, , Klllb I .'., F.C.S. LEATHER INDUSTRIES LABORATORY BOOK LEATHER INDUSTRIES . LABORATORY BOOK 4 OF ANALYTICAL AND EXPERIMENTAL METHODS BY H. R. PROCTER, M.Sc. F.I.C. F.C.S. ' t PROFESSOR OF APPLIED CHEMISTRY (LEATHER MANUFACTURE) AT THE UNIVERSITY OF LEEDS SECOND EDITION, REVISED AND ENLARGED Xonfcon : E. & F. N. SPON, LIMITED, 57 HAYMARKET fiefo gorfe: SPON & CHAMBERLAIN, 123 LIBERTY STREET 1908 t ' -> 1 \ PREFACE TO THE SECOND EDITION. THE following notes on Analytical Methods employed in Leather Manufacture were primarily intended mainly for the use of my own students, and merely to supplement verbal teaching ; and although the scope of the book has been con- siderably expanded, it is occupied mainly with the practical details of analytical processes, and makes no attempt to teach either chemical theory, or the principles of leather manufacture. There is, of course, but little original in the methods described ; but with scarcely an exception they have been carefully tested and found satisfactory in my own laboratory, and in many instances they have been modified to suit the special purposes for which they are employed. The present edition has been thoroughly revised, and many new methods added, which, it is hoped, will prove useful to the practical chemist. My thanks are due to many of my friends for hints and suggestions received in the course of my work, but especially to Mr. J. T. WOOD for assistance in revision of the section on Ferments and Bacteriology. HENRY R. PROCTER. UNIVERSITY OF LEEDS, 1908. 333813 CONTENTS. PAGE INTRODUCTION xvii SECTION I. GENERAL CHEMICAL METHODS. Weighing .. .. .. .. .. .. .. .. .. I Measuring .. .. .. .. .. .. .. .. .. 3 Hydrometers .. .. .. .. .. .. .. .. .. 8 Distillation .. .. .. .. .. .. .. .. .. 9 Ignition .. .. .. .. .. .. .. .. .. 11 Precipitation .. .. .. .. .. .. .. .. ..12 Manipulation of precipitates .. .. .. .. .. .. ..12 Calculation of results .. .. .. ... .. .. .. -.13 Volumetric analysis .. .. .. .. .. .. .. .. 14 SECTION II. ALKALIMETRY. Indicators .. .. .. .. .. .. .. .. .. 16 Standard solutions .. .. .. .. .. .. .. .. 18 Testing of commercial acids and salts .. .. .. .. ..22 SECTION III. TECHNICAL WATER ANALYSIS. Temporary and permanent hardness .. .. .. .. ..31 Free alkalies .. .. .. .. .. .. .. .. 35 Magnesia hardness .. .. .. .. .. .. .. ..36 Chemicals required for softening .. .. .. .. .. .-37 Free carbonic acid .. .. .. .. .. .. .. 39 Chlorides .. .. .. .. .. .. .. .. .. 39 Iron .. .. .. .. .. .. .. .. .. ..40 Organic matter .. .. .. .. .. .. .. ..41 Sulphates .. .. .. .. .. .. .. .. .. 46 Statement of results .. .. .. .. .. .. .. ..48 viii LEATHER INDUSTRIES LABORATORY BOOK. SECTION IV. DEPILITANTS. PAGE Solubility of limes 53 Available lime .. .. .'. .. .. .. .. .. .. 54 Sodium sulphide .. .. .. .. .. .. .. .. 55 Arsenic sulphides .. .. .. .. .. .. .. .. 57 SECTION V. THE ESTIMATION OF AMMONIA AND THE DETERMINATION OF NITROGEN. Ammonia in lime liquors .. ' .. .. .. .. .. .. 62 Kjeldahl's method for nitrogen applied to lime liquors, hide substance, etc 64 SECTION VI. THE MATERIALS USED IN CURING, BATING, Etc. Common salt .. .. .. .. .. .. .. .. .. 69 Ammonium salts and ammonia .. .. .. .. .. .. 71 Sulphurous acid and sulphites .. .. .. .. .. .. 73 Boric acid and the borates .. .. .. .. .. .. .. 74 Carbolic acid and creolins .. .. .. .. .. .. .. 78 Aldehydes .. .. .. .. .. .. .. .. .. 79 Coal tar bates .. .. .. .. .. .. .. .. .. 84 SECTION VII. CONTROL METHODS FOR LIMING, BATING, Etc. Soaks .. .. .. .. .. .. .. .. .. .. 85 Limes and liming .. .. .. .. .. .. .. .. 85 Analysis of lime liquors .. .. .. .. .. . .. 87 Bacteriological examination .. .. .. .. .. .. .. 90 Puers .. .. .. .. .. .. .. .. .. .. 