutty. Special Tests. — Bellier's test (p. 30) ; Renard's test (p. 34). Usual Addlteuants. — Cotton-seed, poppy, rape, maize and sesame oils. Liability to Rancidity. — Slight, and keeps well in the absence of meal, particularly that derived from the brown skin. Uses op Oil. — Every conceivable edible purpose to which a liquid oil can be put, more particularly as a salad oil. Uses op Non-Fatty Poetion. — The cake is very extensively used for cattle feeding, as it contains a large proportion of digestible protein. Analysis of Ground Nut Cake [Decorticated). Moisture .... 10-68 Oil . 5-81 Albuminoids* . 45-12 Digestible carbohydrates 80-49 Woody fibre 3-84 Mineral matter (aBh)t . 4-06 100-00 * Containing nitrogen, 722 per cent, t Including sand, 023 per cent. Arachis Oil 206 Arachis Oil. Determination, TJanal limits. Tjrpioal specimen. Melting-pointj ° C, incipient fusion „ „ „ complete fusion — — Solidifying point, ° C. 0° to 5° — Saponification value 190 to 196 194-8 Refractive index at 40'' C. (Zeiss 55 to 57-5 55-5 butyro-refractometer) Iodine value (Wijs) 87 to 98 91-23 Specific gravity, -1-1° C. 0-916 to 0-918 0-9170 9.8'= O, >i )y 15 ^' ■ Free fatty acids (as oleic) Varies from 0-5 % to 20%, rarely exceeds 10 % 5-24% Unsaponifiable matter . 0-4 to 0-8 % — Melting-point of fatty acids, ° C. . 27° to 30° ~ 206 Fatty Foods THE ALMOND OIL GROUP. This group of oils includes the true almond oil, expressed from the seed of the bitter almond, Prunus amygdalus (var. aTnara), which is found growing in the same localities as the sweet almond, P. amygdalus (var. dulcis), viz. Morocco, the Canary Islands, Portugal, Spain, France, Sicily, Syria and Persia, and also a number of very similar oils derived from the kernels (seeds) of such fruits as peach, plum and apricot. The fruits are drupes, and in the case of the almond it consists of : (1) A green, velvety epicarp. (2) A somewhat firm mesocarp. (3) The seed or almond, covered with a cinnamon- brown testa. The epicarp and mesocarp fall off when the fruit is ripe, and have always been removed before arrival on the market. The proportion of shell to kernel varies considerably in different species. Though the sweet variety contains more oil than the bitter, the latter is usually employed to obtain oil, as the residual cake is of more value. The oils of this group are remarkably similar in all their properties and their detection, in the presence of one another, may be taken to be almost a practical im- possibility. This is all the more unfortunate, as true almond oil is an expensive oil, whereas the others are remarkably cheap. Of the many tests which have been described for the purpose of achieving the impossible, probably Bieber's test is the least unsatisfactory, and is recommended by Lewkowitsch, who has paid much attention to this Almond Oil Group 207 subject (Analyst, 1904, vol. xxix, p. 106). The test is carried out as follows : Reagent. — Equal parts by weight of concentrated sulphuric acid, fuming nitric acid and water. It must be freshly prepared. Five c.c. of the oil and 1 c.c. of the reagent are mixed and thoroughly shaken and the appearance of a colour change noted. The following reactions are given : Almond oil . . No change. Apricot kernel oil . Pink colour. Peach kernel oil . A faint pink on standing. This test serves to differentiate the oils one from another, and 50 per cent, of apricot kernel oil may be detected in almond oil, but any smaller proportion than this is uncertain. The table (p. 208), taken from the above-mentioned paper by Lewkowitsch, gives the various values which he has obtained for some of these oils, and shows very clearly the great difficulty of detecting the admixture of one with the other. The most distinctive difference is the higher iodine value given by the kernel oils in contra-distinction to almond oil, but plum kernel oil (infra) has practically the same iodine value as almond oil. Plum Kernel Oil. Specific gravity. Free fatty acids. Saponification value. Solidifying point. Iodine value. At 15° C. 0-9127 to 0-9195 0-3 per cent, (circa) 191-5 °c. -5 to -9 100-00 208 Fatty Foods 4 s V m Colourless. Colourless at first, then pink. Pink colouration. Slightly pink. Very slightly pink. ad OOOOOQO oo o (M oq M cq (N Jq oq (N oq (n lU o 00 ■* 00 00 00 rH 00 Op OO OOOS0309003 C30b Oi Nr-trHiHrHiHTH rHrH r-f Free fatty acids.* rH CD Oi CO CD »0 iH CD r^ ,H CO ■* m « ■* CO US rt ■# CD Cq.^OiHiHOrH rH.^ O Butyro- refracto- meter at 40° C. loipoujopio Op O ^^*-(bt^^t* dot* OD LOIO10U3U310LO kOlO lO Iodine value. 99-4 103-6 100-3 102-5 96-65 104-2 95-24 107-4 107-9 108-7 ■as. Ill 5- iH m Qo 6i -^ At (N 00 o OOSOOOOOOOiOS Oi Gi OS OqrHiHiHi-Hr-liH rHiH t-H 11 11 0-91995 0-9182 0-9178 0-9180 0-91907 0-9183 0-9198 0-9200 0-9172 0-92026 ■s Almond oils expressed from .- 1. Valencia sweets 2. Blanched Valencia sweets 3. Sicily sweets . 4. Mazagan bitters 5. Small Indian almonds . 6. Mogador bitters 7. Peach kernel oil 8. Apricot kernel oil . 9. Apricot kernel oil from Mogador kernels 10. Califomian apricot kernel Rape Oil Group 209 RAPE OIL GROUP. This group consists of the following oils ; all of which are derived from plants of the natural order Cruciferse. (1) Rape Oil. — Obtained from the seeds of Brassica campestris, sub-species napus. (2) RuBSBN Oil. — Distinguished in France as being the oil obtained from the seeds of Brassica campestris, var. rapa. (3) Jamba Oil. — The oil obtained from Bast Indian uncultivated seed. English rape oil is obtained from the cultivated seed. (4) Ravison Oil. — The variety obtained from seeds produced in the vicinity of the Black Sea. 14 210 Fatty Foods Rape Oil. Teade Namk. — Colza oil. Source. — True rape oil is obtained from the seeds of Brassica caTnpestris (sub-species napus, etc.), which is extensively cultivated in Buropd and East India, large quantities being imported from Calcutta, Ferozepore, Guzerat, etc. The seeds, which are very small, are round, and vary in colour according to the sub-species from which they are derived, some being yellow, others brown or black. Weight of 100 seeds = 0-37 grm. Content of Oil. — in whole seed : European seed. Guzerat. Ferozepore. 39—45 per cent. 44 per cent. 40 per cent. Dbsceiption of Oil. — Appearance. — Though the crude oil is a dark brown and may contain rather a large proportion of free fatty acids, the refined oil is a pale yellow. Consistency. — Liquid at ordinary temperatures and depositing " stearine." The oil is very viscous. Taate. — Unpleasant. Smell. — Characteristic, which is more pronounced in the "extracted" oil. Possible Adulterants. — Ravison, jamba, linseed, cotton-seed, hemp-seed and particularly mineral oils. Special Tests. — The lowest saponification value of all edible oils (not containing a marked quantity of un- saponifiable matter). Viscosity very pronounced, and may be used to detect adulteration by comparing with pure rape oil, or, according to Archbutt, with glj'cerine Rape Oil 211 of sp. gr. 1"226 at 15"5° 0. — as castor oil alone has a higher viscosity. The hexabromide test is characteristic in mixtures of rape oil, with liquid non-drying vegetable oils, and lin- seed and fish oils may be detected in rape oil by the same test. For the lead-salt-ether method as applied to the detection of rape oil in admixture with other fats see p. 38. Uses oj? Oil. — The refined oil (usually cold drawn) is sometimes used for edible purposes, but inferior qualities find an outlet as lubricants and illuminants. Uses op Non-Fatty Poetion. — Eape-seed cake is used as a cattle food. Gr. Jorgensen, (Landw. Vers-Stat., 1910, Ixxii, pp. 1—14; Journ. Soc. Chem. Ind., 1910, vol. xxix, p. 512, abs.) emphasises the need of a laboratory examination of this cake, as mustard-seed often becomes mixed with rape seed. The harmlessness or toxicity of the cake may be ascertained by treating the sample with white mustard and water, when if more than certain quan- tities of mustard oil and thiosinamine are developed in a given time the cake would be poisonous. For details, however, the original paper must be consulted. Samples of cake yielding mustard oil have a high percentage of nitrogen and a stronger smell than normal samples. A rough test to distinguish between those likely to prove injurious to cattle and those which would be harmless, is to treat 1 grm. of the sample with water and a small quantity of white mustard and to allow the mixture to stand in a well-closed tube for one hour ; in the former case the odour is much more pungent and penetrating than in the latter. 212 Fatty Foods Ravison Oil. SouucE. — The seeds of a wild variety of Brassica campesiris from the Black Sea district. Content op Oil. — In whole seed, about 35 per cent. Desceiption op Oil. — Appearance. — Amber yellow in colour, often lighter than rape oil. Govsistency. — Liquid at ordinary temperatures, and somewhat less viscous than rape oil. Taste and smell. — Similar to rape oil. Special Tests. — The oil is very similar to rape oil, for which it forms an adulterant. Its iodine value is, however, higher. Uses op Oil. — Practically those of rape oil, but it is not so suitable for lubricating purposes. Uses of Non-Fatty Poetion. — Ravison cake is used as a cattle food. Analyais of Ravison Cake. Moisture . . . 9-04 Oil . . - . . . 10-50 Albuminoids . 34"37 Digestible carbohydrates . 27'85 Woody fibre . . . 12-07 Mineral matter . . .6-17 100-00 Jamba Oil 213 Jamba Oil. SouKCE. — The seeds of an uncultivated variety of East Indian rape. Content of Oil. — Usually somewhat less than from cultivated rape seed. Description of Oil. — Appearance.—VaAe yellow. Consistency. — Liquid at ordinary temperatares. Taste and smell. — Very strong and pronounced, as well as distinctly unpleasant. Special Tests. — As rape oil. Uses op Oil. — No use of the oil as an edible oil is known to the authors, but there is no reason why it should not be so used, if sufficiently refined. The following figures give the usual limits for rape and ravison oils. Bape. Ravison. Solidifying point, ° C. -10° to -12 -8° (Jean) Saponification value .... 171 to 177 175 to 179 Refractive index at 40° C. (Zeiss butyro- 59 to 60 — refractometer) Iodine value . 97 to 105 109 to 122 (Archbutt) Specific gravity at 15° C. 0-9135 to 0-918 to 0-922 0-9160 (Arohbutt) Free fatty acids (as oleic ■> Up to 2 % Up to 5 % Unsaponifiable matter 1-4. to 3-0 % (Le-w- ko-witsch) 1-5 to 1-75 % Melting-point of fatty acids, ° C. 18-5° to 21-0° (Arohbutt) Optical activity (ang. degrees in 200 -0-8 to -0-5 About -0-5 mm.) The. figures for jamba oil follow those of rape oil very closely. Cotton-seed Oil Group 215 COTTON-SEED OIL GROUP. 1. OOTTON-SBED OlL. 2. Kapok Oil. 3. Sesame Oil. 4. SojA Oil. 5. Maize Oil. 216 Fatty Foods Cotton -seed Oil. Trade Names. — The pure oil is usually sold in bulk under its own name, but smaller consignments under such names as "butter-oil," etc. Products consisting mainly of the oil, are sold as " Cottolene," etc. Source. — The seeds of various species of G-ossypium, the cotton-plant. The cotton fibre, with which the small black seeds are covered, having been removed by suitable machinery, without separation of the husk, the remaining seed may or may not be separated from the ^usk before pressing, thus forming, after pressing, the 1 " decorticated " or " undecorticated " cake. The removal of cotton^bre is more difficult in the case of so^e seeds than others. It follows, therefore, that fibre niay appear in larger or smaller quantities in some undecorticated cake. The husk constitutes about 40 per cent, of the whole seed. Weight of 100 seeds = 7 grm. (including a small amount of fibre which adheres to seed). Content op Oil. — (a) In oleaginous portion, about 37 per cent. ; (&) in whole seed, about 20 per cent. DESORiP'i'ioN OP Oil. — Appearance. — The crude oil is very dark in colour, but it is usually partially or com- pletely refined at the mill by a simple removal of either a proportion or all of the free fattj' acids by treatment with caustic soda, which at the same time removes the bulk of the colouring matter in the form of an almost purple-black dye, the resulting oil being then of a golden-yellow hue. The colour can be still further improved by bleaching. Certain tintometer standards are used in America. Cotton-seed Oil 217 Consistency. — A liquid oil which deposits on standing, particularly at low temperatures, a solid fat known as cotton-seed " stearine," for the separation of which the oil is allowed to stand in tanks, and the clear super- natant liquid having been syphoned off, the " stearine " is obtained from the deposit by filter-pressing. As more " stearine " naturally forms during the winter months, the terms "winter" and "summer" oils hare become used commercially as a distinction. The " winter " oil is more particularly used for salad purposes owing to the fact that it is less liable to pro- duce a deposit when bottled, and as the " stearine " does not keep so well as the liquid portion, the " winter " oil is preferred for margarine and other edible purposes, though in spite of this fact the " stearine " itself is actually refined and finds a use in edible products. Taste and .fw.ell. — Strong and distinctive ; they can seldom be so removed that they cannot be detected by the skilled palate and nose. Special Tests. — Halphen's reaction is characteristic (kapok oil gives a similar colour) . Note that the reaction is destroyed by heating the oil to a high temperature, though practically all the commercially refined oils give the reaction, with the exception of those treated with powerful reducing agents. (See " Hydrogenised " Oils, Appendix C.) The fat of animals fed on cotton-seed cake may give the reaction, but always to a limited extent. In the case of heated oils, see nitric acid test (p. 101). The refractive index is on the low side for liquid vegetable oils, and the titer test is exceptionally high (33 to 36). Uses of Oil. — The whole oil is principally used as a salad oil, and also very largely in margarines. The " stearine " is employed for lard substitutes to a limited extent. The oil also finds a large use for 218 Fatty Foods cooking and' frying and in cakes and biscuits. Its popularity is largely due to its cheapness, though, how- ever well refined, it is always inclined to develop its peculiar taste and smell sooner or later. Uses of Non-Fatty Portion. — ^Apart from the well- known use of the cotton fibre, the oil cake, obtained after pressing the seed, forms one of the most popular cakes for cattle feeding. Typical Analyses of Cotton-seed Oil Cake. Decorticated. XJnde- Chinese. American, corticated. Moisture 10-52 . 8-65 . 11-32 Oil . 6-29 . 7-93 . 616 Albuminoids* 43-50 . 40-25 . 23-75 Digestible carbohydrates ; 22-79 2606 . 31-33 Woody fibre 10-67 . 10-] 6 . 21-80 Mineral matter (ash)t . 6-23 . 6-95 . 5-64 100-00 100-00 100-00 *Containing of nitrogen 6-96 6-45 . 3-80 t „ „ sand 0-12 . 0-23 . 0-42 Cotton-seed Oil 219 OOTTON-SEED OiL. Determination. TTsuaV limits. Typical specimen.* Cotton "stearine." Melting - point, ° C, incipient _. _ _ fusion Melting - point, " C, complete — — — fusion Solidifying point, ° C. . 3° to 5° 3° 19° , Saponification vahie . 192 to 195 192-6 196-2 Eefi;aotive index at 40° C. (Zeiss 58 to 59 58-4 -^ 1 butyro-refraotometer) Iodine value (Wijs)t 105 to 115 112 91-2 Specific gravity ii° C. . 0-922 to 0-925 0-9228 0-919 9£0 Q „ W . 1 5 ^" Free fatty acids (as. oleic) Varies 0-25 % 0-08% TJnsaponrSable matter 0-8 to 1-8% 0-85 % — Melting-point of fatty acids, ° C. 35 to 38 — * Eefined edible oil. t Winter oil should not have an iodine value lower than 110 per cent. 220 Fatty Foods Kapok Oil. Trade Name.— Bastard cotton oil. Source. — Seeds of Eriodendron anfractuosum and Bombax malnharicum. Both trees are widely distri- buted in the tropics, and E. anfractuosum (which seems to be the chief source of kapok oil) is found in great abundance in Java, West Indies, Africa, etc. In 1 Fig. 20. — 1. Half open pod of Eriodendron anfractuosum. Prom • Kordofan (| natural size). 2. Seeds (natural size). some parts the tree is known as the " silk cotton tree," owing to the fact that the seeds, which are enclosed in an oval brown pod, are entirely covered with a cottony growth from the inner wall of the pod The seeds themselves, which are quite free from hairs (unlike cotton), are small, round, and black in colour, the hard shell constituting about 40 per cent, of the whole.^ The imports to Holland and America are steadily increasing. Weight of 100 seeds = 6-0 grm. Lewkowitsch. Kapok Oil 221 CoN'j'isNT OF Oil. — On whole seed : Ceylon seed, 22 per cent. ; Calcutta seed, 24—25 per cent. Desckiption of Oil. — Appearance. — Darkish, yellow. Consistency. — Liquid at ordinary temperatures. Taste and smell. — Not unpleasant, and resembling cotton-seed oil in these respects. Spbciai, Thsts. — Halphen's reaction positive, but the oil is distinguished from cotton oil by means of Becchi's test (modified) . It is also possible to detect 1 per cent, of kapok oil in other liquid oils by this test (see p. 54). Uses of Oil. — As cotton-seed oil, and probably large quantities are either sold as cotton-seed oil or in admixture with it. It is used as an edible oil in Holland. Analysis of Seeds and Cahe.^ Seeds. Cake. Moisture 10-85 13-80 Oil 22-01 7-47 Albuminoids 22-87 26-25 Digestible carbohydrates 18-65 23-19 Woody fibre 20-02 23-19 Mineral matter • 5-60 6-10 100-00 100-00 ' Smetham (abs. Analyst, 1910, vol. xxxv, p. 58). 222 Fatty Foods Kapok Oil. Determination. Usual limits. Typical speci- men. Melting point, " C, incipient fusion ' „ „ „ complete fusion — — Solidifying point . — — Saponification value 190 to 197 192-5 Refractive index at 40" C. (Zeiss buty ro- refractometer) 53 to 57 56-2 Iodine value (Wijs) 95 to 110 97-54 Specific gravity, ;j° C Free fatty acids (as oleic) 0-921 to 0-923 0-9217 Varies 7-54% TJnsaponifiable matter . — Melting-point of fatty acids, ° C. 29° to 36° — Optical rotation (angular degrees in — -005 200 mm.) Sesame Oil 223 Sesam^ Oil. Trade Names. — Sesame oil, Gingelly oil. Native Names. — In India the white seed is known as "suffet-til" and the black as "til" or "tillie"; other names are " beni," " benne," etc. SbuKCE. — The seeds of Sesamum indicum (and other varieties of the group Pedalinese), an annual plant grown in many warm countries, such as India, China, Asia Minor, Japan, etc. In India alone (excluding Eastern Bengal and Assam) the yield was estimated at 421,900 tons, from an area of about 4f million acres, during 1910-11. The fruit consists of a capsule about 1^- in. long con- taining numerous small seeds. These seeds are flattened and oval with a point at one end; they are either black or white in colour, but sometimes brown or grey. The best quality oil is obtained from the white seeds, though the black variety yields a somewhat larger quantity. Trade rules require that " white seeds " shall not contain more than 15 per cent, of black seeds and various other rules are also recognised. The Levant seeds yield the best flavoured oil. Weight of 100 seeds = 0-28 grm. Content of Oil. — In whole seed, 48—54 per cent.; occasional samples, 58 per cent. Desceiption of Oil. — Appearance. — Pale yellow in colour. Consistency. — Liquid at ordinary temperatures; merely a trace of " stearine " is deposited on standing. Taste. — Not unpleasant, rather suggestive of sea- weed. 224 Fatty Foods Smell. — Both taste and smell can be completely removed and the refined oil keeps well. Optical Activity. — Dextro-rotatory. Possible Adulterants. — Governed by market price. Usually rape and cotton-seed, and, more rarely, arachis oil. Special Tests. — The Baudouin test is distinctive, and the reaction is not given by any other commonly occurring oil. The cause of the reaction is probably not completely removable technically, but appears in the liquid fatty acids when separated. TJnsaponifiable matter contains phytosterol sesamin and a viscid brown oil. Uses op Oil. — The oil is used for similar purposes to cotton-seed oil, but not to such a great extent in this country as on the continent, where (in certain countries) it is required by law to form a constituent of margarine, in order to facilitate its detection in butter. It is also used as an adulterant of olive oil. The oil is very greatly used in India for various purposes, such as in cookery, as an unguent, for soaps, etc., and also as an adulterant of ghee. Uses of Non-Fatty Poetion. — Sesame cake is largely used for cattle feeding, more especially on the continent. It is hard and heavy, aud though in many respects it resembles linseed cake, it is lighter in colour. ' The nutritive value is approximately equal to that of decorticated cotton cake. In feeding trials carried out in this country sesame cake seemed, if anything, to give slightly better results than cotton cake, though the difference is practically negligible. The proportion of fat in the milk was identical in both cases.