92 Bates .. .. .. .. .. .. .. .. .. .. 92 Drenches .. .. .. .. .. .. .. .. .. 93 Deliming acids .. .. .. .. .. .. .. .. 97 Control of hides and skins .. .. .. .. .. .. .. 98 Lime in hides .. .. .. .. .. .. .. t .. .. 101 Acid in hides , 102 CONTENTS. ix SECTION VIII. THE CHEMISTRY OF TANNINS AND THEIR DERIVATIVES. PAGE Chemical nature and classification .. .. .. .. .. 107 Preparation and purification.. .. .. .. .. .. . J 10 Methods of examination . .. .. .. .112 Decomposition by heat .. \ .. .. .. .. .. 113 Decomposition by dilute acids .. .. .. . ^9 Gallic acid .. .. ''...." .... ..120 Ellagic acid .. .. 1.21 Phlobaphenes .. ..^ .. 127 Constitution of the tannins .. .. .. .. .. 13 Natural colouring matters allied to tannins .. .. .. .. 14 SECTION IX. THE QUALITATIVE RECOGNITION OF TANNING MATERIALS. Reagents .. .. .. .. .. .. .. .. .. 148 Bromine water.. .. .. .. .. .. .. .. .. 149 Iron alum .. ,. .. .. .. .. .. .. ..150 Cupric sulphate and ammonia .. .. .. .. .. ,.150 Nitrous acid .. .. .. .. .. .. .. .. .. 151 Stannous chloride and hydrochloric acid .. .. .. .. .. 152 Phloroglucol reaction.. .. .. .. .. .. .. ..152 Sodium sulphite .. ... .. .. .. .. .. .. 152 Concentrated sulphuric acid.. .. .. .. .. . .. 153 Lime water .. .. .. .. .. .. .. .. .. 153 Formaldehyde .. .. .. .. .. .. .. .. .-.'. 156 Special reactions .. .. .. . .. .. .. .. 157 Tables , . k_ 157-167 SECTION X. ANALYSIS OF TANNING MATERIALS. HISTORICAL NOTICE. Davy's method .. .. .. .. .. .. .. .. 168 Wagner's method .. .. .. .. .. .. .. .. 169 Trotman and Hackford's method .. .. .. .. .. ..169 Gerland's method .. .. .. .. .. .. .. ..170 Carpene's method .. .. .. .. .. .. .. .-.171 Jean's method .. .. .. .. ,. .. .. .. .. 172 Monier's method .. .. .. .. .. .. .. .. 172 LowenthaPs method .. .. .. .. .. .. .. .. 172 The " indirect gravimetric " method .. ; ... . .. .. .. 173 a 5 x LEATHER INDUSTRIES LABORATORY BOOK. Simand, Weiss, Eitner, Murkatz, Hammer, Miintz, Ramspacher .. 174 Procter .. " , 74 Yocum .. .. .. .. .. .. tt j-- Parker and Payne . . . . . . . . tm j ~ . Wislicenus .. .. .. .. SECTION XI. SAMPLING, GRINDING AND EXTRACTION OF TANNING MATERIALS. Methods of sampling .. .. .. .. .. .. .. .. 177 Tools .. .. .. .. 178 Gambler .. .. .. .. .. .. . .. ..178 Valonia .. .. .. .. .. .. .. .. ..179 Myrobalans .. .. .. .. .. .. .. .. ..179 Grinding .. .. .. .. .. .. .. .. .. 180 Extraction .. . .. .. .. .. ..* .. .. io Real's press .. .. .. .. .. .. .. .. .. 182 Koch's extractor .. .. .. .. .. .. .. 183 Procter's extractor .. .. .. .. .. .. .. .. 185 Filter papers .. .. . - .. .. .. 186 Von Schroeder's extractor .. .. .. .. .. ..190 Vienna extractor .. .. .. .. .. .. .. .. 190 SECTION XII. TANNIN DETERMINATION BY THE HIDE POWDER GRAVIMETRIC METHOD. Total dry matter .. .. .. .. .. .. .. .. 192 Drying apparatus .. .. .. .. .. .. .. .. 193 " Total soluble " matter .. 194 " Soluble non-tanning matter " .. .. .. .. .. ..195 Filter bell method .. .. .. .. .. .. -.. . 195 Preparation of hide- powder for filter bell .. ..198 Statement of results . . . . . . . . . . . . . . . 200 American shake-method . . . . . . . . . . . . . . 20 j Kopecky's method . . . . . . . . . . . . . . . . 205 SECTION XIII. THE OFFICIAL METHOD OF TANNIN ESTIMATION. International Commission .. .. .. .. .. .. ' ' .. 211 General regulations .. .. .. .. . .. .. ..211 I. A. L.T.C. official method .. .. .. .. '*.. .. 212 A.L.C.A. official method .. ... .. .. .. .. 216 Advantages of the former .. . .. .. ... .. .. 219 CONTENTS. xi SECTION XIV. VOLUMETRIC METHODS OF ESTIMATING TANNINS. PAGE Lowenthal's method - 22 4 Modifications of Neubauer .. . .. 22 4 Simand .. 22 4 Procter . ... .. 22 5 Hunt .. ..225 Kathreiner .. ..-..*.. ,. .225 Von Schroeder .. 