^ It is note- worthy that in recent experiments carried out with sesame cake at the Agricultural College at Wye, the > Bull. Imp. Inst., 1909, iii, p. 309. Sesame Oil 226 butter from the cows so fed, did not give the Baudouin reaction. In some cases the whole white seeds are used as food, and in Turkey a sweetmeat called " Halvas " is made by grinding them with sugar and then cooking. Analysis of Sesame Cake. Moisture Oil . Albuminoids* - Digestible carbohydrates Woody fibre Mineral matter (ash) 8-19 8-60 38-70 25-22 5-17 14-12 100-00 *. Containing of nitrogen, 6-2 per cent. Sesame Oil, Determination. Usual limits. Typicftl specimen. Melting point, ° C, incipient fusion _ „ „ „ complete fusion — — Solidifying pointj ° C. - 3° to - 4° — Saponification value .... 188 to 193 191-7 EefractiTC index at 40° C. (Zeiss butyro- 59 to 60-5 59-5 refractometer) . Iodine value (Wijs) 103 to 110 104-2 Specific gravity, 4-f C. 1.90 p 0-922 to 0-924 0-9235 Free fatty acids (as oleic) "Varies 2-61 % Unsaponlfiable matter 0-8 to 1-2 % 1-02% Melting-point of fatty acids, ° C. . 25° to 31'- — Optical activity (angular degrees in + 1 to + 3 + 1-35 200'mm.) 15 226 Fatty Foods Soja Oil. Trade Names. — Bean oil; Soya oil. SouECE. — The seeds of Glycine hispida and numerous varieties of this plant. Natural order^ Leguminoseae. Extensively cultivated in China, Japan, Manchuria and elsewhere. In the latter country the production is rapidly increasing in order to cope with the growing European demand. Practically the first shipment of the beans arrived in Hull as recently as 1908, and since then enormous quantities have been imported. During the year 1910 the total imports into this country amounted to 421,531 tons, China and Japan furnishing 174,415 and 149,958 tons respectively, Kussia and other countries the remainder. In 1911, owing to increased local demand and short crops, the total imports fell to 222,657 tons. From China during the years 1910-1911 the total exports amounted to 859,407 and 650,324 tons respectively. The whole fruit consists of a hairy pod, containing the small, round, pale yellow seeds about the size of the ordinary pea, but, if anything, rather smaller. The yellow seeds are usually employed, but green and black varieties exist, and are often found mixed with the former. Weight of 100 seeds = 8-5 grm. Content of Oil. — Whole seed or bean, 16-19 per cent, (commercial yield 13 per cent.) Description of Oil. — Appearance. — Deep brown, not much improved by alkali refining except when bleached. Consistency. — Liquid at ordinary temperatures, and not depositing " stearin© " on standing. Soja Oil 227 Pia. 21. — 1. Pod of Glycine hispida (Soja bean). 2. Prtiiting branch (i natural size). 3 Seeds (natural size). 228 Fatty Foods Taste and smell. — Very slight and not unpleasant. They are difficult to remove, and the keeping properties of the oil are not so good as in the case of many liquid oils, as it shows a tendency to develop a nauseous taste. Possible Adultekants. — Cotton-seed oil is the only likely adulterant on account of the cheapness of the oil. Spkcjal Tests.— None. Iodine value and refractive index high. Uses of Oil. — It finds a use as an edible oil, but to a limited extent for the reason given above. It will be found in feeding cake and in biscuits made from soja meal. According to Korentschewski and Zimmer- mann (Chem. Zeit., 1905, xxix, pp. 777—778), practical tests showed that the oil was readily absorbed by the system and possessed a high food value. Uses op Non-Fatty Portion. — An excellent feeding cake extensively used on account of its high nutritive value, as it contains about 40 per cent, pf protein (stated to be closely allied to casein) ,of very digestible nature. When first introduced in large quantities about three years ago for cattle-feeding, some sickness from over- feeding was caused before its high nutritive and diges- tible value was realised. Soja flour is usually made from extracted meals, and is often adulterated with wheat, though guaranteed pure. Biscuits are made from soja meal. In its native country soja cheese and other products are made from the whole seed, both the black and white varieties being utilised in China and Japan in various ways for edible purposes. Soja Oil 229 Analyses. Soja bean Moisture . 1! -00 Oil . . ,17-03 Albuminoids* • _ . . 3811 Dig^^iiiblp parbohyirafces . 24-24 Woody fibre 4-82 Mineral mattpr ■ ,., 4-80 lOO'OO *Nitrogen per cent, on sample 6' 10 Soja cake. 6-01 9-63 42-00 31-8,6 5-11 5-39 100-00 G-72 Soja Oil. : Determination. Usual limits. Typical speci- men. Melting point, °' C. incipient fusion „ „ „ complete fusion — — Solidifying point, ° C. -7° to -12° -11° Saponification value . • .' 190 to 193 191-8 Refractive i-ndex.at 40" C. (Zeiss butyro- 62-5 to 63-5 62-9 , refractpmeter) Iodine' value (Wijs) 126 to 135 132 5 Specific gravity, t°" CI. Freo-Jprtty acids (as oleic) - . . 0-924 to 0-926 0-9251 Seldom 0-25% exceeds 3 % TJnsaponiflable, matter . . — 0-32% -Melting-point of ffttty acidSj ° C. 26 to 29 — 1 230 Fatty Foods Maize Oil. Trade Name. — Corn oil. SouECE. — The fruits (caryopses) of Zea Mays (Indian corn), a large number of which are united with a fleshy stalk to form the " cob." The fruit is commercially divided into three parts : (1) Bran, tp. (testa + pericarp). (2) Hominy, e. (endosperm). (3) Germ, s. (scutellum). See diagram. Oil is obtained from the germ alone which is separated from the endosperm by various processes (washing or sifting) before pressing. It is obtained as a bye- product in starch-making. A very acid type of oil is obtained from alcoholic fermentation vats, but this is not now often met with and cannot be used for edible purposes. Weight of 100 fruits = 24 grm. Content op Oil. — (a) Whole fruit, 3—5^ per cent. ; (fe) oleaginous portion (commercially separated germ), 30-35 per cent. Desckiption of Oil. — Appearance. — Golden-yellow colour. Consistency. — Liquid, giving a small deposit of " stearine " in cold weather. Taste and smell. — A very strong tenacious taste and smell, characteristic of the original fruit ; both are removed with difBculty. Liability to Rancidity. — Keeping qualities are fairly good when once refined, but the crude oil is rapidly hydrolised, if meal be present. Possible Adulterants. — Only cotton-seed oil, when price permits. Maize Oil 231 Special Tests. — Iodine value and refractive index very high. Low solidifying point, "both of oil and fatty acids. It is stated to give a small but marked Eeichert- -tp a Fia. 22.— Fruit of Zea Mays. (Longitudinal section x 15.) t^. Pericarp + testa (bran), e. Endosperm (hominy), s. Scutellum (germ). 6. Embryo, a. Aleurone layer. I. Epithelial layer. Meissl value, but the authors^ experience does not confirm this (see figures). Unsaponifiable matter usually over I'O per cent. Certain varieties of the oil are characterised by a cloudi- ness, consisting of microscopic needle crystals, so fine as to be difficult of removal by filtration. 232 Fatty Foods The melting-point of the phytosteryl acetate is higher than usual (127°-137°0.). Titer test (16''-18°C.) lower than other oils of this group. Uses op Oil. — In America the oil is extensively used for salad purposes and in margarine. Some of the better refined qualities reach a high degree of excellence as regards taste, smell and keeping pro- perties, and therefore find a use for cake and biscuit making as well as for smearing bakers' pans. The amount of refined edible maize oil produced in this country is more limited, and the quality does not usually reach the excellence of the American product. Uses op Non-Fattt Poetion. — The uses of the de- germinated portion (hominy — i. e. starchy portion and bran) are too well known for comment. A variety of products, such as flaked maize, etc., are also made. The pressed germ forms the well-known maize cake or meal, extensively used for cattle and poultry feed- ing, and is a favourable vehicle for the administration of molasses, as it supplies a high protein content, together with starch and oil. The cake is most palatable. Analysis of Maize Germ Cake. Moisture ..... 9'47 Oil* 12-23 Albuminoidst . . 20-3() Digestible carbohydrates . . 46 'IS Woody fibre . . . .9-38 Mineral matter J . . . .2-43 100-00 * Free fatty aoids (pei' cent, on the oil), 38. t Nitrogen (per oant. on the sample), 3'23. J Sand, 040 per cent. Maize Oil 233 Maize Oil. Determination. Melting-point ° C, incipient fusion, „ „ „ complete fusion Solidifying point, ° C. . . . Saponification value .... Refractive index at 40° C. (Zeiss butyro- refractometer) Iodine value (Wijs) Specific gravity, ii° C. » <• TT "-'• Free fatty acids (as oleic) Unsaponifiable matter . Melting-point of fatty acids, ° C. Special values : Eeichert-Meissl Polenske ...... Iodine value of liquid fatty acids t ■ * De Negri and Fabris. + Tortfelli and Euggeri. Linseed Oil Group 235 LINSEED OIL GROUP. 1. Linseed Oil. 2. Sunflower Oil. 3. PoppT Seed Oil. 4. Safflowee Oil. 5. NiQEE Seed Oil. 6. Candle Nut Oil. 7. Para Eubbeb Seed Oil. 8. Walnut Oil. 9. Perilla Oil. 236 Fatty Foods Linseed Oil. SouECE. — Seeds of Linuni usitatissimum (the flax plant). India^ America, Russia, Argentina, etc. The variety known as Baltic Sea seed produces the purest oil, principally because the seed itself is kept free from admixture with other seeds. Black Sea seed, on the other hand, is quite commonly mixed with small quantities of such seeds as sesame, hemp, rape, etc. Content op Oil. — In whole seed : The figure varies somewhat according to the origin of the seed, but is usually between 36 and 40 per cent. DsscEifTiON OF Oil. — Appearance. — Golden-yellow in colour. Consistency. — Liquid at ordinary temperatures. Taste and smell. — Strong and distinctive. Possible Adulterants. — Fish oils, cotton-seed oil, rosin oil and rape oil. Special Tests. — lodiiie value very high (should not fall below 170 per cent. — LewkowitsQli). Should not contain more than 2"0 per cent, unsaponi- fiable matter, or have a specific gravity below 0'931. Hexabromide test strongly positive. Fish oil may be detected by the smell on saponifica- tion and by the melting-point of the ether-insoluble bromides (see "Fish Oils," p. 274). Rosin by Liebermann-Storch reaction (p. 55). Riipe oil by Tortelli and Fortini's method (p. 38), and by reduction in saponification value (in absence of mineral oils). Traces of other oils in Black Sea oils may be due to admixture of other seeds before pressing. Uses of Oil. — The refined oil when practically Linseed Oil 237 tasteless and odourless finds a limited but somewhat increasing use as an edible oil, chiefly in margarine, but sometimes as a salad oil. Uses op Non-Patty Poktion. — This portion forms one of the most extensively, used cattle cakes, though the seed contains, a oyanogenetic glucoside, as well as aji, enzypae capa,ble of decomposing it. .This, 'hoyreyer, is, destroyed by the high temperature: to which the meal is subjected in the course of pressing. ■- •'■ -Analysis of Linseed Cake. " (English average. Smetham, Analyst, 1910, vol. XXXV, p. 56, abs.) -. Moisture . . . . 1M6 Oil . . . . . 9-50 Albuminoids . . . 29-50 Digestible carboljydrates . 35'54 Woody fibre , . ' . . . 9-10 Mineral matter .... 5'20 Linseed Oii . ] 00-00 Determination. Usual limits. { JMeHwgrpointi " C , incipiejnt fusion ' ' ■, — ■:■, ' • ■■ /' ■„,. „ ,..„ complete fusion. - 16' to - 20? (Glaessner) ' Solidifying poiiit, " C. '. ' '. -27°{C!liateaiiV ' Sapopifioation value .; . ., . . ... . 189.td 195 . , :. Refractive indejc ?it 4,0° C.X^^eiss Ijntyro- 73 to 75 ■ reff aotomet'ei-) ' ■ -' - ' ■• ibdinetvalu$,(Wijs) ' . 175 to .200 Specific gravity, \f C. . i i . : ■ , J 1 , ' ; . . oo. /-I 0-931 to 0-938 , '^xee\^t%y ajjids^ (as. oleic) ... Usually under 5%. :. tfnsaponifiaijle matter . . 0-8 to 2 % r , ,' Meliing-point oiF fatty alcida .' ■ . ' . ' , ; ' , ; J I Optical rotation (angular degrees in -0-1 200- mm.) , , i';,^-;- : ■-; j. \, :.- i, : • .. - 238 Fatty Foods Sunflower Oil. Source. — The fruits (achenes) of Helianthus annuus (Common Sunflower). The plant is indigenous to Mexico, but is now extensively cultivated in South Russia, China, Hungary, South Africa, etc., in fact, in the last-named country it is so plentiful that it is used to mark out the boundaries of the fields. It is an ex- tremely easy plant to grow and the production of fruits e -- - t Fia. 23. — Fruits of Helianthus annuus (Common Sunflower, x 3). 1. Whole fruit. 2. Longitudinal section, p. Pericarp, t. Testa (or seed coat), e. Endosperm. is enormous, but attempts to introduce it into India and the United States have not proved entirely satisfactory. The production of Rhodesian seeds is increasing, while quite recently Spain has been suggested as a most suit- able country for their cultivation. In Great Britain, though the climate is eminently suitable for the growth of the sunflower plant, the crop is extremely small, chiefly due to the fact that the value of the crop is but little known and much prejudice has to be over- come. The fruits (as shown in sketch) vary in colour from Sunflower Oil 239 white to brown or black, and are about 1 cm. in length. The smooth shell (or pericarp) which encloses the seed constitutes some 50—55 per cent, of the whole fruit, the remaining 45—50 per cent, being made up of the olea- ginous seed. Weight of 100 fruits = 6|- grm. Content of Oil. — (a) whole fruit 22—25 per cent. ; (6) seed (t. e. oleaginous portion) 45—50 per cent. The Hungarian seeds contain the most oil, often over 50 per cent. Desceiption of Oil. — Appearance. — A clear pale yellow, closely resembling olive oil in many respects. Consistency. — Liquid at low temperatures. Taste and smell. — Pleasant, but not very pro- nounced. Special Tests. — None. Uses op Oil. — The oil is used very largely on the continent as a salad and margarine oil, but in this country the use is limited. Uses of Non-Patty Portion. — The non-fatty portion is of special value. The cake is largely used abroad as a feeding stuff owing to its high nutritive value. In the Kuban district alone nearly 40,000 tons of oilcake were produced in the year 1911, 90 per cent, of which was shipped to Denmark. Apart from the seeds, the stalks and leaves are of commercial use, the stalks being burnt to produce an ash rich in potash, which can be advantageously employed as a fertiliser. Some idea of the extent of this industry may be gathered from the fact that seven factories alone pro- duce from £2,000,000 to £3,000,000 worth of potash annually from sunflower stalks. The stalks also contain a tough fibre. Smetham (Analyst, 1910, vol. xxxv, p. 59, abs.) gives the following analyses of sunilower cake ; 240 3 Fatty Foods Whole fruit. Husked Moisture ., , . 7-10 7-75 Oil . 7-43 10-03 Albuminoids 19-01 37-00 Digestible carbohydrates . 28-93 21-14 Woody fibre . 30-03 16-53 Mineral matter 7-50 7-55 100-00 10000 SoNPLowiiB Oil. Determination. Melting-point, ° C, incipient fusion TTsual limits.* Typical speci- men.* „ „ complete fusion — — Solidifying point, ° C — -16 Saponification value .... 190 to 194 191-2 Befractive index at40''C. (Zeiss butyro- 60 to 63 62-7 refraotometer) Iodine value (Wijs) 125 to 140t 133-4 Specific;grsivity, i|° C. 0-924 to 0-926 0-9253 „ ■„ H" C — — Free fatty acids {as' oleic) 10 to 5 % 1-75% TJnsaponifiable matter .... 0-3 to 0-8 % Melting' point of fatty acids, ° C. . 20^ to 23° 22° Optical rotation (angular degrees in — 0-0 200 mm.) * From- black seeds. t A sample of oil extracted from white seeds by the authors gave an iodine- value of 106 and a refractive index of 58'. Poppy Seed Oil 241 Poppy Seed Oil. Source. — Seeds of Papaver sonmiferum (tlie Opium Poppy) and other varieties. The seeds^ which vary in colour from white to brown or bluish black, are con- tained in the well-known poppy capsule, which is shown both whole and in section in the accompanying photograph. The plant is largely grown in France, Russia, India, Asia Minor, Persia, Egypt, etc. The oil obtained from the different varieties of poppy seed have been divided for commercial purposes into two classes : (1) " Huile d'oeillette," obtained from grey or blue indigenous European seed. (2) " Huile de pavot," furnished by white, brown or mottled seeds of foreign origin . " Huile d'ceillette ' ' is superior to " huile de pavot," and is of more value commercially. In a general way these two oils can be distinguished in the following manner : On violently shaking these oils with air in a bottle, it will be noticed that (1) gives a fine emulsion of air- bubbles, so that the oil is rendered turbid, and this is not the case with (2). The froth is also more persistent than with (2). In colour also there is a slight varia- tion : " huile d'ceillette " is a bright golden-yellow as compared with " huile de pavot," which is a pale straw colour, and often requires to be coloured for general use (L. Yuaflart, Annales des Falsifications, 1909, ii, pp. 276-278). Content of Oil. — In seed, 45—50 per cent. Accord- ing to Halphen the yield of oil varies from 36—40 per cent. 16 242 Fatty Foods Description of On,. — Ajjiearanci'. — Pale straw to golden-yellow, according to source (see above). Consistency. — Liquid at ordinary temperatures. Taste and smell. — Vary according to source (see above). /.lud nj', u>^ian «l. ,. ,.,,, ^ ft i. hiiiL- V..I fHLrTS Iff ihe OPIUM fttfiager ti'mni/tni POPPV «. L Fig. 24. According to Halphen the cold drawn oil has a mild and pleasant odour and does not easily develoii ran- cidity. The Russian oil, on the contrary, has a harsh disagreeable taste and its smell resembles that of linseed. Poppy Seed Oil 243 PossiBLis Adulteeants. — This oil becomes mixed with sesame oilj arachis oil, etc., not actually for the purpose of adulteration, but owing to the fact that they are often pressed in the same mill and batches of the one run through before the machinery is clear of the other. As poppy oil is usually a high priced oil, anything more than -traces of other oils must be regarded as adulteration. Special Tests. — (See " Walnut Oil.") Hexabromide testj no deposit. Iodine value and refractive index very high. Titer test low (16° 0.). Uses of Oil. — The finest oil is largely used for edible purposes, both mixed with olive oil and as a salad oil. Uses ofNon- Patty Poetioij. — The cake is used for cattle feeding and is rich in protein. For analysis of cake see p. 24 i. Poppy Seed Oil. Detsrmination. Usuul limits. Typical speci- men." Melting-point, ° C, incipient fusion „ „ „ complete fusion — Solidifying point, ° C. -1.5 to -18° — Saponification value 192 to 196 194-0 Refractive index at 40^ C. (Zeiss butyro 63-0 to 645 63-3 refraotometer) Iodine value (Wijs) 130 to 138 135-6 Specific gravity, ^° C. Free fatty acids (as oleic) 0-924 to 0-926 0-9247 Varies, but 0-82% seldom ex- ceeds 10 % tTnsaponifiable matter About 0-5 % — Melting-point of fatty acids, ° C. . 20° to 21° — Optical rotation (angular degrees in -0-02 to — 200 mm.) -0-25 244 Fatty Foods Analysis of Pofity Seed Cuke. Moisture 10 97 (Ml 8-63 Albuminoids* 35 90 Digestible carboliydrates 20 80 Woody fibre 9-95 Mineral matter 13-7.5 100-00 * Including nitrogen. o 75 per cent. Safflower Oil 245 Safflower Oil. Teade Name. — Carthamus oil. Source. — The seeds of Carthamus tincturitis, one of the Compositffi. It grows in China, India, Egypt and the Levant, but it is also cultivated in many European countries for the yellow dye obtained from the flowers (this last is often confused with saffron, which is, however, quite distinct).' The seeds (known as Kurdee seeds) have a hard, shining, yellowish-white husk, the decorticated seed being said to yield a better quality oil (Tylaikow). They are 6-8 mm. long and 4—5 mm. broad (Fendler;. Weight of 100 seeds = 5 grm. The kernels represent 54 per cent, of the whole seed (Fendler) . Content op Oil. — In seed, 25—30 per cent. Dbsceiption of Oil. — Appearance. — Golden yellow. Consistency. — Liquid at ordinary temperatures. Taste aiid smell. — Very slight. Liability to Rancidity. — Decided. Special Tests. — Gives the hexabromide reaction, and has distinct acetyl value. Uses of Oil. — Utilised in India for edible purposes. It serves a similar purpose as linseed oil, but its drying powers are not so great. The following analyses of the cake are due to Smetham (Analyst, 1910, vol. xxxv, p. 54, abs.). : Whole seed. Decorticated. Moisture 8-55 . n-60 Oil . ' . . 9-73 7-70 Albuminoids 20-25 . 47-88 Digestible carbohydrates . 25-12 19-72 Woody fibre . 32-95 6-20 Mineral matter 3-40 6-90 100-00 100-00 246 Fatty Foods Sapflowee Oil. Determination. Usual limits. Melting-point, " C, incipient fusion „ „ „ complete fusion Solidifying point, ° C. Saponification value Eefractive index at 40° C. (Zeiss butyro- refractometer) Iodine value (Wijs) Specific gravity, ^° C. BOO P Free fatty acids (as oleic) tTnsaponifiable matter . Melting-point, ° C, of fatty acids . - 5° (Fendler) Below - IS' 190 to 194 63 to 65 135 to 145 0-926 About 6 % (varies) 0-7 % (Fendler) 16° to 17° Special value — Acetyl value (true) 161 (Lewkowitsch) Niger Seed Oil 247 Niger Seed Oil. SouECE. — The achenes of Guizotia abyssinica (Natural order Composifcse), indigenous to many districts of tropical Africa, and cultivated in other parts of the world. The achenes are black and highly polished, about 4 mm. long and from 1-|- to 2 mm. broad. Weight of 100 fruits = 0-35 grm. Content op Oil. — On whole fruit, 40-50 per cent. The shell constitutes about 20 per cent, of the achene. Desceiption op Oil. — Appearance. — Yellow. Consistency. — Liquid at ordinary temperatures. Taste. — Pleasant and nutty. Smell. — Very slight. Si'ECrAL Tksts. — None, but is differentiated from rape oil, for which it is used as an adxilterant, by its higher iodine and saponiBcation values. Uses op Oil. — As an edible oil, but inferior varieties find a use in soap making. Uses op Non-Fatty Poetion. — The cake may be used for cattle feeding, and is very rich in albuminoids. The following analysis is due to Smetham (Analyst, 1910, vol. XXXV, p. 54, abs.). Moisture . . . 8-90 Oil . . . . 14-03 Albuminoids . . 34'06 Digestible carbohydrates . 21-80 Woody fibre .... 9-26 Mineral matter .... 11'95 10000 248 Fatty Foods NiGKE Seed Oil. Determination. Usual limits. Melting-point, ° C, incipient fusion „ „ „ complete fusion Solidifying point, ° C. Saponification value Refractive index at 40° C. (Zeiss butyro- refraotometer) Iodine value (Wijs) Specific gravity, -14° C. Free fatty acids (as oleic) Unsaponifiable matter Melting-point of fatty acids, ° C. -8=" to -10' 189 to 192 63 to 64-5 128 to 134 0-92o to 0-927 Varies from 2 % up- wards 26" to 28' Candle Nut Oil 249 Candle Nut Oil. Other Names. — Kekune oil. Country Walnut oil, etc. SouiiCE. — Seeds of Aleuritcs triloba, a large tree of Seeds Fio. 25. — Aleurites triloba (| natural size). 1. Fruit. 2, Section of fruit. 3. Seed (candle nut). 4. Section of seed. tropical countries, large quantities being exported from Hong Kong, Fiji, as well as Australia and New Zealand. The seeds are about the size and appearance of a 250 Fatty Foods walnut, except that they are a mottled brown and white colour. The kernel is enclosed in a somewhat tough hard shell (see sketches). Another oil derived from A. moluccana is also called "candle nut oil," and the two are often sold indiscriminately. Weight of 100 seeds = 900 grm. Content op Oil. — In kernel, 62-66 per cent. Descejption oe Oil. — Appearance. — A pale yellow, almost white, limpid oil. Consistenr.i/. — Liquid at ordinary temperatures. Fig. 26. — Fruit of Aleuritei Fordii (| natural size). Taste and smell. — The cold pressed oil has not the unpleasant taste and smell of the hot pressed oil. Special Tests. — High iodine value. Hexabromide test positive. Uses op Oil. — The seeds of Aleuntes tiiloha are said to produce a pure edible oil, but that from A. moluccana (as has been pointed out by Lewkowitsch) has purging properties, so that care must be exercised in distinguish- ing the seeds and the products. The failure to recog- nise this distinction probably accounts for the wide variation in published analytical figures. There is also a sharp line of distinction to be drawn between the oil from A. triloba, and that from ^4. Fordii and A. cordata, Candle Nut Oil 251 which are closely allied species, the fruit of A. Fordii being shown in Pig. 26 for the sake of comparison. Both these species yield a poisonous oil which is employed commercially for varnishes, etc. as tung oil or Chinese wood oil. Candle nut oil resembles linseed oil in many respects, and, seeing that it can be procured in such- large quantities, is used for the same technical purposes. Uses of Non-Fatty Poetions. — Smetham (Analyst, vol. XXXV, 1910; p. 55, abs.) gives ah analysis of the cake as follows : Water 7-10 Oil . . 16-01 Albuminoids 42-87 Digestible carbohydrates . . 19-92 Woody fibre . . . ^. 5-95 Mineral matter 8-15 100-00 It will be noted from the above analysis that it is exceptionally i-ich in albuminoids. The authors have extracted samples of the oil from tlie seeds, for which they have obtained the following fisrures : AleuriteB moluccana. AUurUex triloba * Saponification value Iodine value (Wijs) Refractive index at 40" C. (Zeiss butyro-refraotometer) Free fatty acids (as oleic) . 190-3 164-0 65-7 20-1% 202-5 143-8 61-8 1-0% ■ The oil remained liquid below zero without depositing " stearine.' 262 Fatty Foods Rubber Seed Oil. SuURCE. — Seeds of Hevea hrasiliensis (the Para rubber tree). A large euphorbiaceous tree indigenous to the forests of the Amazon valley, but now verj' extensively cultivated in Ceylon, Borneo, Malay States, Java, etc. The fruit, as shown in sketch, is typical of this order ; it encloses several seeds which are about the size and shape of a large damson. Each seed consists of a dark brown, smooth shell, characteristically marked with black, which shell loosely encloses the oleaginous kernel, which is covered with a cotton-like, fibrous coat. Weight of 100 seeds = 360 grm. ; weight of kernels = 180 grm. Content of Oil. — (a) In whole seed, 22—25 per cent. ; {li) in kernel, 45—48 per cent. Desceiptiok of Oil. — Apjtearance. — Deep brownish- j-ellow, about the same shade as average linseed oil. Consistency . — Liquid at ordinary temperatures. Taste and smell. — Almost identical to those of lin- seed oil. Special Tjssts. — Exceptionally high iodine value and refractive index. When determining the saponification value it will be noticed that an appreciable quantity of a " gutta "-like body separates, which is insoluble in petroleum ether and is therefore not included in the unsaponifiable matter as usually estimated. Properties. — A fairly good drying oil, producing a hard clear transparent film on exposure to the air for a period of twelve days (Pickles and Hayworth, Analyst, 1911, vol. XXX vi, p. 491). These authors also draw attention to the fact that in the liquid fatty acids (which Rubber Seed Oil 263 make up 86 per cent, of the total present), there is nearly 51 per cent, of linolic acid. 2 13 Fio. 27. — 1. Fruits of Hevea brasiliensis. 2. Seed of ditto. 3. Section of seed. (Natural size.) Uses of Oil. — At the present time no quantity of the oil is prepared commercially owing to the fact that 254 Fatty Foods the seeds are all required for planting purposes, but in view of the extensive area over which this tree is now being cultivated, it is inevitable that a very large quantity of the seeds will eventually be produced for which a use will have to be found, and it is therefore to be expected that in the near future the oil will appear on the market in considerable quantities, in which case it will form a substitute for linseed oil, and no doubt attempts will be made to utilise it for edible purposes, but as to its suitability in this direction we do not wish to make any definite statement — particularly since any oil intended for this purpose would have to be very carefully ex- tracted in order to avoid as far as possible the action of the lipolytic enzyme present in the seed. According to Dunstan (Ohem. Soc. Proc, 1907, vol. xxiii, pp. 168, 169) the seed kernel when ground with water evolves small quantities of hydrocyanic acid and acetone, thus indicating the presence of a cyanogenetic glucoside. There is also evidence of an enzyme capable of decom- posing this glucoside. Uses op the Non-Fatty Portion. — This portion is hardly likely to be utilised for edible purposes unless its poisonous properties can be counteracted, and in any case it will be necessary for extensive trials to be carried out before its value as a cattle food can be ascertained. ^Analysis of Seed. Moisture . . .11-00 Oil . 24-87 Albuminoids . . 9-63 Digestible carbohydrates 22-55 Woody fibre . . 30-40 Mineral matter , .1-55 100-00 Smetham (Analyst, 1910, vol. xxxv, p. 59, abs.). Rubber Seed Oil 255 Paea Edbber-Seed Oil. Determination. Usual limits. Typical specimen. Melting-point, ° C, incipient fusion _ _ „ „ complete fusion — — Solidifying point. °C — — Saponification value, ° C. 190 to 195 191-7 Eefractive index at 40° C. (Zeiss butyro- 61 to 64 62-7 refractometer) Iodine value, (Wijs) 135 to 140 138-7 Specific gravity, ^^'' C. Free fatty acids (as oleic) 0-924 to 0-927 0-9258 Varies 5-26% greatly Unsaponifiable matter . — 0-53 % Melting-point of fatty acids, ° C. . — — Optical rotation (angular degrees in — - 0-04 200 mm.) Note. — The figures for Ceara rubber oil are not greatly different from the above. 266 Fatty Foods Walnut Oil SoQBCE. — Seeds (walnuts) of the well-known tree Juglaiis regia, indigenous to Western Asia, but now in cultivation throughout Europe, especially in Southern France. The kernel from which the oil is obtained is too well known to require any description. The best oil is obtained by pressing kernels when they are two to three months old, as before this time a good deal of albuminous matter separates with the oil ; but if, on the other hand, the kernels are kept too long, con- siderable hydrolysation will take place. Content op Oil. — In kernel, 60-64 per cent. Description op Oil. — Appearance. — Pale greenish- yellow to almost colourless. Consistency. — Liquid at ordinary temperatures. Taate and smell. — Pleasant and nutty. Special Tests. — High iodine value and refractive index. Very low solidifying point. Bellier's test for poppy-seed nil (Ann. chim. appliq., 1905, Hi). — 1 c.c. of the oil is warmed with 5 c.c. of a 16 per cent, solution of KHO in 92 per cent, alcohol till clear, the tube closed and kept at 70° for twenty minutes. To the solution is added sufficient 25 per cent, acetic acid to neutralise the KHO used (ascertained by previous trial). The tube is again corked and cooled to 25° 0. and finally placed in water at 17° to 19° C, shaking frequently. Pure walnut oil only deposits a very small quantity of fatty acids after long standing, whereas poppy seed gives an abundant precipitate. It is doubtful if less than 20 per cent, of poppy-seed oil can be' detected. Walnut Oil 257 Ilalphen's test for linseed oil (Bull. Soc. China., 1905, xxxiii, p. 571). — 0'5 c.c. of the oil are dissolved in 10 c.c. of ether and 3 c.c. of hromine solution (made by adding sufficient bromine to carbon tetra-chloride to increase the volume by one half) added. The tube is closed and the contents mixed by inversion and kept at 25 0. In the presence of linseed oil turbidity occurs in less than two minutes. Poppy seed does not interfere. Usual Adui^teeants. — Principally poppy seed and linseed oils and other liquid vegetable oils. Uses of OiL.-^Owing to the demand for whole kernels the amount of oil expressed is somewhat limited, and is largely used in the manufacture of the finest artists' colours. From time to time, however, it appears as a salad oil, and sometimes as an adulterant of olive oil when the price permits. Uses oe Non-Fatty Portion. — As a feeding stuff for cattle, but only to a limited extent. Analyses , Moisture Kernel. 4-21 Cake (calculated) 13-0 Oil . . . 61-90 7-0 Albuminoids . 18-68 44-0 Digestible carbohydrates Woody fibre Mineral matter . . 11-94 1-26 2-06 28-2 2-9 4-9 100-00 100-0 17 268 Fatty Foods Walnut Oil. Determination. Usual limits. Typical specimen. Melting-point, incipient fusion _ „ „ complete fusion — — Solidifying point, ° C — -20 Saponification value 193-197 195-4 Eefractiveindexat40°C. (Zeiss butyro- 64-5 to 67-5 64-7 ref ractometer) ... • Iodine value (Wljs) Specific gravity W C. . » S3 TT *-'• • Free fatty acids (as oleic) 138 to 148 140-8 i 0-925 to 0-927 0-9256 Varies 3-76% tJnsapouifiable matter .... 0-5 to 1 % ] Melting-point of fatty acids, 'C. . 15° to 20° Perilla Oil 269 Perilla Oil. SouEOE. — Seeds of Perilla ocymoides (Natural Order Labiatas) . Indigenous to N.W. Himalayas, China, and Japan. The seeds are very small, being about the size and shape of rape-seeds. Cont'ent of Oil. — In seed, 33—35 per cent. Description of Oil. — Appearance. — Dark brown. Consistency. — Liquid at ordinary temperatures. Taste and smell. — Like linseed oil. Special Tests. — The iodine value is higher than that of any known fatty oil. Uses op Otl. — Largely used for edible purposes in the country of origin, but its use for this purpose in Europe is not known. It is, however, sometimes added to Japan wax to assist in the expression of the latter, and in this way may find its way into food products. Uses op the Non-Fatty Portion. — According to Sir George Watts' Dictionary of Economic Products of India, the seeds are used as a food, and from this it may be inferred that the residual cake from the preparation of the oil may also be used as a feeding-stufE. The following figures are due to Wijs (abs. Journ. Soc. Chem. Ind., 1903, vol. xxii, p. 805). Specific gravity |-§-° 0. Saponification value . Iodine value Free fatty acids Melting-point of fatty acids 0-9306 189-6 206-1 0-48 - 5-0° C. Castor Oil Group 261 CASTOR OIL GROUP. 1. Oastoe Oil. 2. Chaulmugea Oil G-roxif. 262 Fatty Foods Castor Oil. Phaemaceutical Name. — Oleum ricini. Source. — Seeds of Ricinus communis. Many varieties of the plaut are grown in India, which country is also the largest producer. Large quan- tities of seeds are also produced in the British African Possessions, Bast and West Indies, North and South America, French Indo-Ohina, Mexico, etc., and recently samples of castor seeds from Uganda, Anglo-Egyptian Soudan, Ceylon and Fiji have been examined and the Fig. 28. — Ricinus communis (Castor oil tree). 1. Seed. (1^ times natural size.) 2. Fruit enclosing several seeds. (Natural size.) oil found to be equal, and in some cases superior, to that of the Bombay seed.^ Seeing that there are such a large number of species cultivated, the imported seed varies very much in size, shape and markings. The fruit is enclosed in a characteristic spiny husk, which is shown partly ruptured in the sketch. This fruit contains several seeds, the markings on which resemble those on the Para rubber seed (see sketch, p. 253). The soft kernel is enclosed in a thin shell, 1 Bull. Imp. Inst., 1911, ix, pp. 17-35. Castor Oil 263 which amounts to about a fifth of the whole seed. This shell contains practically no oil. The Indian Trade Journal, 1908, vol. ix, p. 309, and 1909, vol. xi, p. 278, mentions a process for destroying the toxic principle of castor seeds. The decorticated seeds are ground with water to a cream and poured into boiling water in order to effect the coagulation of the proteins. The solid residue, which is strained, dried, and freed from oil, gave the following composition : Moisture . . 9-2 Oil . . 2-6 Ash . . 7-3 Protein . 71-7 Fibre . . 5'0 Carbohydrates 4'2 Weight of 100 seeds = 48-7 grm. Content of Oil. — (a) In Avhole seed, 45-50 per cent. ; (&) kernel, 55—60 per cent. DisscKiPTiON OP Oil. — Appearance. — Very pale straw colour, varying to a greenish tint. Consistency . — Liquid, but very viscous. Taste. — Characteristic and unpleasant. SmeZZ.— Slight. Optical Activity. — Strongly dextro-rotatory. Possible Adultbeants. — Cotton seed, rape and rosin oils. Special Tests. — Miscible in all proportions with absolute alcohol and glacial acetic acid. It should dissolve completely in three volumes of 90 per cent, alcohol at 20"^ C. The oil, when shaken with petroleum ether, takes up an equal volume of the solvent (in which it does Qot dissolve), the excess of petroleum separating out. The presence of another oil, however, causes the whole mixture to be homogeneous, so that this insoluble property of the oil cannot be used to detect adultera- tion. The specific gravity, acetyl value and viscosity are exceptionally high, the first being the highest of 264 Fatty Foods any known, fatty oil. The saponification value is low for a liquid oil. These characteristics, together with its optical activity, render its recognition a matter of no great difficulty. Uses op the Oil. — The refined oil is never completely tasteless or odourless, as any attempt to carry the refine- ment too far depreciates its medicinal value very considerably. It is most extensively employed for medicinal purposes on account of its well-known. aperient properties. Uses op Non-Fatty Poetion. — This portion is of no use for cattle feeding or any edible purpose owing to the presence of the poisonous alkaloid ricin. As previously mentioned, processes have been devised for the removal of this substance, but the idea does not seem to be greatly favoured. At present the only use of the non-fatty portion is as a manure. The lipase from the seeds is commercially employed for " fat-splitting," and the recent experiments of Dunlap and Gilbert (Journ. Amer. Chem. Soc, 1911, vol. xxxiii, 1787) on castor oil seeds (freed from fat) seem to show conclusively that the enzyme contained in these seeds is able to synthesise fats, from an emulsion of glycerine and oleic acid. Castor Oil Castor Oil. 265 Determination. Usual limits. Typical speci- men. Melting-point, incipient fusion „ ,, complete fusion Solidifying point, ° C. . Saponification value . . Refractive index at 40° C. (Zeiss butyro- refractometer) Iodine value (Wijs) Specific gravity, y|-° C. „ f«°c. . . Free fatty acids (as oleic) TJnsaponifiable matter . Melting-point of fatty acids, ° C. . Optical rotation (angular degrees in 200 mm.) - 12° to - 18= 175 to 185 68 to 70 83 to 90 0-960 to 967 Usually un- der 2 % 0-3 to 0-6 % ■^ 7-5 to -H 9-0 -18° 180-2 68-9 87-8 963 0-72% 0-35 % + 8-8 Special values : Reichert-Meissl value Acetyl value (true) 1 to 2-5 146 to 150 — 266 Fatty Foods CHAULMUQRA OIL GROUP. This group of oils is distinguished by the poisonous properties of its members and also by their exceptional dextro-rotatory powers. It consists of the following : (1) Chaulmugea Oil. — Obtained from the seeds of Taraldogenos Kurzii. (2) LuKRABO Oil.' — Obtained from the seeds of Hydnncarpus an ihelmin thica. (3) Hydnocaepds Oil. — Obtained from the seeds of Mydnocarpvs IVightiana. Lendrich, Koch and Schwartz, who have made an investigation of the margarines which caused the poisoning cases in Hamburg in 1910 (Zeit. Unters. Nahr. Genussm., 1911, xxii, p. 441), point out that these fats contain chaulmugric and hydnocarpic acids, and that their action is similar to that of croton oil. Chaulmugra (or Chaulmoog^ra) Oil. Source. — Seeds of Taraktogenos Kvrzii, a forest tree of Burmah, Chittagong, Assam, etc. The fruits (according to Lendrich and others [s?6prn]) are about the size of an oi'ange, the seeds being em- bedded in the pulp. The dried fruit and seed have a strong, unpleasant odour. Content of Oil. — In seed, about 40 per cent. Description op Oil. — Appearance. — White to whitish- yellow. Consistency. — Similar to butter but somewhat softer. Taste. — The fat is poisonous and a vesicant. Smell. — Pronounced. Chaulmugra Oil Group 267 Optical Activity. — Strongly dextro-rotatory. Special Tests. — The saponification value is high but the optical activity is distinctive. Uses oi' "Oil. — The oil is used medicinally, and is simply referred to here on account of the fact that it has found its way into edible fats. For values, see p. 269. Lukrabo Oil. Native Name op Skiods. — Ta-fung-tsze. TuAUE Names.— Marotty or Maratti oil. SooECE.^Seeds of ffi/dnocarpus anthelminthica, indi- genous to Siam and Northern Cochin-China, whence they are exported to China. Content of Oil. — In whole seed, 17'6 per cent.' (The kernels form 31 per cent, of the whole seed.) For values, see p. 269. Power and BarrowclifF, Journ. Chem. Soc, 1905, p. 893. 268 Fatty Foods Hydnocarpus Oil. Source. — Seeds of Hijdnocarpus Wightiana, indi- genous to the Western peninsula of India from South Fig. 29. — 1. Seed of Taraktogenos Kurzii (Chaiilmugra). (Natural size.) 2. Truit of Hydnocarpus Wightiana. (| natural size.) 3. Seed of ffydnocarpus venenata. (Natural size.) Concan to Travancore. (The kernels represent 75 per cent, of the whole seed.) Content of Oil. — In whole seed, 41 '2 per cent.' Luki-abo and Hydnocarpus oil are somewhat similar to Chaulmugra oil, and are used for the same purposes. ' Power and Barrowoliff, Journ, Cheni. Soc, 1905, p. 886. Chaulmugra Oil Group Determination. Chaulmugra oil (expressed) . Lukrabo oil. HydnocarpuB oil (expressedl. Melting point, ° C. . Free fatty aoids Saponification value . Iodine value 22° to 23° 23-9% 213-0 103-2 + 52-0 24° 11-25% 211 5 81-6 + 44-1 22° to 23° 3-8% 207-0 101-3 ^ + 57-7 Authority . Power and Grornall Bolton and Eevis Power and Barrowcliff For a very careful investigation into the oils of tliis group, see Power and Gornall (Chem. Soc. Journ., 1904, vol. Ixxxv, p. 838), and Power and Barrowcliff [ibid., 1905, vol. Ixxxvii, p. 884). 