22 5 Jean's method .. .. .. 2 3 The Parker-Payne method .... ..232 SECTION XV. ESTIMATION OF COLOUR IN TANNING MATERIALS. Lovibond's tintometer . . . . . . . . . . . . . . 236 Tanning colour tests . . . . . . . . . . . . . . 239 Rotating churn apparatus . . . . . . . . . . . . . . 240 SECTION XVI. ANALYSIS OF USED LIQUORS AND SPENT TANS. Total solids 241 Total dissolved matter , , 241 Suspended matter .. .. .. .. .. .. .. .. 241 Ash and organic matter .. .. .. .. .. .. .. 241 Tanning matter .. .. .. .. .. .. .. .. 242 Free acids, and their determination . . . . . . . . . . 243 Procter's method 243 Koch's method . . . . . . . . . . . . . . . . 243 Kohnstein and Simand's method . .. .. .. .. .. 244 Simand's charcoal method .. .. .. .. .. .. .. 245 Volatile acids .. .. .. .. .. .. .. .. .. 246 Microscopic examination .. .. .. .. .. .. .. 246 Valuation of liquors and tannage .. .. .. .. .. .. 247 Spent tanning materials .. .. .. .. .. .. .. 248 SECTION XVII. ANALYSIS OF MATERIALS USED IN ALUM AND CHROME TANNAGES. Alum tannages .. .. .. .. .. .. .. .. 249 Determination of alumina .. .. .. .. .. .. 250 Determination of iron .. .. ., .. .. .. ..251 xii LEATHER INDUSTRIES LABORATORY BOOK. PAGE Determination of acids .. .. .. .. .. .. .. 253 Chromium compounds used in tanning .. .. .. .. .. 254 Determination of chromium in solutions of chrome salts .. 262 Determination of chromium and alumina in leather .. .. .. .63 Basicity of chrome liquors .. .. .. .. .. .. .. 266 SECTION XVIII. DETECTION AND ESTIMATION OF GLUCOSE. Fehling's solution .. .. .. .. .. .. .. .. 267 Simple volumetric method .. .. .. .. .. .. .. 267 Von Schroeder's gravimetric method .. .. .. ,, .. 269 SECTION XIX. ANALYSIS OF SOAPS. Determination of water .. .. .. .. .. .. .. 277 Fatty matter and alkali .. .. .. .. .. .. .. 278 Alcohol method .. 280 SECTION XX. OILS AND FATS. General chemistry of oils and fats .. .. .. .. .. , . 281 A. DETERMINATION OF TOTAL FATTY MATTER AND NON-FATS. Determination of water in oils and fats .. .. .. .. .. 288 Determination of ash .. .. .. .. .. .. .. .. 288 Detection and estimation of free mineral acids .. .. "... .. 289 B. PHYSICAL EXAMINATION OF FATS. Specific gravity .. .. .. .. .. .. .. .. 289 Sprengel tube .. .. .. .. .. .. .. .. .. 290 Mohr's hydrostatic balance .. .. .- .. .. .. .. 290 Melting and solidifying points ., .. .. .. .: ..291 Index of refraction .. .. .. .. .. .. .. .. 294 Zeiss refractometer .. .. .. .. .. .. .. .. 295 Specific temperature reaction .. .. .. .. .. .. 299 C QUANTITATIVE ANALYSIS OF FATS. Chemical fat constants .. .. .. .. .. , .. .. 301 Acid value .. .. .. .. .. .. .. *.. .. 301 Saponification value .. .. .. .. .. ... .. .. 302 CONTENTS. xiii PAGE Ether value .. .. .. .. .. .. 33 Iodine value 36 Unsaponifiable matter .. .. .. 39 Fatty acids .. .. .. .. .. 3H Oxidised fatty acids .. .. .. .. .- - 3H Resins .. .. .. .. .. .. .. 3 I( > Lactones .. .. .. .. .. .. .. 3 J 7 Acetyl value .. .. .. .. .. .. .. 3 [ & D. SPECIAL APPLICATIONS OF FAT ANALYSIS. Tallow 3 21 Neatsfoot oil .. .. .. .. .. .. .. .. .. 325 Olive oil 326 Fish-fat 3 2 7 Cotton stearine .. .. .. .. .. .. .. .. 327 Wool-fat '. 327 Distilled grease stearine .. .. .. .. .. .. ..329 Cod oil .. .. .. .. .. .. .. .. .. .. 330 Other fish oils .. .. .. .. .. .. .. .. .. 334 Whale oil .. 336 Seal oil .. .. .. .. .. .. .. .. .. .. 337 Degras and sod-oil .. .. .. .. .. .. .. .. 337 Resin oils .. .. .. .. .. .. .. .. .. 343 Mineral oils .. .. .. .. .. .. .. .. .- 344 Mineral waxes.. .. .. .. .. .. .. .. .. 345 Paraffin wax .. .. .. .. .. .. .. .. .. 345 Ceresin wax .. .. .. .. .. .. .. .. .. 346 Animal and vegetable waxes .. .. .. .. .. .. .. 346 Beeswax .. .. .. .. .. .. .. .. .. 