270 Fatty Foods MYRISTICACE/E GROUP. Of this group some values are given for the fats derived from the seeds of the following : (1) Myristica fro grans (nutmeg). — Occurring in the Moluccas, East and West Indies, 13anda Islands, Brazil, etc. (2) Myristica surinamensis (Cuajo nut). (3) Myristica guatemalensis. (4) Scyphocephalinm, Ochocoa (Ochoco nut). When cut in transverse section the seeds of this group Fio. 30. — Scyphocepiialium, Ochocoa ; transverse section of seed. (Natural size.) show a characteristic structure, that is, the peculiar ramifications of the inner seed coat into the endosperm in the form of irregular brown strips, giving an appear- ance somewhat resembling marble. In longitudinal section these ingrowths are seen to arise from the chalazal end of the seed. This is typically shown in the Ochoco seed (Fig. 30), and it is this type of structure which leads to the fat being always a dark brown to black in colour, the endosperm itself, and consequently the oil, being quite white. The fat derived from these seeds is characterised by a high percentage of the glyceride of rayristic acid. They yield hard fats of tenacious character and MyristicaccEe Group 271 generally of a yellowish-brown colour, with the ex- ception of ochoco fatj which is brown to black {supra). The fats possess somewhat high melting-points and sharp and definite solidifying points. They often contain a considerable percentage of an essential oil, which, being unsaponifiable, has led to the publication of very vary- ing figures for this value. The true saponification values are probably high and due to the presence of much myristin, which also accounts for appreciable Reichert-Meissl and Polenske values. With the ex- ception of nutmeg butter, iodine values are low and refractive indices high. Practically pure "myristin can usually be isolated from the fats by crystallisation from ether, and can be identified by preparing the myristic acid and determining the molecular weight, melting- point, etc. (molecular weight = 228 and melting-point 53°-54°0.). The following figures will be a guide to the recognition of the fats of this group : 272 Fatty Foods s 59-2 (B and fi) 45-48 2S8-5 1-72 0-65 40 4 O 1 4 59-0 50-0 41-0 38-0 223-229 51-0 11-5-18-3 0-8862 14-0-19-0 % 1-1 5-9 (O 1 g 1 'o .s g o pq If ■S a 38-40 38-51 150-180 60-67 45-65 0-898 9-15 % § 1 Fat, % in kernel . „ seed Melting-point, ° C. . . . Solidifying point, ° C. Saponification value . ... Eefraotive index at 40° C. (Zeiss butyro- refractometer) Iodine value . . . . Specific gravity yy ° C . Free fatty acids (oleic) Keichert-Meissl value Polenske value Fish and Marine Animal Oils 273 FISH AND MARINE ANIMAL OILS. This group of oils does not strictly come within the category of edible oils, but brief mention is made of them, since, as we have already said, it is by no means unlikely that some of them may, in the near future, be pressed into the service of the margarine manufacturer, when the present processes of de-odourisation shall have been sufficiently perfected. Evea at the present time tasteless and odourless fish oils have been prepared. The following are the main characteristics of the group : (1) The oils are usually liquid at ordinary tempera- tures, but generally deposit " stearin e " when cooled. (2) Their iodine values are usually high, varying from 100 to 160 per cent, or more, being due to the presence in tlieir glycerides of fatty acids of greater unsaturation than oleic. In fact, the degree of unsaturation sur- passes that of the vegetable drying oils, in that many of the fish oils form octobromides, whereas the vegetable oils rarely form higher compounds than the hexabromides. For this reason the hexabromide test (p. 42) becomes of great value in the detection of the oils of this group. {■i) The other values which are obtained for these oils closely follow those of many vegetable oils, and do not allow of their detection in admixture with them. As just mentioned, the hexabromide test will detect these oils in other oils which do not give any reaction with that test. The reaction given by the vegetable drying oils alone would not be mistaken after a little experience for that given by the oils of this group as the precipitates are easily distinguished, but the test 18 274 Fatty Foods naturally does not admit of the detection of oils of this group in admixture with the vegetable drying- oils (more particularly with linseed oil) ; for which reason they form a favourite adulterant of these oils. The presence of fish and marine animal oils, however, may be detected by the following methods : (a) A simple qualitative test may be carried out hj saponifying the oil, when even the most highly refined fish or marine animal oil develops a characteristic fishy odour, though this may have been quite undetectable in the sample before saponification. (6) It has been pointed out by Lewkowitsch that a real distinction can be obtained by determining the melting-point of the ether-insoluble bromides prepared from the fatty acids derived from the oil. The method of carrying out the test as slightly modified by the authors is as follows : 0-3 grm. of the fatty acids are dissolved in 5 c.c. of glacial acetic acid and 20 c.c. of methylated ether, and the solution cooled in a closed flask to 5° C. Bromine is then added, drop by drop, until its colour is perma- nent, the solution being kept at 5° C. during the process. If the colour on standing becomes pale yellow, a few more drops of bromine must be added. The llask is now allowed to stand at this temperature for three hours, shaking occasionally, after which the precipitate is filtered off through a folded filter and washed six times in succession with 5 c.c. of ether at 5° C, the residue after the last washing, while still moist, being placed on a piece of porous tile. The ether-insoluble bromides obtained in this manner from linseed oil melt at 180°— 183° C. (uncorrected), whereas the similar compounds derived from the fish oils have no proper melting-point, but blacken at 200° C. without melting. If, therefore, the ether-insoluble bromides derived from a linseed oil Fish and Marine Animal Oils 276 melt at a temperature distinctly above 183" C. and show signs of blackening, the presence of fish oils may be regarded as certain. The bromides from linseed oil always melt to a char liquid, and the darkening of this liquid due to the presence of even as small a quantity as 10 per cent, of fish oil can be readily seen. (c) The oils prepared from the jaws of certain marine animals, particularly dolphin and porpoise, yield extremely high I},eichert-Meissl values, largely due to the presence of valeric acid. 276 Fatty Foods S ^ o ^ o n (N ■4 o CO T-t •T o 5 o H CD QO CO t Ci ■N t-, iH 05 > . CO ?0 m r-t ca 00 OS T-H ? CO N (M 05 o> c: O O c! O O o R O 6 O) W -- o e i> ^1 00 CO en ■* OS .— 1 CO 9^ 00 00 05 OS Ci rH t-H tr- c^ ci m 13 TJ Id ^ rt ID T3 ?n dD / /f / 1 J 1 7 ~-~ / --^ "~~~^ ao 40 00 80 1 PCKCENIAGES or COMPONENTS oo (L) (N) (l) = Levigatlon curve. (f) = Fibre curve. (n) =Nilrogi>li curve. (See Fig. 32.) Cocoa 307 Calculation of Amount of Shell and Nib. The accompanying curves will assist the analyst to determine approximately the proportion from the analytical figures obtained for the nitrogen and fibre in the fat-free-dry material, and more particularly from the results of the levigation process. They are based on a slight modification of the limits in the text. Fig. 32 is the nitrogen curve, on a larger scale for readier reading. It is not to be expected that the various estimations will give the same proportion, and a mean should be taken, rather more reliance being placed on the nitrogen figures than on those of the fibre, and most certainty being allocated to the levigation results. The results so obtained are then calculated on the original sample. The above assumes that there is nothing present but " cocoa-matter." ' If there be added starch, etc., the calculations are somewhat more complicated and the actual cocoa-matter present must be obtained from the cold water extract figure. The methods of arriving at this in such cases are given under " Chocolate.'" Chocolate. As we have pointed out above, this substance will contain all the constituents of cocoa in varying propor- tions, and will also contain cane-sugar, and in many cases small quantities of starches other than those derived from the cocoa bean. The methods of examination are practically those used in the analysis of cocoa. The determination of the cold water extract becomes of greater importance as furnishing the only basis from which the proportion of fat-free-dry cocoa matter may be calculated after 308 Fatty Foods Flour. extract. Wheat 7-7 Barley 5-1 Maize (cornflour) . . 0-8 Rice . 0-9 Sago . 1-98 Arrowroot . 0-4 correcting it for the amount of sugar present, and, if there be any appreciable quantity of added starch, for the cold water extract of this starch. The following table, due to Booth, Cribb and Richards, gives the cold water extract and nitrogen figures for likely starches. Nitrogen. 1-97 1-2 0-14 1-23 0-0.3 0-13 In order to arrive at the amount of cocoa matter and the added starch in a chocolate the following method will be found of some advantage : In the fat-free-dry sample the tgtal nitrogen, total fibre, total starch, and total cold water extract are estimated, together with the amount of cane-sugar. The nature of the added starch is found from the microscopical examination, and the true values for the percentage of actual cocoa matter and added starch can then be found by applying the following equations : In the fat-free-dry sample let — X — percentage of true cocoa matter (required). Y = percentage of added starch (required). Q = percentage of cold water extract of added starch (found from table). T = percentage of total water extract in sample (deter- mined and corrected for sugar only). *S(, = percentage of starch in cocoa matter. 8 — percentage of total starch in sample (determined). N = percentage of total nitrogen in sample (deter- mined) . N^ = percentage of nitrogen in added starch (found from table). Chocolate 309 F = percentage of fibre in sample (determined). Then assuming an approximate value of 10 for the value /S„. V2400 - 8^.0/ / 24 8 - T.8\ \2400 - 8,.g)' of 8a be desired, /T- 3-73 N - 0-89 J?" \ \ C~^ 3-73 N, /' and Y = 100 Or, if no assumption of 8^ be desired, then Y= 100 andX = 100 T. - G.Y 24 assuming in the above equations that fat-free-dry " nib " contains 5 per cent, of nitrogen and 6 per cent, of fibre, and that fat-free-dry " shell " contains 2'5 per cent, of nitrogen and 16"5 per cent, of fibre. The actual percentage of nitrogen in the cocoa matter {X) may now be determined by subtracting that due to the added starch (F), whence from the fibre and nitrogen values of the cocoa matter the relative pro- portions of nib and shell may be found from the curves given on pp. 305—6, and the result should be in substantial agreement with the figures obtained from the levigation process. All values are then calculated to the original sample.* • The corrections for nitrogen and cold water extract are only necessary when wheat or barley starches are present. The methods of analysis are, as stated, similar to those of cocoa, with some slight modifications : The sample should be scraped or shredded with a knife. Determinations of moisture, nitrogen, fat and cold water extract are exactly the same as in the case of * For example of method of calculating results, see Appendix D (p. 362). 310 Fatty Foods cocoa. For fibre 15 grm. should be used. For the starch 5 grm. are used, and care must be taken that all cane-sugar is removed by washing. When the polarimetric method is employed, this may be subsequent to the acetic acid treatment. Washing is quite simple if the large Buchner filter be used. For estimation of the sugar two separate quantities of a quarter and half the normal weight of the fat-free-dry material are dissolved in 100 c.c. of water, and cleared after fiiUing to the mark with solid basic lead acetate (0"1— 0'3 grm.), or by the addition of 5 c.c. of plumhi snhacetatis Hquor B.P. (see below) and 5 c.c. of alumina cream, allowing, in the latter method, for the dilution. Multiply each reading by two, when the true reading (= percentage of sugar in fat-free-dry material) equals the product of the readings divided by their difference. For Ventzke scales (J512 grm. and 13"024 grm. will be employed, and the true reading (if in saccharimeter degrees) is direct percentage of sugar (angular degrees X 28"82 = saccharimeter degrees). If Pehling's method (p. 314) be employed a 2-5 per cent, solution of the chocolate should be made, and no correction for the volume of the cocoa matter is neces- sary. The procedure is that given under "Cocoa" (p. 298). Preparation of Basic Lead Acetate Solution. If the B.P. preparation be not available, the solu- tion may be made by boiling 430 grm. of normal lead acetate and 130 grm. of recently ignited litharge with 1000 c.c. of water. After cooling and settling the clear liquid is syphoned off, and diluted to sp. gr. 1-25. Alumina cream may be made by adding ammonia to a strong solution of alum till there is only a slio-ht excess of the alum left. Milk Chocolate 311 Milk Chocolate. The analysis of milk chocolate presents greater difficulties than that of plain chocolate, and becomes very complicated when both nib, shell and foreign starch are present. We think it advisable to deduce the amount of milk-solids present from an accurate determination of the lactose in the fat-free-dry sample, confirming the percentage of milk-solids so obtained by an estimation of the actual casein present. The non-fatty solids present in the mixture due to the original milk employed can then be calculated on the very fair assumption that, of the non-fatty solids of that milk, thirteen twenty- fourths are anhydrous lactose and eight twenty-fourths are casein. The total milk solids in the fat-free-dry sample so deduced are subtracted from the total solid matter of the fat-free-dry sample, when the remainder will be cocoa matter plus foreign starch plus cane sugar. The cane-sugar is determined at the same time as the lactose by the method to be described, when the mixture left will resolve itself into coc.oa-matter and foreign starch, the relative proportions of which may be determined as already described under " Cocoa." That is to say, (a) the total starch by diastase or Ewers' method, after washing out the cane-sugar, (/;) the total nitrogen in the original fat-free-dry sample- corrected for the nitrogen due to the milk proteins (the amount of which will have been already arrived at), (c) the total fibre, which may be estimated on the original fat-free- diy sample, and ('?) the proportion of shell by the levigation process. These \arious determinations being made, the formula given under ''Cocoa" are applied and the percentage of shell, nib and added starch will then be arrived at. It will, however, be obvious that, except in the case 312 Fatty Foods of milk chocoLite made from pure nib, dried or con- densed milk and cane-sugar, the analysis is beset with difficulty on account of the wide variations in the com- position of the various constituents and the relativelj"^ small proportion in which some of them may be present. The investigation has, therefore, rather the character of a scientific than of a technical analysis, but the appli- cation of care in the analytical details and of common- sense in their interpretation will in most cases lead to satisfactory results. The following are the special methods of analysis connected with milk chocolate. Cane and Milk Sug-ars. Ten grm. of the fat-free-dry residue are placed in a 100 c.c. flask, dissolved by shaking with 70 c.c. of water, and 15 c.c. of acid mercuric nitrate solution carefully run in, allowing the liquids to mix as little as possible. The contents are adjusted to the mai-k with water quickly ^nd without disturbance, and then rapidly and thoroughly mixed by vigorous shaking. The con- tents of the flask are cooled to about 1 0° C. (in order to avoid incipient inversion of the cane-sugar by the acid mercuric nitrate) and then filtered at this temperature, and a direct polarisation made at room temperature as quickly as possible. Fifty c.c. of the filtrate are placed in a small flask, and counterpoised. The flask is then heated in boilino- water for seven minutes (evaporation being avoided as much as possible), cooled, re-adjusted to tlie original weight, filtered if necessary and polarised at the same temperature (t) as that of the direct polarisation. In order to allow for the bulk of the preoipilate produced by the mercuric nitrate, an exactly similar Milk Chocolate 313 experiment is carried out using 5 grm. of the sample. From the readings obtained in the two experiments the true values are found in the following manner : If Rj^ is the reading of the 5 grm. dilutionj and iJgis the reading of the 10 grm. dilution^ then the true reading for the 10 grm. dilution is given by the formula, /(*. /? B = — ^ — 5_ both for direct and invert readings. The true reading having been thus obtained, the percentages of cane-sugar and lactose are calculated in the following manner : (a) For instruments using the Soleil Ventzke scale : , (D-I) 100 N Cane sugar per cent. = '-^ — -^ — — x —= ;S' and t n' 142-68-- z Lactose per cent. j '^ — ^\~) \ 0'3]213 (anhydrous) =i ^ZJ x 100 ; where D and / = the true direct and invert readings for the 10 grm. dilution, N = the normal weight foi the instrument, W = the weight of sample employed (i.e. 10 grm.). (b) For instruments reading in angular degrees : 21-7 [D-I) _ ^ The reading due to cane-sugar = „.. „.. — ^ ^ — -a. (Tuchsmidt's formula), then D — X— reading due to lactose. And as 1 grm. of cane-sugar dissolved in ,100 Hohr c.c. gives a reading of 1"33 angular degrees, then — XxW . = percentage or cane-sugar m sample; and since 1 grm. of lactose monohydrate dissolved in 100 Mohr c.c. gives a reading of r058 angular degrees, then — 314 Fatty Foods (D — X) X If percentage of lactose monohj-drate in 1 OoS the sample, which result must be divided by r053 to get aniiydious lactose per cent. It may not be out of place to remind the observer that if I be negative, then-/ in the above formula is positive. If a polarimeter be not available, the sugars may be estimated by Fehling's method: For this purpose 5 grm. of the fat-free-dry sample are dissolved in water and made up to 100 c.c. ; 25 c.c. are inverted by citric acid (see " Cocoa," p. 298), cleared with basic lead acetate and sodium sulphate, and used for the determination of the lactose and inverted cane- sugar after making up to 50 c.c. Fifty c.c. of the remainder are cleared in the same way and made up to 100 c.c. and used for the estimation of the lactose. The solutions in each case are filtered, the filtrates being suitably diluted before titration if necessary. The sugars are determined against Fehling's solution in the following manner (Ling's method, using Harrison's indicator) : Solutions required : (o) Copper fiulphate solution. — 69"278 grm. of pure crystallised copper sulphate are dissolved in water and made up to 1 litre. (/;) Alhrdine tartrate sohitiun. — 346 grm. of Rochelle salts are dissolved in hot water and mixed with 142 grm. of caustic soda, previously dissolved in water, the whole being cooled and made up to 1 liti-e. (f) Indicator. — 5 grm. of pure potassium iodide are dissolved in about 40 c.c. of water. To this are added 005 grm. ol soluble starch, which has been dissolved in about 5 c.c. of water by boiling and cooled, the whole being then made up to 50 c.c. Milk Chocolate 315 A 10 per cent, solution of acetic acid is also required. Method of analysis. — On a white porcelain tile place a series of mixtures of two drops of the starch solution and one drop of the 10 per cent, acetic acid. A number of pieces of quill tubing about 8 in. long and about |- in. in diameter are placed ready to hand. 20 c.c. of the copper sulphate solution are then measured accurately into a flask and mixed with 20 c.c. of the alkaline tartrate solution. Of this mixture 10 c.c. are accurately measured into a 200 c.c. Jena flask, and brought to the boil. The sugar solution is now run in from a burette (2 c.c. at a time), boiling between each addition until the blue colour has nearly dis- appeared. A drop of the solution is now removed by one of the quill tubes and mixed with the indicator on the tile. (No filtration is necessary, and a fresh tube should be used for each test.) The sugar solution is then added 0'5 c.c. at a time, testing after each addition, until no colour is produced in the indicator after the lapse of at least a minute. The experiment is then repeated, running in the sugar solution as before down to the point which last produced a colour in the first experiment, and then by successive additions of 0'2 c.c. until the " no colour " point is again obtained. The number of c.c. of the sugar solution required in this last titration is read off and the lactose calculated as follows : Since 10 c.c. of the Fehling solution are reduced by 0"067 grm. of lactose monohydrate, then if X CO. of the sugar solution are required : Percentage of anhydrous lac-) 0'067 x 100 x 40, tose in the original sample) ~ X y. r053 since 100 c.c. of the sugar solution contains 2"5 grm- of the original fat-free dry chocolate. If the sugar solution has to be further diluted the factor 40 must be proportionately increased. 316 Fatty Foods The original filtrate df the chocolate would in most cases give the right concentration for the above determi- nation, though if less than 20 c.c. be required to reduce 10 c.c. of Fehling it should be suitably diluted until about 20 c.c. are required for that purpose. That portion of the filtrate which has been inverted with the citric acid should now be diluted until about 25 c.c. are required to reduce 10 c.c. of the mixed Fehling. (A rough experiment is made to determine the necessary dilution.) The total sugars from the latter titration are calcu- lated as lactose in an exactly similar manner, allowing for the extra dilution employed. The percentage of reaJ lactose is subtracted from the percentage of apparent total lactose so obtained, and the difference multiplied by 0"709 will give the percentage of cane-sugar. For general purposes it may be taken that 10 c.c. of mixed Fehling solution is reduced by 0"05 grm. of pure dextrose. In practice, however, it is not really sufficient to rely on this, and the Fehling solution should be set against pure dextrose, but as this is not easily obtainable it is practically always standardised against a solution of invert sugar of such a strength that the quantity of invert sugar required to reduce 10 c.c. of the mixed Fehling solution is contained in about 25 c.c. The following method of making this invert sugar solu- tion and the tables from which the amount of dextrose or invert sugar in each 100 c.c. of the test solution can be ascertained are given by Ling and Rendle (Analyst, 1905, vol. XXX, p. 185) and Ling and Jones (ihid., 1908, vol. xxxiii, p. 165). Pure sucrose (0"95 grm.) is dissolved in water (150 c.c.) and boiled with N/2 hydrochloric acid (30 c.c), the mixture being maintained in ebullition for one minute, cooled, neutralised by the addition of N/2 sodium Milk Chocolate 317 hydroxide (30 c.c), and made up with water to 500 c.c. This solution contains 0'2 grm. of invert sugar per 100 c.c. Volume of solution Dextrose in Invert sugar required by 10 c.c. 100 c.c. in 100 c c. Pehllng'a solution. of solution. of solution. 0.0. Grm. Grm. 20 0-2427 — 21 0-2332 0-2412 22 0-2226 0-2311 23 0-2138 0-2218 24 0-2056 0-2132 25 0-1981 0-2052 26 0-1911 0-1980 27 0-1846 01910 28 0-1784 01846 29 0-1728 0-1787 30 0-1675 0-1731 : 31 0-1625 0-1678 32 0-1577 0-1629 33 0-1532 0-1583 34. 0-1490 01539 35 0-1450 0-1497 ; 36 01412 0-1458 3V 01377 0-1421 38 0-1343 0-1385 39 0-1310 0-1849 40 0-1279 0-1319 41 — 0-1288 42 — 0-1259 It is also advisable to standardise the Fehling solution against pure lactose, for which purpose a water solution of the pure substance should be evaporated to dryness when the sugar is obtained in the anhydrous condition. Casein. The authors have employed a method of estimating the casein in milk chocolate based on its solubility in sodium oxalate or citrate. As, however, the method has been published in a somewhat improved form by 318 Fatty Foods Baier and Neumann (Zeit. Untersnch. Nahr. Genussm., 1909, xviii, p. 13), their method is here described : Ten grm. of the fat-free-dry chocolate are rubbed up in a mortar with a small quantity of 1 per cent, sodium oxalate solution, the paste washed into a 250 c.c. flask with about 200 c.c. of the oxalate solution, the whole heated to boiling and the flask filled nearly to the mark with hot sodium oxalate solution. The mixture is allowed to stand for eighteen hours, shaken occasionally, and is then filled to the mark with cold oxalate solution, mixed and filtered. To 100 c.c. of the filtrate are added 5 c.c. of uranium acetate solution, and then 30 per cent, acetic acid solution is dropped in slowly till the casein commences to precipitate. Five drops more of the acetic acid are then added and the precipitate sepa- rated, preferably by sedimentation or centrifugalisation (as it filters slowly) or on a Buchner filter (p. 297), and washed with a solution containing 5 grms. of uranium acetate and 3 c.c. of 30 per cent, acetic acid per 100 c.c. till the washings are free from oxalate. The filter and precipitate is dropped into a flask and the nitrogen is then estimated by Kjeldahl's process, using the factor G-39 to calculate the nitrogen to Feeding Stuffs 319 CHAPTER VIII. FEEIJING STUFFS. The analysis of feeding stuffs is carried out in a somewhat avbitrai-y manner, the vai'ious constituents being classed together under certain headings, without, as a rule, any differentiation being made among the various substances present which are included under these headings. The headings under which the con- stituents of feeding-stuffs are almost universally returned are moisture, ash, total protein, (alhuminoids) , fat, digestible carbohydrates, and crude fibre. 'I'his method of return does not give a real idea of the true feeding value of the substance analysed. For instance, there is no discrimination between the various nitrogenous bodies present (and on which the food value of the material largely depends), nor is any idea given of the actual availability of the constituents. This is a very great drawback to the value of such analyses, but custom has attached to them an impor- tance which will probably continue to be held. The actual availability of the various constituents can only be ascertained by proper digestion experiments carried out on animals, from the results of which factors of digestibility for each important class of constituents of a very large number of well-known feeding stuffs have been calculated. From, the analytical figures of a feeding stuff and its digestibility co-efficients, it is 320 Fatty Foods possible to ascertain the relative amounts of various food materials which will make up a satisfactory ration for animals under various conditions. The bases of these calculations are the so-called standard rations, and though the value of these standard rations has often been called in question, there is no doubt that under unknown conditions they furnish a valuable basis for experimental rations, both to the analyst and the stockman. A brief explanation of the method of using these standard rations is here given, and thougb while we would warn the analyst against calculating strict rations for animals unless he has had practical experi- ence of the subject, there is no reason why he should not be able to give guidance to the stock-keeper as to a satisfactory experimental ration, given the available feeding stuffs. For the purpose of calculating rations, feeding stuffs are divided into two classes : (a) Roughage, under which heading are included those feeding stuffs which provide bulk, such as hay, straw, grasses, silage, roots, etc. (b) Concentrates, under which heading are included such feeding stuffs as contain various nutrient materials, particularly protein and fat, in relatively large proportions. By the careful adjustment of these two classes of feeds, satisfactory rations of a most variable character which will fit in with available feeding stuffs can be made. There are various types of standard rations and of methods of arriving at the mixtures which will satisfy these. Some, however, are unnecessarily com- plicated, and the method here given is that chiefly used in America, and is based on what are known as the " Wolff-Lehmann " rations, and has proved both satisfac- tory in result and easy in calculation. For full details Feeding Stuffs 321 li . per' day per lOOO lb. live weight Digestible nutrients. Animal. Dry Matter. Nutritive ratio, 1: -> Digestible '■ Protein. - carbo- hydrates. . Oil. Omen — ■ At rest ; - '. ' 18 0'7 8-0 ' 0-1 11-8 Medium work 25 20 11-5 0-5 e;5 Fattening iditle — ' 1st ;^eriod . ; . 30 2-5 15-0 0-5 6-5 2nd period . 30 30 14-5 0-7 5-4 _ 3rd period 26 2-7 15-0 0-7 6-2 Milth cow&^- . Griving„16-6 lb. of milt . . ' .' 2l\ 20 11-0 0-4 60 ■GriVing" 22 lb. of milk ^ 29 2-5 130 , 0-5 5-7 Sorses — Light work . 20 1-5 9-5 0-4 7-0 lledium- work :24 20 11-0 0-6 6-2 Heavy work . 20 2-5 13-3 0-8 6-0 Growing cattle — Dairy breeds. is. "-iSS 2- 3 . 150 23 40 13-0 20 4-5 3-6 . 300 24 3-0 . 12-8 10 5-1 . 6-12 500 27 2-0 12-5 0-5 6-8 12-18 . 700 . 26 1-8 12-5 0-4 7-5 18-24 . 900 26 ■ 1-5 12-0 0-3 8-5 Beef breeds. 23 • 4-2 13-0 2-0 4-2 2-3 . 160 3- 6 . 330 24 3-5 12-8 1-5 4-7 6-12 . 550 25 25 .13-2 0-7 60 12-18 . .750 24 2-0 .12-5 0-5 6-8 18-24 . 950 . ,24 1-8 120 0-4 7-2 of this method and the tables connected with it, refer- ence should be made to the works of Henry (Feeds and Feeding, 1898) or of Armsby (Manual of Cattle Feeding : Wiley, New York.) 21 322 Fatty Foods The table on. p. 321 gives a selection of the Wolff- Lehmann standard rations in certain particular instances. These standards are for animals under normal con- ditionSj and are to be used, as guides only. The figures given under " dry matter " may be allowed to vary quite 10 or 12 per cent, if necessary. If animals are below normal size, it may be necessary to increase the nitro- genous nutrients by 0"3 lb. and the non-nitrogenous nutrients by 1'5 lb. per day per 1000 lb. of live weight. In the case of the standards for growing animals, allowance is made for a moderate amount of exercise ; if no exercise be given the figures are to be reduced 15 per cent., or if much exercise be given they are to be increased by 15 per cent. The " nutritive ratio," which is the ratio of the amount of digestible carbohydrates and oil to the amount of digestible protein, is calculated as follows : " The percentage of oil is multiplied by 2*4 (to make its heat value similar to that of the carbohydrates) and the product added to the percentage of digestible carbo- hydrates, the sum being then divided by the percentage of digestible protein." For instance, if a feeding stuff contains 10-2 per cent, digestible protein, 69"2 per cent, digestible carbohydrates, and 1-7 per cent, of oil, then the nutritive ratio equals — (1-7 X 2'4) + 69-2 „ „ ^ . iTTo ~ ' ^ -^^ expressed as — nutritive ratio = 1 : 7"2. If the nutritive ratio be high it is said to be " wide," and implies a large excess of carbohydrates over protein, but if the ratio be very small it is said to be " narrow," and points to a protein content practically equal to that of fat and carbohydrates combined. The value of the feeding stuff is sometimes expressed in "food units," which are calculated as follows : Calculation of Rations 323 The content of oil and albuminoids' are added and multiplied by 2-5, and to the product is added the content of digestible carbohydrates^ the result being "food units." The method of calculating rations is as follows : The analytical figures for various kinds of feeding stuffs being ascertainedj it is necessary to know the factors which represent the average digestibility of these constituents, which factors -have been ascertained by digestion trials. The various constituents are then multiplied by the corresponding factor, when the com- position of the feeding stuff in percentages of digestible nutrients is obtained. A selection of the available foodstuffs being made, they are combined in such pro- portions as shall give a composition corresponding as nearly as possible to that of the particular standard ration required. The following example will show the method of carrying this out : " It is required to provide a ration for a cow giving 22 lb. of milk per day, hay being available as roughage and flaked maize as a concentrate, together with oil -cake and some product such as middlings or offal, which are generally necessary to provide against constipation." On reference to the tables below (p. 325), the following analytical figures will be found for these sut)stances. Feeding stuff. Dry matter. Protein. Digestible Oil. carbohydrates. Hay . 85-7 . 9-3 . 41-3 . 15 Flaked maize 93-7 106 . 81-6 . 09 Middlings . 88-2 13-1 . 58-9 . 3-3 DigesHMlity Oo-efficients. Hay . ... 57 . 64 53 Flaked maize . . 84 , 85 90 Middlings . 82 . 85 . 85 ' This word is still used commercially instead of total protein. 324 Fatty Foods Digestible Nutrients calculated from above. Hay . . 5-3 . 264 . 08 Flaked maize . . 89 69-5 0-8 Middlings . 10-7 50-0 . 2-8 As not more than 2 lb. of middlings can be given per day on account of their purgative properties, a tentative ration of say 20 lb. of hay, 8 lb. of flaked maize and 2 lb. of middlings may be made up. The values of these are obtained from the above figures as follows : In 20 lb. of hay there are : 20 Dry matter . . 85-7 x— -=17-1. •' 100 20 Digestible protein . 5"3 x =1"06. ^ ^ 100 20 Digestible carbohydrates 26'4 x = ,5*3. 20 Oil . . . . 0-8 X =0-16. 100 The values for the maize and middlings being calcu- lated in a similar way, the value of the ration is given as follows : Dry matter. 17-1 75 1-8 Digrestible nutrients. Tentative ration. 201b. hay. 8 lb. flaked maize 2 lb. middlings Protein. 106 0-71 0-21 Carbohydrates. 5-3 5-6 10 Oil. 016 0-06 005 26-4 1-98 11-9 0-27 Standard ration 29-0 . 25 . 13-0 . 0-5 It will be seen that the ration falls short in each particular but principally in protein and oil, and this must be corrected by the addition of a foodstuff rich in these such as oil-cake, which will not increase the carbohydrates unduly at the same time. Linseed cake will meet these requirements, and from the figures for this foodstuff, given in the table below, the change brought about by adding 3 lb. of linseed cake may be ascertained. Feeding Stuffs 325 1 1 8 § I S § ■^QXlOOirHCQlOCOCOOlO cooScDooooosoocoaioiioco CO .H CO (M CO 1 00 lO CO CO -^ lO 1 Oi •e « '1 CD CO 5 lo 3 Oi oa ■i " 1 1 t- rH O O »0 CD O lO iH CO IM U3 t- OS g o 1 o o I § OJ>coocoeot-coococDco iH (H lo Oi CO lo (N W Tp CO CO CD Cq Ah O 00 tb t- OD lb ■* CO CO CO CO 1 0(McDoa)Ocq-^ocoj>io Cq CD -^ lO lO (N »S CO Oi O (N t- CO p , 6i cq "^ CO th »H tH i lOOW30W30iW3lO'tflOO>0 05t-»o rH 00 CO CO i-H .-( lb S fe ce t> S * .2 -^ g CO t-t ^ Eh «'" 2 a " o ■■^ ■" oi ^ fn OS .—I c * , — . .S 03 C U 03" g ° a X t-" CO CD 2 S '=' o . _. c f-' S § 2 03 3 t4 CtJ ^ -fe'S ■s <«1 Feeding Stuffs 327 Digestible nutrients. Dry matter. Protein. Carbohydrates. Oil. Ration as above 26-4 . 1-98 . 11-9 0-27 3 lb. of linseed cake 27 . 0-60 . 09 0-27 291 2-58 12-8 054 This compares quite closely with the standard ration, and gives a nutritive value of 1 : 5'4 as compared with 1 : 5"7 as required by the standard ration. This ration would therefore form a satisfactory basis for working purposes. The tables on pp. 325—6 gives the analytical figures and the digestibility co-efficients of a few well-known feeding stuffs. The analytical figures are a selection from a very great number published by Smetham (abstract, Analyst, xxxv, 1910, p. 54). The digestibility co-efficients applicable to them are taken from those published by Henry (Feeds and Feeding, 1898, p. 625). Molasses Cake. Some feeding stuffs are made by mixing molasses with various oil-cakes and maize in particular, or are a simple admixture of the saccharine material with peat moss. The sugars may be extracted by maceration with water and estimated in the usual manner by the polari- metric or Fehling's method [q. v). The following table gives the variation in composi- tion of molasses as given by Babington (Canadian Inland Revenue Dept., Bull. xxv). Polarisation (direct) 80 to 50° V. (invert.) -10 „ - -24° V. Sucrose by Clerget 32 „ 52 per cent. Reducing sugar 13 „ 24 „ Moisture 29 „ 32 „ Ash . . . 0-5 „ 4-0 „ These figures will enable an approximate idea of the content of cane-sugar molasses present to be obtained. 328 Fatty Foods A number of other analyses of feeding cakes will be found in Chapter V under their respective oils and fats^ in the preparation of which they form a bye-product. Analysis of Feeding Stuffs. As feeding stufEs which are sold under " 'J'he Ferti- lizers and Feeding Stuffs Act" have to conform to certain regulations, the methods laid down under the Act, as well as those which are generally , suflicient and applicable, are given. The ofBcial determinations apply only to the determinations of oil, moisture and albumin- oids, and prescribe certain limits of errors of sampling, and within these limits the composition, as stated in the invoice, must lie. The limits of error for oil and albuminoids ai'e given in Schedule II of the Fertilisers and Feeding Stuffs (Limits of Error) Regulations, 1906, as follows : Feeding stuff. Cake or meal : Decorticated cotton TJndecorticated cotton Earth nut or ground nut Palm kernel or palm nut Coconut Niger seed Sesame seed. Sunflower seed Hemp seed Kurdee or safflower Compound . Linseed Eape Maize .... All other feeding stuffs, not othcTwise specified in this schedule .... Limits of error. One tenth of the percentage of oil and one tenth of the percentage of albuminoids stated in the invoice. One eighth of the percent- age of oil and- one eighth of the percentage of albu- minoids stated in the invoice. One fifth of the percentage of oil and one fifth of the percentage of albuminoids stated in the invoice. Feeding Stuffs 329 Method op Sampling (Official). (tt) If the sample be already in a fine condition^, e. g. a meal, it should be thoroughly mixed and a portion for the estimation of the moisture taken at once. (6) If the sample be not in a fine condition, e. g. a cake, it should be carefully pulverised until the whole passes through a sieve with perforations 2 to 3 mm. in diameter. It should then be thoroughly mixed, and a portion for the determination of moisture taken at once. [The authors find that, for practical purposes, the simplest method is to pass the roughly broken cake through an ordinary cofEee mill (care being taken that the mill is not driven so fast as to heat the material), the residue inside the mill being b™shed out and mixed with the ground sample.] (c) Prom the sample thus prepared a portion not less than 100 grm. in weight shall be taken and further powdered, if necessary, and passed through a sieve with perforations of about 1 mm. diameter [which is quite unnecessary if a coffee-mill has been used]. (d) If the original sample be appreciably damp, or if the operations of pulverisation and mixing be likely to result in loss or gain of moisture, the moisture shall be determined in this portion as well as in the sample prepared as in' paragraph (6) in order that the results of the analysis may be corrected to correspond with the sample in its original condition as regards moisture. (e) Materials which cannot be conveniently pul- verised or passed through a sieve should be thoroughly mixed and sampled by the most convenient means. (/) The prepared portion of the sample shall* be placed in a stoppered bottle, and from it the portions for analysis shall be weighed. 