346 Carnauba wax .. .. .. .. .. .. .. .. 347 Japan wax .. .. .. .. .. .. .. .. .. 348 Tables of constants .. .. .. .. .. .. .. .. 349 SECTION XXI. ANALYSIS OF COMMERCIAL PRESERVED EGG- YOLK. Sampling .. .. .. .. .. .. .. .. ..351 Determination of water and fatty matters .. .. .. .. ..352 Determination of common salt .. .. .. .. .. --353 Determination of ash.. .. .. '.. .. .. .. 353 Typical analysis .. .. .. .. .. .. .. .. 354. SECTION XXII. ANALYSIS OF LEA 1 HER. Preparation for analysis .. .. .. .. .. .. .-355 Mineral ash .. .. .. .. .. .. .. .. .. 356 Moisture .. .. .. .. .. .. .. .. .. 357 xiv LEATHER INDUSTRIES LABORATORY BOOK. 1*A(.K Fats .. ,; .. 357 Water soluble matter .. .. .. .. .. .. .. .. 358 Glucose and dextrine .. .. .. .. .. .. .. .. 360 Analysis of leather ash .. .. .. .. .. .. .. 363 Examination for acids . . . . . . . . . . . . . 366 Free mineral acids .. .. .. .. .. .. .. .. 367 Procter and Searle's method .. .. .. .. .. .. 369 Balland and Maljean's method .. .. .. .. .. .. 370 VVuensch's method .. .. .. .. .. .. .. 371 Stiasny's method .. .. .. .. .. .. .. .. 372 SECTION XXIII. DYES AND DYE STUFFS. Natural and artificial dyes .. .. .. .. .. .. .. 374 Methods of classification and identification .. .. .. .. 375 Systematic dye-tests .. .. .. .. .. .. .. .. 382 SECTION XXIV. THE USE OF THE MICROSCOPE. Selection of microscope .. .. .. .. .. .. 391 Objectives 392 Eye-pieces .. .. .. .. .. .. .. .. .. 393 Immersion objectives .. .. .. .. .. .. .. .. 396 Illumination .. .. .. .. .. .. .. .. .. 397 Abbe illuminator .. .. .. .. .. .. .. .. 398 Focusing .. .. .. .. .. .. .. .. .. 398 Air bubbles 399 Oil globules 399 Preparation and mounting of objects .. .. .. .. .. 399 Starches .. 40 Use of iodine .. .. .. .. .. .. .. .. .. 401 Use of polariscope .. .. .. .. .. .. .. .. 402 Examination of fibres and powders.. .. .. .. .. .. 402 Cutting section? .. .. .. .. .. .. .. .. 403 Drawing and measuring .. .. .. .. .. .. .. 405 SECTION XXV. THE MICROSCOPIC STRUCTURE OF SKIN. Epidermis and corium, distinct origin in ovum .. .. .. .. 4*3 Rete malpighi or mucous layer .. .. .. .. .. ..413 General structure of skin .. .. .. .. .. '- .. .. 4 I 3 Development of young hair, and of epidermis .. .. .. .. 4H CONTENTS. xv PAGE Structure of hair ,. .. -. . 4 ! 5 Hyaline or glassy layer .. .. .. 4 r 5 Sebaceous and sudoriferous glands .. .. .. .. ..416 Erector pili .. W Nerves 4*9 Preparation of skin for microscopic examination .. .. .. 4 22 Voluntary muscle . . . . . . . . . 4 2 3 Yellow or elastic fibres .. .. .. 4 2 3 Preparation of skin for section cutting .. .. .. .. .. 4 2 4 Cutting sections .. .. .. .. .. 4 2 5 Staining sections .. .. .. .. .. 4 2 5 Mounting sections .. .. .. ... .. 4 2 5 SECTION XXVI. BACTERIOLOGY AND MYCOLOGY. Bibliography .. .. .. .. .. .. .. .. 427 Classification of bacteria .. .. .. .. .. .. .. 430 Spirillus.. .. .. .. .. .. .. .. .. .. 430 Aerobic and anaerobic bacteria .. .. .. .. .. .. 432 Micrococcus .. .. .. .. .. .. .. .. .. 432 Bacillus .. .. .. .. .. .. .. .. .. 433 Saccharomycetes .. .. .. .. .. .. .. .. 437 Ascomycetes .. .. .. .. .. .. .. .. .. 438 Phycomycetes .. .. .. .. .. .. .. .. .. 439 Bacteriological methods .. .. .. .. .. .. .. 440 Sterilisation .. .. .. .. .. .. .. .. .. 441 Cultures.. .. .. .. .. .. .. .. .. .. 443 Pasteur's fluid .. .. .. .. .. .. .. .. .. 442 Liquid media .. .. .. .. .. .. .. .. .. 443 Gelatine media .. .., .. .. .. .. .. .. 443 Tubes and plate cultures .. .. .. .. .. .. .. 444 Microscopic examination .. .. .. .. .. .. .. 447 Staining .. .. .. .. .. .. .. .. .. 