330 Fatty Foods Moisture (Official). A weighed quantity of the sample shall be dried at 1 00° 0. [The authors employ for this purpose large flat stoppered weighing bottles, 2^ inches in diameter by li inches high, into which are weighed 3 to 5 grm. of the material according to its bulk. 3 Pat oe Oil (Official). A weighed quantity of the sample shall be placed in a Soxhlet thimble, which shall then be placed in the Soxhlet extraction tube and extracted with washed re-distilled ether. At the end of three to four hours the thimble shall be removed from the Soxhlet tube, dried, and its contents finely ground in a small mortar previously rinsed with ether. The substance shall then be returned to the thimble, the mortar being washed out with ether, and extraction continued for another hour. After evaporation of the solvent the oil shall be di'ied at 100° C. and weighed. The oil shall be re- dissolved in ether, and any undissolved matter shall be weighed and deducted. In the case of samples con- taining saccharine matter, such as sugar -meals, the weighed portion in the Soxhlet thimble is to be washed twice with water and then dried previous to extraction. [The method given on pp. 9—11 will be found to give identical results in a very much shorter time, and since petroleum ether is used, no correction of the extracted fat is necessary.] Total Peotein (Albuminoids) (Official). The percentage of albuminoids shall be ascertained by multiplying the percentage of nitrogen bv 6'25. Feeding Stuffs 331 The determination of the nitrogen is carried out officially as follows : A weighed portion of the sample shall be transferred to a Kjeldahl digestion flask, 10 grm. of potassium sulphate and 25 c.c. of concentrated sulphuric acid shall be added, and the flask heated until a clear liquid, colourless or of a light straw colour, is obtained, A small crystal of copper sulphate or a globule of mercury may be added to the liquid in the digestion flask. The quantity of ammonia shall be determined by distillation into standard acid after liberation with alkali, and where mercury has been used, with the addition also of sodium or potassium sulphide solution. The materials used shall be examined as to their freedom from nitrogen by means of a control experiment carried out under similar conditions with the same quantities of the reagents which have been employed in the actual analysis, 1 grm. of pure sugar being used in place of the weighed portion of the sample. The quantity of standard acid used in the control experiment shall be deducted from the total quantity of acid found to have been neutralised in the distillation of the sample. The following is an exact description of the Kjeldahl- Gunning method as carried out by the authors : One grm. of the substance which has been sampled in the manner described above is weighed into a 500 c.c. Jena flask. In the case of fine meals and flours this is conveniently done in a small thin glass test-tube, one inch long and half an inch in diameter, the whole tube being dropped into the flask. Twenty-five c.c. of pure concentrated nitrogen-free sulphuric acid are added, taking care that the contents of the tube have been well shaken out before adding the acid. A small crystal of copper sulphate (about 0"2 grm.) is introduced, and the flask heated, in an almost horizontal position, over 332 Fatty Foods a very small flame until all spirting has ceased, when the flame is increased until the acid is seen to be slightly Fig. 34. distilling above the black mass. The flame is turned out, and after the mixture has sliglitly cooled, ]0 grm. of pure, finely powdered potassium sulphate are added Feeding Stuffs 383 and the heating continued, the flask being now placed on a guard-disc of asbestos board, iii which a hole has been cut of .sufficient size to prevent the flame from touching the sides of the flask above the liquid. The heating is now pushed as vigorously as possible, until the liquid has become quite clear, any black pai-ticles on the sides of the flask being washed down by carefully swinging the contents of the flask, the heating being continued until the contents are a clear bright blue. The heating is then further continued from this point for thirty minutes, as conversion into ammonia is not always complete under this time. The contents of the flask are allowed to cool,. 150 c.c. of water added, and the flask fitted with a double bored cork carrying a tap funnel and still head as shown in the illustration. The still head is then connected to a condenser, fitted with a bulb adapter, dipping into a flask in which have been placed 50 c.c. of N/10 sulphuric acid. A knife-point full of zinc dust is dropped into the tap funnel which is immediately filled with the necessary quantity of caustic soda solution, this being then allowed to run slowly into the flask, carrying with it the zinc dust. It is well to note that bubbling should commence in the standard acid immediately the alkali enters the flask, thus indicating that all connections are air-tight. The contents of the flask are brought rapidly to the boil, care being exercised that frothing does not take place as ebullition commences. The distillation is carried out as rapidly as possible and in general all ammonia will have distilled over in fifteen minutes. The adapter is then removed and a drop of the liquid falling from the condenser tested for freedom from ammonia with litmus paper. When free, the tap of the funnel is opened,; and the gas turned out. The adapter is rinsed, inside, and. out,, with a little water, and the dis- 334 Fatty Foods tillate titrated with N/lO alkali^ using methyl-orange or cochineal as indicator. A blank experiment is carried out in a similar manner, using 1 grm. of pure sugar in place of the sample. From the result so obtained the nitrogen is calculated as follows : The number of c.c. of alkali used plus those required for the blank are subtracted from the 50 c.c. of the standard acid used. The remainder, multiplied by 14 and by a factor of conversion of nitrogen into albu- minoids, gives the percentage of albuminoids in the sample if exactly 1 grm. has been taken. The following points should be noted : The con- denser should have a Jena glass inner tube. A globule of mercury may be used instead of the copper sulphate, in which case, after the acid digestion is finished, about 50 c.c. of water should be added and then 1 grm. of sodium hypophosphite, the solution being warmed (if necessary) to between 60° and 70° C, when all mer- cury will be deposited. After this 100 c.c. more water are added followed by the alkali and zinc dust as above described. If methyl orange solution be used it should first be tested for sensitiveness, as a distinct colour change must be observed when one drop of N/10 acid is added to 100 c.c. of distilled water containing two or three drops of the indicator. For many reasons cochineal is preferable, particularly as titrations may be carried out by gas-light. The alkali solution used for liberating the ammonia is made by dissolving 3 lb. of ground caustic soda insufficient water to fill a Winchester quart bottle. It is useless to attempt to filter this solution, but if it be left to stand for a week or so, it may be syphoned off easily into a bottle fitted with a syphon and guard tube, and may be drawn off as required. The quantity necessary Feeding Stuffs 336 for each experiment is ascertained by placing 5 o.c. of the digestion acid in a beaker and running in the alkali until neutralised. This amount multiplied by 5 will be the amount necessary to liberate the ammonia for distillation. Factors for Conversion of Nitrogen to Protein. Ordinary feeding stuffs . . . 6-25 Casein ...... 6'39 Cereals ...... 6'25 Si'AECH. (a) Polarimetric Method (Ewers). This is always determined on the fat-free-dry sample. If ready-formed sugars be not present no further pre- paration of the substance will be necessary. If sugar be present, stir 5 grm. in a beaker with 50 c.c. of cold water for one hour. Pour on to a filter and wash with 250 c.c. of water. The filter is then pierced and the contents washed into a 100 c.c. flask with hydrochloric acid (containing 1'124 per cent, by weight of HCl) con- tained in a wash bottle fitted with a fine jet. When 50 c.c. of the acid have been used the flask is placed in boiling water for fifteen minutes, during the first three of which the contents of- the flask are gently rotated. About 40 c.c. of water are then added and the contents cooled to 20° C, when 10 c.c. of a 4 per cent, solution of phosphotungstic acid are run on to the surface without mixing, and the flask filled to the mark. After mixing, the contents are filtered clear and polarised. 336 Fatty Foods The mean specific rotation of starch is taken by Ewers' as— ' '■ [a] ;» = 181-5, but if the Ventzke scale be used and a- 200 mm. tube the reading obtained is multiplied by 1-912, when the percentage of starch is obtained without further cal- culation, if 5 grm. of material are employed. Ewers gives the following specific rotations for various starclies : Wheat . . 182-7 Rye . 1810 Barley . . 181-5 Oats . 181-3 Rice . 185-9 Maize . . 184-5 Potato . 195-4 When no sugar is present the meal may introduced directly into the flask, in which case it is well shaken in the flask with 25 c.c. of the acid, and the sides are then washed down with a further 25 c.c. before heating in the water bath. (6) Starch by Diastane. About 5 grm. of the fat-free-dry material are washed on a filter with 150 c.c. of 10 per cent, alcohol and then with 10 c.c. of 95 per cent, alcohol. The residue on the filter is washed into a 250 c.c. graduated flask witb about 50 c.c. of water, and the flask immersed in boiling water with constant shaking for at least fifteen minutes. It is then cooled to 55° C. and 0-1 grm. of diastase (mixed with a little water) added, and the mixture maintained at 55°C. for an hour. It is then rapidly heated to boiling, again cooled to 55° C, and the diastase treatment repeated for one hour. The mixture is then cooled, made up to the mark and.filtei-ed; 200 c.c. of the filtrate are mixed with 20 c.c. of hydrochloric acid (sp. gr. 1-125) aiid feeding Stuffs 337 heated in boiling water under a reflux condenser for two and a half hours. The mixture is then cooled and sodium hydroxide solution added until the reaction is only faintly acid, and the volume then made up to 500 c.c. The dextrose is determined volumetrically by Fehling's method (p. 314 et seq.). The amount of dextrose found, Via. 35. multiplied by 0'9, will give the amount of starch, which is then calculated on the 5 grm. taken. Crude Fibre. ,One' grm. of the fat-free-dry material is weighed into; a 300 c.c. flask and 100 c.c. of boiling 1"25 per cent, sulphuric acid added. The flask is connected to a reflux condenser and brought rapidly to the boil, 22 338 Fatty Foods and the boiling continued for thirty minutes exactly. The contents of the flask are then rapidly filtered on a pump plate covered with a piece of fine well-washed linen cut one eighth of an inch larger than the dia- meter of the platBj the pump being in full operation before any liquid is poured on to the linen. The flask is rapidly washed out with boiling water, which is poui-ed over the material on the plate (about 100 c.c. of water usually suffices). The funnel is then placed in the mouth of the digestion flask, the linen lifted with forceps, the plate removed and the residue washed back into the flask with 100 c.c. of boiling 1'25 per cent, sodium hydroxide solution contained in a wash bottle. The flask is then again connected to the reflux condenser and the contents boiled for thirty minutes exactly. 'J'he contents are then filtered on a plate as before, using acircular tared filter-paper instead of the linen (the best quality ashless filter-paper must be used for this purpose). The whole of the contents of the flask are waslied on to the filter with boiling water, and the washing continued until the residue is free from alkali. Ten c.c. of 10 per cent, acetic acid are then poured over the residue on the filter, which is washed once or twice more with hot water, and then once with hot, strong, alcohol. The filter is carefully lifted off the plate, placed in its weighing bottle, any particles on the sides of the funnel being carefully removed and added to it, and the whole dried to constant weight. The filter is then ashed, and the weight of the residue less this ash is calculated in percentages on the original sample and is returned as " crude fibre." In order to prevent frothing, a gentle stream of air should be introduced through a small tube placed down the centre of the condenser to just inside the neck of the flask. A simple method of effecting this is shown in Fig. 35. Feeding Stuffs 339 The acid and alJcali uned in the above method must he accurately standardised hy titration, and the alhali should he free from carbonate. Ash. Two to three grm. of the material are burnt in a platinum dish at the lowest possible temperature. Where difficulty is experienced in obtaining a white ash (as, for example, in the case of maize germ cake), it is better to just char the substance, and to extract the char with water, which is poured through a small filter. The filter is then placed in the dish and the whole burnt to a white ash. The water extract is then added, evaporated to dryness on the water bath, and the whole again gently ignited. After weighing the total ash, it is dissolved, as far as possible, in hot dilute hydrochloric acid. The in- soluble residue is filtered off, ignited and weighed. This is returned as sand, but it will be understood that this is not the total silica, which, if required, may be estimated in the acid filtrate from the sand in the usual manner. The following determinations are less often required than the foregoing : Albuminoid Nitrogen (Soluble). (American official method.) Place 0"7 grm. of the substance in a beaker, add 100 c.c. of water, and heat to boiling (or if the substance be rich in starch heat on the water bath for ten minutes), and add so much of Stutzer's reagent as shall contain 0'5 grm. of cupric hydroxide ; stir thoroughly, allow to cool, and when cold, filter, wash the residue on the filter well with cold water^ 340 Fatty Foods and estimate the nitrogen by dropping the filter and contents into a flask, proceeding as described above under " Total Protein," it not being necessary in this case to add any copper. When examining oil-cake or other substances rich in alkaline phosphates, it will be necessary to add 1 or 2 c.c. of a concentrated solution of potash alum (free from ammonia) immediately before the addition of the cupric livdroxide. Stutzer's reagent is prepared as follows : 100 grm. of pure CuSO^ are dissolved in 5 litres of water, adding 2'5 c.c. of glycerol. A dilute solution of sodium hydroxide is then run in until the liquid is alkaline, the precipitate is allowed to settle, the super- natant liquid poured oif and the precipitate rubbed up with water containing 5 c.c. of glycerol per litre and washed by decantation until free from alkali. The precipitate so obtained is rubbed up with water con- taining 10 per cent of glycerol into a thin gelatinous mass which can be just measured with a pipette. The cupric hydroxide per c.c. of the preparation is then determined. Amide Niteoqen. A weighed quantity, about 1 grm., is boiled with 100 c.c. of 5 per cent, hydrochloric acid for thirty minutes under a reflux condenser. The solution is then just neutralised to methyl orange with a solution of sodium carbonate, and, after the addition of an excess of magnesium oxide, distilled into standard acid in the ordinary way. The ammonia so obtained is usually calculated as aspurayin C3H3 (NHg) (CONHj) COOH, and returned as amido compounds. 1 molecule NHj = 1 molecule of asparagin. Feeding Stuffs 341 Phosphoeic Acid and Potash. 10 grm. of the fat-free-dry substance are weighed into a 500 c.c. Jena flask and a mixture of 30 c.c. of concentrated nitric acid and 5 c.c. of concentrated hydrochloric acid added. (If the mixture is not then liquid the quantity of acid must be increased.) The flask is gently heated over a small flame, which is gradually increased until a clear liquid is obtained. The solution is cooled, cautiously diluted with water and made up to 250 c.c. In this solution phosphoric acid and potash are determined. Phosphoric acid (Pemberton's method, A.O.A.C. modification). — 25 c.c. of the acid solution (unless the percentage of phosphoric acid is expected to be more than 5, in which case not more than 10 c.c. should be used) are taken. To this are added 5 to 10 c.c. of nitric acid, and then ammonium hydroxide until nearly neutral, and the whole diluted to 70 or 80 c.c. and heated in a water-bath to 60°— 65° C. Twenty to 25 c.c. of freshly filtered molybdate solution are added and the mixture stirred and allowed to stand for fifteen minutes. The liquid is then rapidly poured ofE through a Gooch filter having as thin a layer of asbestos as will be sufficient to retain the precipitate, and the precipitate in the beaker washed twice with water (25 c.c), stirring well and allowing to settle before pouring off the washings through the filter. The precipitate is now washed on to the Gooch with cold water and the washing continued till 20 c.c. of the washings do not decolourise the colour produced by 1 c.c. of the standard sodium hydroxide solution with phenol- phthalein. The asbestos wad and the precipitate are carefully picked out and dropped into a flask, the 342 Fatty Foods Gooch placed in a small funnel in tlie month of the flask and washed with 10 c.c. of the standard sodium hydroxide solution to remove any particles of the precipitate left in the Gooch. The Gooch and funnel are then thoroughly washed into the flask with water, the flask closed with an india-rubber stopper, and the contents shaken with a rotary motion until the asbestos is thoroughly broken up. The standard alkali is then added until the yellow precipitate is completely dissolved, about 5 c.c. being then added in excess of the necessary quantity. The .contents of the flask are thoroughly mixed, and after the addition of a few drops of phenolphthalein, titrated back with the standard acid. The total number of c.c. of standard sodium hydroxide used to dissolve the precipitate, less the number of c.c. of standard acid required to titrate back X O'OOl = grm. of PnOg in the number of c.c. of the original solution used. The reagents required are made as follows : MoLybdate solution. — 100 grm. of molybdic acid dis- solved in a mixture of 144 c.c. of ammonium hydroxide (sp. gr. 0'90) and 271 c.c. of water, are added slowly and with constant stirring to a mixture of 489 c.c. of nitric acid (sp. gr. 1"42) and 1148 c.c. of water. The mixture is kept in a warm place for several dnys and decanted for use. For the above method 100 c.c. of this solution are mixed with 5 c.c. of nitric acid (sp. gr. 1"42) and filtered just before use. Standard sodium hydroxide solution. — Dilute 323'8 c.c. of N/NaOH solution to 1000 c.c. 1 c.c. of this sol ution=: O'OOl grm. of P20g. In order to avoid the presence of carbonate in the standard sodium hydroxide solution, it is advisable to place about twice the required weight of solid caustic soda in a sepai-ator and to add just sufficient water to Feeding Stuffs 343 wash the surface of the sticks. This water is rapidly run off and the process repeated. By this means the carbonate, which is on the outside, is washed away. Sufficient water is then added to dissolve the residue in the separator, and after solution has been effected the liquid is run out into a graduated flask, made up to a definite bulk, and its strength ascertained by titration against standard acid. It can then be adjusted by dilution to the required strength. Standard acid solution.. — This solution is made exactly equivalent to the above alkali solution, using phenolphthaleiii. The acid is to be run into the alkali in setting the solution. Any mineral acid may be used. Potash. 100 c.c. of the solution of the substance, made as directed under phosphoric acid, are evaporated to dry- ness, taken up with a little water and hydrochloric acid, and again evaporated to dryness, any silica which separates being filtered off, and washed till free from chlorides. The combined filtrate and washings are made alkaline with pure sodium carbonate, and boiled till free from ammonia. The solution is washed into a beaker and made up to 250 c.c. approximately with water, 20 c.c. of hydrochloric acid cautiously added and the whole brought to the boil. A solution of barium chloride (containing 127 grm. BaCl2.2H20 and 125 c.c. of pure concentrated hydrochloric acid per litre) is dropped in with constant stirring until all sulphates are precipitated. The solution is boiled for a few minutes, cooled, filtered and evaporated to dryness. The residue is taken up with 20 c.c. of hot water and sufficient perchloric acid added to combine with one and a half times the bases present (5 c.c. of acid sp. 344 Fatty Foods gr. 1-5 will probably be sufficient). The mixture is evaporated - with stirring to a syrupy consistency, dis- solved in a little hot water, 5 c.c. of perchloric acid added and evaporated till white fumes of the acid are given off strongly, again diluted, and 2 c.c. of perchloric acid added and again evaporated. The residue while still warm, is well stirred with 20 c.c. of acid alcohol (97 per cent, alcohol, containing 0'2 per cent, of per- chloric acid). The undissolved portion is allowed to settle, and the liquid decanted through a tared Gooch crucible with an asbestos pad, the residue in the dish being warmed with another 20 c.c. of acid alcohol, thoroughly disintegrated, and the liquid again decanted, this treatment being once more repeated. The residue in the dish is then dissolved in 10 c.c. of hot water, a little perchloric acid added, and the whole again evaporated until it fumes strongly. The residue is mixed with 5 c.o. of the acid alcohol and the mass transferred to the Gooch crucible with further quantities of acid alcohol. (Not more than 50 c.c. of the acid alcohol should be required for all the above washings.) The Gooch crucible is then rinsed with 1 or 2 c.c. of 97 per cent, alcohol anil dried at 130° C. to constant weight. Weight of KOIO^ x 0-3402 = KgO in the original 100 c.c. used. In using the above process, which is really much more rapid than the description would imply, great care should be exercised in thoroughly disintegrating the mass of perchlorates with the acid alcohol, in order that those other than potassium perchloi-ate should be properly dissolved out. The process is quite as accurate as the platinum method, and pure perchloric acid is prepared by Kahlbaum at a price which brings this process much below the cost of the platinum method. The above method presupposes the use of perchloric acid of sp. gr. 1"5 to TB. The acid is perfectly safe and non-explosive. Milk 345 CHAPTER IX. MILK. It is not intended in this cliapter to enter into the full details of the analysis of milk and cream, etc., but to give such methods and tests as will probably be useful to the margarine manufacturer in controlling the milk employed by him. The control of milk will include an ordinary analysis for chemical composition, and for the presence of pre- servatives which would be deleterious to the sub- sequent "souring," tests fbr any previous pasteuri- sation, and also for the control of pasteurisation employed in actual manufacture, and some simple methods of judging the bacteriological condition of the milk supplied, in order to avoid as far as possible the development of " taints " in the soured milk or in the margarine itself. Analysis or Milk. Total Solids. 