447 INDEX 449 INTRODUCTION. LEATHER manufacture has long been conducted by " rule ot thumb," rather than by the more exact methods of modern science, not so much from any preference of manufacturers for the former, as because, from the inherent difficulties of the subject, the latter has been scarcely possible. Another obstacle which has somewhat impeded progress has been a misap- prehension, perhaps justified in the past by adventurous in- ventors who would " revolutionise the trade," as to the part which chemistry is destined to play in the tanning of the future. The term " chemical tanning " became one of reproach and contempt ; and it has been too much forgotten that in every tan-yard, chemical processes are being carried on of a com- plexity so great that as yet they are very imperfectly understood, even by those chemists who have devoted themselves most thoroughly to the task. The first service which chemistry must render to the trade is not to supersede the processes of manufacture which have stood the test of experience, but to explain the principles on which they depend, and so enable us to carry them out with more exactness and invariable success. To do this, we must apply the usual tools of chemical research the balance and the measure ; but to do it successfully, methods have to be elabo- rated to determine the nature and quantities of the very varied constituents of all the materials used. In this direction there is yet much to be done, but it has been my effort in this little book to sum up the best methods which are yet known, and to put them in a handy form for use and reference. Those who can look back on the progress of the last twenty years will feel that, xviii LEATHER INDUSTRIES LABORATORY BOOK. however much remains to be done, the progress already made has been very substantial. It is now possible to determine the strength and value of most tanning materials with an accuracy and certainty which then was quite unknown ; waste liquors and tans can be analysed with approximate accuracy, and the loss in leaching and tanning ascertained, and many new methods have been introduced for the control of processes and materials ; and only the systematic application of these methods is required to learn much which we yet need to know about the economies of the tanning process. In liming, bating and drenching, much information has been gained as to the losses of hide substance which take place, and the way in which these are influenced by the materials employed, and the mode in which the process is conducted ; and new materials have been introduced and accepted by the trade, for which means of valuation and analysis have been devised. Bacteriology has taught us at least some of the methods by which fermentative processes such as puering and drenching can be regulated, and has strengthened the hope that we may be made independent of the old and disgusting materials ; while the removal of lime by purely chemical means is perfectly understood, and a matter of every-day practice to many tanners. In the tan-house we no longer hear the vague talk about " gallic acid " as an enemy to be avoided, but the amounts of acetic and lactic acid in the liquors are estimated in well-managed sole-leather yards as a matter of routine ; their influence on the proper tanning of the goods is well understood ; and it is recognised that they are derived, not from the de- struction of tannin, but from the sugary matters of the tanning materials. In chrome tanning an entirely new process originating in the laboratory has firmly established itself, and is slill finding new applications ; and the accuracy which is essential to its success has involved the invention of new schemes of analysis, and done much to familiarise tanners with chemical methods. In the analysis of oils and fats, the old and unreliable tests have been largely superseded