2 '5 grm. of the samples are weighed out into glazed porcelain dishes about 2| in. in diameter, flat-bottomed, and about i— | in. in depth. One c.c. of acetone is added to each, and the dishes, placed on a rapidly 346 Fatty Foods . boiling water bath for thirty minutes, and then placed in the baih for forty-five to sixty minutes longer, the outsides being first carefully wiped with a towel. They are then weighed in the following manner : To avoid the use of a desiccator, the dishes are placed in the oven in order of ascending weights. The first is taken out and placed on a wire tray beside the balance. While still gently warm the dish is placed on the balance and rapidly counterpoised. It is then re- placed on the wire tray and the second dish removed from the oven. The rider is now shifted up 1—2 mgrm., and immediately the first dish is just cool (as ascertained by touching against the cheek) it is replaced on the balance, and the needle allowed to swing. If the divisions on the scale corresponding to a difference of 1 mgrm. are known (2 is a suitable number in order to get rapid swings) the weight is easily ascertained to the nearest milligramme without further alteration of the rider. By this time the second dish will be sufficiently cool to deal with, and so on. The weights of the dishes themselves, when empty, are ascertained in an exactly similar manner, and need re-determining every three to four weeks. Using 2'5 grm. the actual weight of milk solids multiplied by 40 gives the percentage of total solids in the sample. Fat. For routine purposes the Gerber method is the sim- plest and quickest. The necessary butyrometers and centrifuge may now be obtained anywhere, and the chemicals employed are simple and as easily procured. Reagents. — Sulphuric acid (commercial) : Sp. sv. 1'820— 1 825. Acids of a higher gravity may be diluted, but the gravity must not bo less than 1-820. Milk 347 Arnyl alcohol. — The fraction boiling between 124° and 130° 0. and having a specific gravity of 0'815— 0'818 is employed. It should be practically colourless. Making the test. — Ten c.c. of the acid are placed in the butyrometer and 11 c.c. of the sample run in on to the acid down the side of the butyrometer^ so that the acid and milk do not mix. One c.c. of the alcohol are run in on to the milk, and the mouth of the butyrometer closed by screwing in a tight-fitting india-rubber stopper. The tube is then violently shaken up and down in the direction of its length till all the milk is dissolved, and then placed in a bath of water at 160° F. for five to ten minutes (not more than twenty). It is then placed in the centrifuge and whirled for five minutes at about 2000 revolutions per minute. The fat, which has collected in the graduated neck, is read off, the readings being in direct percentages of fat, the larger divisions being units. The fat layer should be quite clear, and there should be no plug of undissolved material below it, which latter indicates insufiicient shaking and heating. It is advisable to make a blank experiment with the acid and alcohol, using water instead of milk, in order to see that no insoluble matter is obtained from them. It is also very advantageous to check the readings of the tubes against some standard method of fat analysis. Such a method is the — Gottlieb Process. For this the following reagents are required : (1) Strong alcohol (95 per cent, by volume). (2) A solution of ammonia (made by mixing 100 c.c. of 0-880 ammonia with 100 c.c. of water). (3) Methylated ether, sp. gr. 0-720. 348 Fatty Foods (4) Petroleum ether, distilling below 40° C, and leaving no residue. in Apparatus. — A strong glass tube about | diameter and holding 50 c.c. fitted vpith an indiarubber Fi0. 36. stopper. Instead of this tube, the appai-atus shown in Fig. 36 will be found of grftat advantage (Eichloff and Grimmer) . Milk 349 In the tube or apparatus are placed 5 grm. of milk, 0"5 c.c. of ammonia solution added, and the contents mixed by swinging the tube in a circular manner. Then 5 c.c. of alcohol are added, the whole again mixed, 12'5 c.c. of methylated ether run in and the stopper inserted. The tube is inverted three times, after which 12"5 c.c. of petroleum ether are run in and the tube again inverted three times (any violent shaking must be avoided). The tube is allowed to stand and settle out ; the ether solution is then blown off into a weighed flask, by means of a wash bottle arrange- ment (shown in place in Fig. 36), down to ^ in. from the aqueous layer. A. fresh quantity of petro- leum ether (or mixed ethers recovered from other tests) is run into the tube, and the contents mixed once by inversion. This ether is then blown off and the process repeated once more. The ether is then distilled off from the fat and the flask dried to constant weight in the water oven. Acidity. The method of estimating the acidity in milk com- monly employed in this country consists in adding N/10 NaOH to the milk until a faint pink colour is given in the presence of phenolphthalein. The test," in the case of milk, is carried out conveniently as follows : Twenty c.c. of the milk are placed in each of two test- tubes (6 in. by 1 in.), and to one of these about five or six drops of 1 per cent, phenolphthalein solution in 50 per cent, alcohol are added, and then N/10 NaOH, about four drops at a time, until, after gently shaking, a very faint pink colour is noted in the milk as compared with the colour of the control. Then " the number of c.c. of N/10 alkali required to produce this result multiplied by 5 gives the so-called degrees of acidity." 350 Fatty Foods 3 ^ ^ ^ Good fresh milk from normal cows in good condition usually has an acidity of 15 to 18, which will not in- crease perceptibly during the next twenty-four hours if the milk has been cooled immediately aftei; milking and subsequently kept at a reasonably low temperature. Milk with 20 to 22 degrees of acidity has a slight sour taste, quite perceptible to the educated palate. When 24 degrees is reached the milk usually coagulates on boiling. When estimating the acidity in milk which has been soured for the purposes of churning, it is pre- ferable to measure out the required quantity into a porcelain dish or a cup, and to carry out the titration in this, rather than as described above, the acidity in such cases being much higher than is the case with fresh milk. In the case of milk soured for churning, it is more usual to express the acidity in percentages of lactic acid, taking 1 c.c. of N/10 alkali to equal 0009 grm. of lactic acid. For simplicity in this case 9 grm. of milk are pipetted into the dish (from a special pipette) when c.c. of N/10 alkali „ , , . . t = percentage ot lactic acid. The following very useful method of titrating the acidity of milk has been devised by Richmond (Analyst, 1912, vol. xxxvii, p. 168), and enables acidi- ties to be determined correctly by artificial light, and obviates any difficulties arising from the colour-blindness of the operator, as only shades of colour are compared and not the colours themselves. It is carried out as follows : To the milk in the control tube is added one drop of a O'Ol per cent, solution of rosaniline acetate in 96 per cent, alcohol, and to the other tube 1 c.c. of 0"5 per cent, phenolphthalein solution. The alkali is run into the latter until the depth of colour in it, on shaking, is Milk 351 identical with that of the control. The figure so obtained is about 1'5 degrees higher than by the usual method, and should be distinguished as "Acidity by the Rosaniline [B.S.) Stanilard." Preservatives. Boron compouudg.- — About 1 c.c. of the milk are placed in a flat dish similar to that employed for total solids, and thoroughly mixed with one drop of strong hydrochloric acid and ' then with five to six drops of saturated alcoholic solution of turmeric. The dish is then placed on a rapidly boiling water bath, and the mixture evaporated to dryness. In the presence of as little as 0"02 per cent, of boric acid the edge of the residue will assume a salvwn-'pink colour immediately it is dry, while, if no boron compounds be present, the residue will be a dirty yellow. Formalin. — The presence of formalin or its com- pounds will be detected during the fat estimation by the G-erber process. In the presence of formalin the junction of the milk and acid before shaking will assume a violet colour, pure milk giving only a brownish- yellow tinge. It is to be noted that excessive quan- tities of formalin do not give the purple colour, but in such cases, apart from the taste of the milk, its presence would be suspected on account of the great diflSculty with which the curd dissolves in the Gerber tube when shaken. Salicylic and benzoic acids. — Not less than 100 c.c. of the milk are heated in boiling water after the addi- tion of 4 c.c. of 10 per cent, sodium carbonate solution per 100 c.c. of milk used. When the mixture has attained the temperature of the water, a volume of 20 per cent, calcium chloride solution equal to that of 362 Fatty Foods the sodium carbonate solution employed is run in, when precipitation of the casein should occur. The contents of the ilask are rapidly cooled and filtered, and the test finished exactly as described on p. Ill for the alkaline extract of butter or margarine. Pasteurisation. It will often be necessary to ascertain whether the supply of milk has been pasteurised previous to delivery, and in any case it will be of importance to control the efficiency of the pasteurising process employed in the factory. Two forms of pasteurisation are commonly employed : (1) That in which the milk passes continuously through a steam- jacketed vessel,, being caused to pass over the heated surface in a thin film by means of a paddle, the milk then passing out on to a cooler. 'J'his method is usually termed " flash " pasteurisation. (2) That in which the milk is run into a jacketed vessel, and, while being steadily stirred, brought to the necessary temperature by carefully admitting steam to the jacket, and subsequently held in the pas- teuriser at the required temperature with constant stirring for the necessary time. In the opinion of the authors the second method is greatly superior in every way to the former for milk which is to be used for souring, as not only can the necessary temperature be maintained throughout the bulk of the milk, but a much lower temperature can be efficiently employed, with great gain in the flavour of the product. The efficiency of the operation also can be much better controlled by chemical tests. Of the methods which have been suggested for the detechion or control of pasteurisation (and they are Milk 353 legion)/ the two following are recommended- as being the most practical and efficient ; (1) Benzidmie acetate test. — This test, which can he rapidly carried out^ is satisfactory for determining wjiether the milk has been heated to a certain tem- perature for a definite time, but while quicker, is not so sensitive as the next method described. It is carried out in the following manner, and in compara- tive work it is necessary to adhere to the quantities given : Five o.c. of the milk are placed in a test-tube, 1 c.c. of a 4 per cent, alcoholic solution of benzidene added, and the whole shaken ; one drop of pure acetic acid is then added- and the whole again shaken. Three c.c. of hydrogen peroxide solution (20 volumes) are then allowed to run carefully on to the surface of the mixtui-e in the test-tube. The following results are obtained : ao„«ia Temperature of Time of -Roo^if aampie. pasteurisation. pasteurisation. ±tesuic. (1) TJnheated Deep azure blue. (2) Pasteurised . 151" P. . Up to 45 min. Slightly Jess than (3) „ 158 to 160° P. . 30 min. As (2), , (4) „ „ 45 min. . Paint blue. (5) „ 162 to 163° P. . 15 min, . Very faint blue. (6) „ „ . 30 min. . Colourless. (7) „ 176° P. 1 to 2 min. . Colourless. The benzidene solution keeps indefinitely in the dark. (2) Para-phenylenediamine-guaiacol test (Hesse and Kooper). — This test, though not so rapid oE application as the former, permits (in the following slightly modi- fied form of the original test) of very exact estimations of the degree of pasteurisation. Reagents. ^-{1) Para-phenylenediamine hydrochloride 2 grm. dissolved in 75 c.c. of 95 per cent, alcohol. (2) Guaiacol (crystallised) 1 grm. dissolved in 15 c.c. of water and mixed with 60 c.c. of alcohol. 23 354 Fatty Foods ^^ y Equal volumes of these solutions arfe mixed fOr use. Method nf carrying out the test. — Forty c.c. of the milk are mixed with 6 c.c. of basic lead acetate solution (see p. 310), well shaken and allowed to stand for five minutes. The mixture is then filtered and 5 c.c. are placed in a 50 c.c. Nessler glass, and one drop of 20 volume hydrogen peroxide and two drops of the mixed reagent added. After allowing to stand five minutes water is added to the 50 c.c. mark and the whole filtered (if necessary). Working iti this way the colour given by the test of the pasteurised milk is readily matched against sueh a volume of the test of the original milk as will give the same colour. The following table has been compiled by the authors. The number of c.c. given in the various columns are the number of c.c. of the test of the original unpasteurised milk, which gave the same depth of colour as the full 50 c.c. obtained from the pasteurised milk; All the pasteurisation tests were carried out by holding the milk in bulk at the tempera- ture named in an experimental pasteuriser. It must be clearly understood that the figures, though pro- bably strictly comparable among themselves, may not be absolutely correct for all actual conditions. A few tests carried out on the actual plant will show what corrections it will be necessary to apply. Temperature of Duration of pasteurisation, pasteurisation. lomin. 30min. 45min. 151° F. . . 45 CO. . 41 c.c. 44 c.c. 156° to 157° F. . 39 „ 40 .35 „ ] 58° to 160° F. . 21 „ . 12 „ . 7 162° to 163° F. . 4-5C.C. . 0„ . 0„ In the above experiments, samples which were with- drawn from the pasteuriser after the specified time, were cooled at once, and the further destruction of the enzymes prevented, which would not be the case in Milk 355 practice. In actual practice the time which elapses while the pasteuriser is emptying on to the cooler must be taken into account. ' TPhe progressive destruction of enzymes as the temperature rises is well illustrated by the following figures : ' Milli was steadily heated until it had reached 150° P., when samples were withdrawn as various higher tem- peratures were attained, the times being noted at which these temperatures were reached, starting from the time at which the milk reached 150° F. Temperature. Minutes elapsed after 150° was reached. Deg:ree of pasteursatiou. 161° F. 5 (39i c.c.) 21 per cent. 163° F. 9 (33 J e.c.) 33 167° P. 15 (20 CO.) 60 167" F. . maintained for fifteen minutes (0 CO.) 100 As the maintenance of a temperature of 165° F. , for twenty to thirty minutes is amply sufficient for pasteur- isation, it will be seen from the above figures how easily this can be controlled, and the variety of ways in which the same results can be produced, taking des- truction of peroxydase and destruction of vegetative forms of bacteria as occurring simultaneously. G-ENEEAL Tests foe Cleanliness and Bacteeiolqgical : Condition. For this purpose also methods have been produced in an increasing stream. By far the greater number are of very doubtful utility and are based oil principles of which the fundamental characters are not yet clearly understood. The following yield information of a definite and useful character and are simple in appli- cation. Key to Plate Y~ Fia. 1. Lactic acid-forming organisms. X 1000. FiQ. 2. Cellular elements in milk sediment. X 1000. PLATE V. ^«^ Fici. 1. / 'i' Fia. 2. 2*0 /ac-e page 356. Milk 357 erinary surgeon. A certain number of these cells are always present in milk, and it is only when an excessive number are present that any deductions as to disease can be made, and in some cases very large numbers of these cells are found in milk from quite healthy cows. The presence, therefore, of even large numbers only justifies a veterinary inspection. (2) Fermentation Test. For the purpose of this test several large test-tubes about 6 in. by 1^ in. and fitted with loose metal covers are required. These tubes, with the covers on, are cleaned and sterilised, either in an ordinary hot-air steriliser, or if this be not available, by thorough rinsing with boiling water. Method of carrying out the test. — In each tube are placed 50 or 100 c.c. of the various milks to be examined, for preference as soon after arrival of the milk as is convenient. The tubes, being closed with the covers, are placed in a water bath or incubator which can be maintained at between 38° and 40° C. After twelve hours' incubation the tubes are removed and examined. Good, fresh, well-cooled and cleanly- handled milk will as a rule not have coagulated in this time, or if it has done so, the milk will have set to a uniform gelatinous curd, free from all bubbles, and from any separation of whey between the cream layer and the milk. The curd will have no unpleasant smell, and should possess a sharp, agreeable, sour taste, leaving no unpleasant after-flavour. It is by no means un- common, however, for a considerable contraction of the curd to take place at the sides of the tube, and conse- quently the appearance of whey even in the case of good milk. In such cases, however, the whey is' always 358 Fatty Foods very turbid and there are no gas^bubbles whpn the tube is gently shaken. The appearance is, however, quite distinct from that mentioned below in which digestion of the casein has taken place, under which circum- stances the whey is clear and at the bottom of the tube. If the mijk has not been so cleanly handled, the curd will be seen to be streaked with bubbles, and a small quantity of whey will probably be noticed between the cream and the curd. In these cases the milk has often an unpleasant smell, and the acidity may be dis- tinctly bitter. Under worse conditions there may be considerable digestion of the casein, with separation of whey at the bottom of the tube, and the cream may be filled with bubbles of gas. Any condition other than the first one described must be considered as pre- judicial to proper souring of the milk for margarine purposes. It must be clearly understood that pas- teurisation is no panacea for uncleanly dairying, as not only is the souring proper prejudiced, but taints, already developed in the milk, will persist and often increase in the finished product. It may be noted that milk which has been pasteurised may produce in twenty-four to thirty-six hours a coagulum which afterwards slowly digests, and is usually alkaline (the curd being a rennet curd), or the reaction described .under the next heading may take place. A small portion of the curd obtained in any case may be spread on a slide, stained, and examined as described in the last section, when it will be found that in the curd of good milk practically the only organisms present are small diplococci and short streptococci usually not exceeding four to six members, though longer chains are frequently found. This appearance is shown in Plate V, fig. 1. When the milk has been dirtily handled the bac- terial flora is usually much more diverse, and " bacil- Milk 359 lary " (rod-shaped) forms are generally to be seen. The curd of pasteurised milk is usually quite free from " coccoid " forms and probably only rods will be present, generally rather large in character. This fermentation test probably furnishes more useful informat)ion than any other, ar[d its extreme simplicity renders it very appropriate for control purposes. It has been considerably elaborated by some investi- gators and some rather fine distinctions drawn as to the results obtained, but for these a manual dealing with milk must be consulted. (3) The Enteritidis Sporogenes Test. This test furnishes, in the opinion of the authors, considerable evidence as to the amount of cow dung which has been allowed to get into the milk, though the milk may have been subsequently cleaned by some efficient type of filtration. The test