by new and exact methods ; and INTRODUCTION. xix tanners rather than chemists must now be blamed that more is not known of the differences of constitution which render an oil fit or unfit for the purposes of the trade ; since, when difficulties arise, the chemist is constantly told that the source of the mate- rials in fault cannot be traced, and that no samples have been preserved for analysis ! In dyeing and staining, the old natural dyes are more and more giving place to the synthetical organic colours of the chemist, which compete with them successfully, not only in price but in brilliancy and fastness. Here, again, the services of the chemist are in demand, not only to test the qualities of the dyes , but to match the tints required with the cheapest combinations, and to recognise the character of unknown colours. In one way, indeed, chemistry must be admittedly unwelcome to the conservative tanner it is the key to mysteries ; and has rediscovered and thrown open to the world many secrets which he believed safely locked within the doors of his factory ! But even so it must be reckoned with, for, like the genii of the stories, it can neither be killed nor bound, but is fortunately the willing servant of all who have intelligence to use it ; and as capable of discovering new secrets as of throwing open the old ones. A few words may here be said as to the origin and scope of the succeeding chapters. They originally, in a much briefer form, were a part of the Author's 'Text-Book of Tanning,' pub- lished in 1887 ; but as time went on, and knowledge increased, it became evident that the amount of analytical chemistry avail- able and earnestly demanded by the more advanced workers was too large to be conveniently incorporated in a book intended rather for the tannery than for the laboratory ; while, on the other hand, the technical chemist would have had difficulty in picking out the methods he required from the larger volume. On these accounts, it was decided to issue the purely laboratory matter in an independent volume, which has now come to its second edition, and the reader may be referred for informa- tion on the theory and practice of leather manufacture to the ' Principles of Leather Manufacture,' which has appeared as a xx LEATHER INDUSTRIES LABORATORY BOOK. separate volume ; * while, on the other hand, the working details of chemical methods of research are given in the ' Laboratory Book ' and merely referred to in the ' Principles.' The ' Laboratory Book ' is not intended to take the place of the ordinary chemical text-books, but to supplement them ; and those readers who have not already a knowledge of the common methods of the laboratory will find it needful to study some of the many Manuals on general Quantitative Analysis. It is also extremely desirable that the chemist should familiarise himself with some of the later developments of chemical theory, such as the laws of chemical equilibria and mass-action, of ionic dissociation, and of osmotic pressure ; and the advances which are now being made in our knowledge of colloidal solution, which have important bearing on the very complex problems of leather manufacture. To those who have already some know- ledge of the analytical methods, it is hoped that the opening section on ' General Chemical Methods ' may prove a useful reminder of the more important details of manipulation. The succeeding chapters assume a knowledge of these, and of the elements both of leather manufacture and chemical theory. * E. & F. N. Spon, Limited, price i&r. net. THE LEATHER INDUSTRIES LABORATORY BOOK. SECTION I. GENERAL METHODS. THE object of chemical analysis is to divide a material into its constituent parts, so that they