is simple in char- acter, requires no bacteriological apparatus, and further, as the organisms concerned in the reaction do not in- crease in number during transit, the findings are true for the milk as dairied. Method of carrying out the test. — A number of test- tubes 6 by I in. are plugged with cotton- wool and steri- lised in" a hot-air steriliser at 150° C. for three quarters of an hour, or in an oven until the cotton-wool begins to. turn brown. With a lO'c.c. pipette, which has been sterilised by rinsing out with boiling water, or by thorough heating in the flame of a bunsen burner, .10 c.c. of each milk are placed in the sterilised tubes, the cotton-wool plugs being singed before removal, and so Jield that the lower ends do not touch anything. About f in. of melted vaseline is now poured on to the surface of, the milk in each tube, the plugs being then replaced. The tubes are now immersed in water at 360 Fatty Foods 80° C. and kept at this temperature for twenty minutes. They are then cooled and maintained at 38° to 40 C. for twenty-four hours. In the case of milk which has been dirtily handled the milk will have undergone a profound change. Very considerable digestion of the casein will have taken place^ resulting 'in the production of whey in which the remnants of curd, torn and disintegrated with gas, float. The vaseline plug is usually forced, by the excessive gas evolution, more or less completely up the tube, and the culture, when opened, has a distinct smell of butyric acid. The characteristic appearance of the curd is well seen in Plate VI, fig. 1, and if some of this curd be stained and examined microscopically it will be found to contain stDut bacilli with rounded ends, as seen in Fig. 2. Milk which has been cleanly handled will either undergo no change or will set to a solid curd, which will gradually contract with the separation of whey, but without any gas-formation. Cream and Separated Milk. These will usually appear in. margarine factories as bye-products, and -practically the only examination of them that will be necessary will be an estimation of the fat. For this purpose the Gerber method is satisfactory in both cases. The estimation of the fat in cream. — About 1 grm. is weighed out in a small funnel, having a small rod inserted in its tube. The cream is then washed into the ordinary Gerber bottle, with two quantities of 5 c.c. of warm water. The tube and contents are then cooled and the usual 10 c.c. of sulphuric acid and 1 c.c. of amyl alcohol added, the test being finished as in the case of milk. If the tube be read on removal from the rotator without re-warming it will be found that : Key to Plate VI. Fio. 1. Characteristic appearance of curd in Enteritidis sporogenes Test. Fig. 2. B. ent&ritidis sporogenes from milk culture. X 1000. Fia. 1. PLATE VI. Fie. 2. C"^"'/ To face page 360. Separated Milk 361 Fat per cent. = Reading x factor Weight of cream used. The value of the factor, as determined by the authors, is as follows . Eor 55 per cent, fat = 11-07 „ 45 „ „ = 11-23 » 20 „ „ = 11-41 For rapid internal control a simple estimation of the total solids of the cream as suggested by Mats WeibuU is quite satisfactory. The following formula is used : Fat per cent. = 1-1 X T — 9-5, where T is the estimated total solids. The total solids are conveniently estimated by care- fully boiling off the water in a flat porcelain dish in a manner exactly similar to the method of Patrick described under butter, or by heating on the water bath as -for the total solids in milk, until constant in weight. The estimation of Fat in Separated Milk. For the control of separators the residual fat in the separated milk may be estimated with quite sufKcient accuracy in the ordinary Gerber tube. In this case, however, the tube, after rotation, is placed in water at 160° F. for two or three minutes and again rotated for five minutes, again placed in warm water and read warm — the reading being taken from the bottom of the fat layer to a point half, way between the lowest point of curvature and the tip of the horns of the meniscus. The results obtained by the Gerber process are usually lower than those obtained by the use of such a rrjethod as the Gottlieb or other extraction process, but as any fat which will not rise in the Gerber tube will in all probability not be thrown out by the separator, it is scarcely fair to judge the efficiency of separators by extraction methods. 362 Fatty Foods Appendix D (to p. 309). Chocolate. Typical example of method of calculation and intei-pretation of the results of analysis of a sample of cheap chocolate. > ' ' ' Analytical data obtained. [^ Percentages on dried sample. Fat .• . 29-00 Mtrpgen , .... 0^78 (N.) Fibre, , . 1-60 (F.) Starch (by diastase method) 6'80 (S.) Gold water extract " - 5224 Oane sugar 48' 00 Cold water extract less cane sugar . 4'24 (T.) Levigation residue . . . 2-05 Microscopical examination revealed maize starch and a fair pro- portion of cocoa shell. The percentages of fat-free-dry cocoa matter (X.) and added starch (F.) are calculated by the formulae on page 309, substi- tuting the above analytical figui-es where indicated, after allowing for cane sugar and fat as follows : Since 100 parts of the dry sample contain 77 per cent, of fat + cgir^e sugar, the figures for (N), (F), (S), and (T) must be multi- plied by ^-^ to obtain the percentage in cocoa matter + starch. The values so obtained are : (]Sr=8-39) ■(F = 6-96) (S = 29-55) (T=]8-43). By substitution :' • , ^°°(SS^5S)=^ = '6percent. /^^^^'MSt5°)=^=22percent. assuming in the above an approximate value of 10 for Sc and taking the value 0-8 for C from the table on p. 308. Employing the other formulae, in which no assumption is made for So, the following values are obtained : 100 ( T -3-73 N - -89 FN ^y--1U \ - 3-73 Ns I ^- ^^■ This value for Y has, of course, no ^raciicaZ meaning, but when substituted in the following formula, gives : 100 T - OY ^ „, =2 = -X = 81 per cent. whence the real value of r= (100 — 81) = 19 per cent. v^ Appendix D 363 [A preferable formula for Y is : \3Sx + C -15nJ in whioh Sy =: 75 for wheat and barley flours, 95 for arrowroot and maize starch, and 89 for rice flour.] As these values are probably nearer the truth than the more approximate ones obtained above, it will be assumed that there is 80 per cent, cocoa matter (nib + shell) and 20 per cent, of maize starch, which when calculated to the original dry sample (by multiplying by tbh). gi^^es : Cocoa matter . 18-4 per cent, and Maize starch . 4-6 per cent. The proportion of nib and shell, are now calculated for 100 parts of cocoa matter, as follows : Prom the nitrogen* . . 18-4 0-78 : : 100 4-23. On reference to the curves on pp. 305-6 it will be seen that the value 4'23 oori-esponds to : Shell . . 33 per cent, and Nib . . 67 per cent. In a similar manner from the fibre the values obtained are : Shell . 26 per cent, and Nib . 74 per cent, and fi'om the levigation curve : Shell 30 per cent, and Nib . 70 per cent. It may therefore be assumed that the true composition approximates to : Shell . 30 per cent, and Nib 70 per cent, which values, when calculated back to the original di-y sample, give : ' Shell 5"5 per cent, and Nib 12-9 per cent. The composition of the dry salmple therefore is : Oane sugar ' 480 Fat . , . 29-0 Maize starch . 4'5 Cocoa nib 130 Cocoa shell i . . 5'6 100-0 * If large quantities of starch lie present or if it "be wheat, barley or rice starch, the nitrogen which they contain jnust he allowed for. In the present case, since maize starch contains only 0'14 per cent, of nitrogen {p. 308), the amount introduced by 5 per cent, (being only 0'007 per cent.) may be neglected. 364 Fatty Foods BOTANICAL INDEX. Aleurites cordata, 250 — Fordii, 250 — moluccana, 250 — triloba, 249 Arachis hypogaea, 202 Attalea Cohune, 157 — funifera, 157 Bassia>butyracea, 183, 188 — latifolia, 183, 192 — longifolia, 183, 194 — Parkii, 183, 184 — toxisperma, 190 Bombax malabarioum, 220 Brassica oampestris, var. napus, 209, 210 var. rapa, 209 Butyrosperum Parkii, 183, 184 Cannabis sativa, 280 Carthamus tinotorius, 245 Cocos nuoifera, 148 Croton Bebiferum, 169 Cyperus esculentus, 278 Blseis guineensis, 153, 173 Eriodendron anfraotuosnm, 220 Garoinia indioa, 181 Grlyoine hispida, 226 G-ossypium, 216 Guizotia abyssinioa, 247 Heliauthus annuus, 238 Heyea brasiliensis, 252 Hopea, 166 Hydnocarpus anthelminthioa, 266, 267 — venenata, 268 — Wightiana, 266, 268 Isoptera tjorneensis, 166 Juglans regia, 256 Linum usitatissimum, 236 Mafotireira oleifera, 176 Mimusops Djave, 183, 190 Myristiea fragrans, 270 — guatemalensis, 270 — surinamensis, 270 Olea europaea, 198 Papaver somniferum, 241 Perilla ocymoides, 259 Prunus amygdalus, var. dulcis, 206 var. amara, 206 Bhus vernicifera, 160 — sucoedanea, 160 Bicinus communis, 262 Sapium sebiferum, 169 Scyphocephalium Ochocoa, 270 Sesamum indioum, 228 Shorea aptera, 166 — ghysbertiana, 166 — robusta, 166 — stenoptera, 166 StUlingia sebifera, 169 Taraktogenos Kurzii, 266, 268 Theobroma Cacao, 164 Trichilia emetica, 176 Vateria indioa, 179 Zea Mays, 230 Index 365 INDEX Acetyl value, determination of 28 true, 29 inferences from, 77 Acid value, determination of, 41 inferences from, 77 Albuminoids, estimation of, 330 Almond oil group, 206 Amide nitrogen, estimation of, 340 Analytical methods, general, 7 Analysis, object of, 2 — ■ preparation of sample for, 8 Animal fats, 94 Annatto, detection of, 113 Apricot kernel oil, 207 Arachidic acid, 34 Arachis oil, 34, 202 Arsenic, detection and estimation of, 71 Ave Lallemeut, barium method of, 121 Azo-dyes, detection of, 113 Bambuk butter, 184 Baryta value, 121 Bassia group of fats, 183 Baudouin's test for sesame oil, 53 inferences from, 79 Bean oil, 226 Becchi's test, 54 inferences from, 79 Beef fat, 94 microscopical examination of, 56 analytical determinations, 95 — " stearine," 95 Bellier's test for arachis oil, 30 inferences from, 77 for poppy seed oil, 256 Benzoates, detection of, in butter and margarine, 110 in milk, 351 Benzidene acetate test, 353 Biscuits, oils and fats employed in manufacture of, 72 — oil extracted from, 87 Bjorklund's test, 58 Borneo tallow, 166 Boron compounds, detection and estimation of, in butter and margarine, 108 detection and estimation of, in milk, 351 BUTTEK — Bambuk, 184 Cocoa, 164 Djave, 190 Galam, 184 " Green," 59 lUipe, 166, 183, 194 Karite, 184 Kokum, 181 Njave. See Djave Phulwara. See Butyraceafat Butter, introductory, 104 Bdttbb, analysis of — Colouring matters, 112 Curd, 107 Examination of fat, 114 Fat, 106 Foam test, 125 Preservatives, 108 Salt, 107 Valenta test, 123 Water, 106 Butter, dried milk in, 107 — interpretation of analytical results, 126 — microscopical appearance of, 125 366 Fatty Foods Butter, sampling of, 105 | — renovated, 107, 125 Butyrabea fat, 188 Batyro-refraotometer, Zeiss, 17 Cakes, oils and fats employed in manufacture of, 72 Candle nut oil, 249 Cane sugar, 298, 310, 312 Caramels, oils and fats employed in manufacture of, 73 Carthamus oil, 245 Casein, estimation of, in milk chocolate, 317 Castor oil, 262 Ceara rubber seed, oil, 255 Chaulmugra oil, 266 Chinese vegetable tallow, 169 Chocolate, analysis of, 307 — milk, 311 — oils and fats employed in manufacture of, 72 Cholesterol and phy tosterol, sepa- ration of, 48, 50 Cocoa butter, 164 sample, guaranteed pure, 84 Cocoa, introductory, 28 Cocoa, analysis of — Ash, 299 Extract, cold water, 295 Fat, 293 Fibre, 295 Moisture, 292 Nitrogen, 293 Phosphates, 298 Starch, 296 Sugar, 298 Cocoa, "levigation" method for estimation of shell in, 303 — microscopical examination of, 301 — shell and nib, calculation of amount of, 307, 362 " Cooolardo," 148 Coconut oil, 148 — shavings oil, 86, 151 -=- desiccated, 150 — "stearine," 152 — " oleine," 152 Cohune oil, 157 Colouring matters in butter and margarine, 112 Colour of oils, 5, 143 Colour tests, 52 Colza oil, 210 Consistency, 4, 96 Copper, detection and estimation of, 70 Copra, 148 Coquilla oil, 157 Corn oil, 230 " Cottolene," 216 Cotton seed oil, 216 bastard, 220 — Halphen's test for, 52 " stearine," 319 Cream, 360 Cuajo nut and oil, 270 Dahcan's method. See Titer test Desiccated coconut, 150 Diastase method for estimation of starch, 296, 336 Djave butter, 190 Dried milk in butter, 107 Earth nut oil, 202 Enteritidis sporogenes test, 359 Ester value, 22 Expression of oils and fats, 141 Extraction and estimation of oil or fat, 9 — commercial, 142 Eat, estimation of, 9 in cocoa, 293 in cream, 360 in feeding stuffs, 330 — — in milk, 346 — — in separated milk, 361 Fat— Butyracea, 188 Latifolia, 192 Longifolia, 194 Mahua, 183 Mowi-a, mowrah, 183 Tangkawang. See Borneo tallow Fats — Animal, 94 Bassia group of, 183 Index 367 Fats (con*.) — " Hydrogenised." See Appendix 0,282 Myristicaoese group of, 270 Tatty acids, separation and deter- mination of liquid, 31 examination of, by frac- tional precipitation, 91 solidifying point of, 13 Feeding stuffs, introductory, 319 Feeding stuffs, analysis of — Ash, 339 Fat (official), 330 Fibre, 337 Moisture (official), 330 Nitrogen, albuminoid, 339 — amide, 340 Phosphoric acid, 341 Potash, 343 Protein, total (official), 330 Starch, 335 Feedin" stuffs, calculation of ratictas, 320 sampling of (official), 329 Fehling's method for estimation of sugar, 314 Fermentation test for milk, 357 Fibre, estimation of, 337 Fish and marine animal oils, 273 Fluorides, detection and estima- tion of,in butter andmargarine, 109 Foam test, 125 " Foots," 145 Formalin, detection of, in milk, 351 Galam butter, 184 Ghee, 138 — a sample of adulterated, 90 Gingelly oil, 223 Glycerine, examination of, 61 Gottlieb process, 347 Ground nut oil, 202 Halphen's test for cotton-seed oil, 52 inferences from, 78 for linseed oil, 257 "Halyas,"225 Hemp oil, 280 Hexabromide test, 42 — — inferences from, 77 Hiibl method, 24 , Hydnocarpus oil, 26? Hydrocyanic aoid, deti^Ction and estimation of, 71 '• " Hydrogenised " fats and oils. See Appendix C, 282 Ioe-wafekS, oils and fats em- ployed in the manufacture of, 73 "lUipe fat," 166, 183, 194 Interpretation of analytical re- sults, 74 Iodine value, determination of, 24 — — inferences from, 76 Iron, detection and estimation of, • 70 Jamba oil, 213 Japan tallow ("wax"), 160 Japanese sardine oil, 276 Kapok oil, 220 Karite butter, 184 Kekune oil, 249 KeeJ)ing properties, 4 Kirschner process, 120 — value, inferences from, 77, 129-133 Kjeldahl-Gunning method for estimation of nitrogen, 331 Kokum butter, 181 Kottestorfer value, 42 Kreiss' reaction, 288 Kurdee seeds, 245 Lakd, 93 — examination of, for adultera- tion, 99 — microscopical examination of, 56 — substitutes, 85, 103 — vegetable, 104; 148 Latifolia fat, 192 Lead-salt-ether method, 31 inferences from, 77 Lead, detection and estimation of, 70 " Levigation" method, 303 Liebermann-Sto'Tch reaction, 55 Lime, detection and estimation of, 70 368 Fatty Foods Linolenic acid, 34 Linolic acid, 34 Linseed oil, 236 Longifolia fat, 194 Lukrabo oil, 267 Mafoubeiea oil and tallow, 176 Mafura oil and tallow, 176 Magnesia, detection and estima- tion of, 70 " Mahogany nuts," 190 Mahua fat, 183, 192 Maize oil, 230 Malabar tallow, 179 Maratti oil, 267 Maboaeine, analysis of — Colouring matters, 113 Curd, 107 Examination of fat, 129 Tat, 106 Preservatives, 108 Salt, 107 Water, 106 Margarine, oUs and fats used in manufacture of, 136 — interpretation of analytical results, 129 — mixtures, 81, 88 — sampling of, 105 Marine animal oUs, 273 Marotty oil. See Maratti Marzipan, oils and fat used in manufacture of, 73 Mee oil, 183 Melting-point, determination of, 12 — comparison with titer test, 15 Microscopical examination of beef and lard fats, 56 of butter, 125 of cocoa, 301 of milk, 356-360 Milk, analysis op, 345 Acidity, 349 Pat, 346 Preservatives, 351 Total solids, 345 Milk, cleanliness of, 355 sediment test, 356 fermentation test, 357 — — bacteriological condition of, 355 Milk, cleanliness of, Enteritidis sporogenes test, 359 — examination of, for pasteurisa- tion, 352 benzidene acetate test, 353 paraphenylenediamine test, 353 — separated, 360 — — extraction of fat in, 361 — sugar, 312 Molasses cake, 327 Mowrah fat, 183, 192 Mustard seed oil in rape cake, 211 Mutton fat, 94 Myristicacese group of fats, 270 NiQEE seed oil, 247 Nitric acid test, 101 Nitrogen, albuminoid estimation of, 339 — amide estimation of, 340 Njave butter, 190 " Nucoline," 148 Nutmeg butter, 272 Nut oil, 202 " Nutrex," 148 " Nutto," 148 OcHooo nut, 270 Oil, extraction and estimation of, 9 — from unknown oil-cake, 87 — cakes, various. See under respective oUs Oil— Almond, 206 Arachis, 202 Apricot kernel, 207 Bean, 226 Candle-nut, 249 Carthamus, 245 Castor, 262 Ceara rubber-seed, 255 Chaulmugra, 266 Chinese wood, 2.'51 Coconut, 148 — >havings, 151 Cohvine, 157 Colza, 210 Coquilla, 157 Corn, 230 Cotton, bastard, 220 — seed, 216 Index 369 Oil (cont.) — Dolphin jaw, 275 Earth nut, 202 Gringelly, 223 Ground-nut, 202 Hemp-seed, 280 Herring, 276 Hydnooarpus, 268 Jamba, 213 Kapok, 220 Kekune, 249 Linseed, 236 Lukrabo, 267 Mafura, 176 Maize, 230 Maratti, 267 Mee, 183 Menhaden, 276- Niger seed, 247 Nut, 202 Oleo, 94 Olive, 198 Palm, 173 Palm kernel, 153 Para rubber seed, 252 Pea nut, 202 Peach kernel, 207 Perilla, 259 Plum kernel, 207, 208 Poppy-seed, 241 Porpoise jaw, 275, 276 Porpoise body, 276 Bape, 210 Eavison, 212 Eice, 279 Eubber-seed, 252 Eubsen, 209 Eufisque, 202 Safflower, 245 Sardine, Japanese, 276 Sesame, 223 Shea nut, 184 Soja, 226 Stillingia, 170 Sunflower, 238 Tiger-nut, 278 Tung, 251 Walnut, 256 — country, 249 Whale, 276 Wheat, 279 Oleaginous material, preparation of, for extraction, 8 Oleic acid, 34 "Oleines," 144 " Oleo oil," 88, 94 — ■ stearine, 95 Olive oil, 198 sample of, guaranteed pure, 83 Optical activity, 17 inferences from, 78 Paokino oils, 73 Palm oil, 173 — kernel oil, 153 " stearine," 155 — — " oleiue," 155 " Palmine," 148 Paraffin wax in lard, 48 Para rubber seed oil, 252 Paraphenyleuediamine test. See Fasteurisation Pasteurisation, 352 Patrick's method for estimation of water in butter and marga- rine, 106 Peach kernel oil, 207, 208 Pea-nut oil, 202 Perilla oil, 160, 259 Phosphoric acid, estimation of, 341 Phulwa, 188 Phytosteryl acetate test, 44 — inferences from, 77 Piney tallow, 179 Plum kernel oil, 207 Poisonous substances, 5, 17, 266 Polarimetric examination of oils, 17 — estimation of cane sugar, 310 ^ of cane and milk sugar, 312 of starch, 297, 335 Polenske process, 117 ' — value, inferences from, 77, 126-133 " Pontianak illipe " nuts, 166 Poppy seed oil, 241 Potash, detection and estimation of, in oils, 70 — estimation of, in feeding stufes, 343 Premier jus, 94 Preservatives in butter and mar- garine, 108 24 370 Fatty Foods Preservatives in milk, 351 Eancidity, 284 — causes of, 285 — tests for, 287 Eape oil, 38, 210 Rations, calculation of, 320 Eavison oil, 212 Refractive index, determination of, IV inferences from, 76 Beichert-Meissl process, 117 value, inferences from, 77, 126-133 Renard's test, 34 — — inferences from, 77 Tortelli and Fortini's modi- fication of, 38 Rice oil, 279 Ricinolic acid, 34 Rosin, detection and estimation of, 55 Rotary power, 17 Rubber seed oil, 252 Rubsen oil, 209 Eufisque oil, 202 Safflowek oil, 245 Salad oils, 200, 204, 239, 243, 257 Salicylates, detection and esti- mation of, in butter and mar- garine, 110 in mUk, 351 Sampling of — Butter and margarine, 105 Feeding stuffs, 329 Glycerine, 61 Saponification value, determina- tion of, 21 — — inferences from, 75 — equivalent, 22 Scbiff's reagent, 287 Sediment test for milk, 356 Sesame oil, 223 Sbea nut oil, 184 " stearine," 187 " oleine," 187 Silver nitrate test (Beochi's), 54 Soda in oils, detection and esti- mation of, 70 Soja oil, 226 Solidifying point, determination of, 13 Soxblet extractor, 9 Specific gravity, determination of, 16 Standardisation of apparatus and solutions, 7 Starch, estimation of, 296, 335 Stearic acid, estimation of, 59 •- inferences, 79 Stillingia oil, 170 " Steai-ines," 144 Sugar, cane, 298, 310 — cane and milk, 312 Sulphur, detection and estima- tion of, 70 " Summer oils," 144 Sunflower oil, 238 Tallow — Beef, 95 Borneo, 166 Chinese vegetable, 169 Japan, 160 Mafura, 176 Malabar, 179 Mutton, 94 Piney, 179 Taste and smell, 5 Thermometer, stem correction for, 47 Tiger nut oil, 278 Tinned food packing oils, 73 Titer test, determination of, 13 comparison with melting point, 15 Toffee, oils and fats employed in manufacture of, 73 Tung oil, 251 Twitehell's process for estimation of rosin, 55 Typical oases, 79 !TJnsaponifiablb matter, deter- i mination of, 48 ' inferences from, 78 Valknta test, 123 " Vegetaline," 148 ;Vie<£nann's reaction, 287 Volatile fatty acids in butter and coconut oil, 115-120, 126-133 Index 371 Walnut oil, 256 — country, oil, 249 Water, estimation of, in oils and fats, 68 in butter and margarine. 106 " Wax " Japan, 160 Whale oil, 276 Wheat oil, 279 Wijs' solution, 25 " Winter oils," 144 Zinc, detection and estimation of, 70 Zulu nuts, 278 ADLAKU AND SON, IMPB., LONDON AND DOKEtlNG,