may be weighed or measured, and their proportions calculated. These are generally expressed as percentages of the whole, and are calculated by multiplying the weight found by IOO, and dividing by the original weight of material used. It is rarely necessary or desirable to go beyond the first place of decimals in technical analyses. Sometimes the weight of the substance which it is desired to estimate is not found directly, but calculated from that of some other substance to which it is known to be proportional. Weighing. Substances must never be weighed directly in the balance pan, but always in a basin or some other vessel, of which the weight is previously determined, and which must be dry, cool, and clean, and must not be touched by hot or damp fingers before weighing. Vessels which have been heated > especially if they contain substances which absorb moisture readily, must be allowed to cool in the " desiccator " or " exsic- cator," a vessel with an air-tight cover, and usually containing dry calcium chloride or sodium hydrate, or pumice moistened with concentrated sulphuric acid. It must be placed on the pan while the balance is arrested, and neither vessel nor weights must ever be touched while the beam is free. Weights are placed on the right-hand pan with forceps, beginning with the largest, and trying each in turn by liberating the balance, till one is found B 2 THE LEATHER INDUSTRIES LABORATORY BOOK. which is lighter than the body which is to be weighed. Smaller weights are then added in the same way, always trying the larger first, and keeping the weight below that of the object. When the weight is within one centigram, the balance case is closed, and the final adjustment is made with the rider. This is complete when the index swings an equal distance on each side of the centre. Many balances are adjusted to swing an even number of divisions to one side or the other for each milligram of over or under-weight, and most can be made to do so by n\ /Ts FIG, i. CHEMICAL BALANCE. a, scale for rider ; b, arrestment for pans ; e, e, arms for moving rider. suitable adjustment. Where this is the case the division at which the pointer turns at each side is noted and that to the left deducted from that to the right, and the corresponding number of milligrams + or -, added to the weight on the pan. It is generally safest to confirm the result by moving the rider to the weight indicated. The balance may be made to oscillate if required by gentle fanning with the hand. The weight should be read off the empty spaces in the box, and entered in the note- book, and checked in taking off the weights. GENERAL METHODS. 3 Measuring is done in glass vessels marked to definite quan - tities. Water, and all other liquids which wet glass, form a hollow upper surface or meniscus, the lowest part of which must be brought level with the mark. This is most accurately done by looking through the liquid at some distant light object nearly level with the eye, the lower surface of the meniscus being darkened, if necessary, by holding some dark object behind 10 cc V Pipettes. c c 15' C -^-'00 j-30 ! fh w W JJ60 I 5 f- 40 f-30 ~ 20 H" 60 'S' 50 Jo f-50 I. I i Cylinders. FIG. 2. MEASURING INSTRUMENTS. Burettes. and slightly below it. A less accurate method is to look at the lower edge of a white card held behind and nearly level with the meniscus, with which the eye must also be level. For dark- coloured liquids in burettes, a Mohr's float should be used, which may be either of glass or porcelain. Care must be taken that the vessel is upright while reading. Flasks and cylinders are usually graduated to contain the amount to which they are marked ; pipettes and burettes to deliver hence, if the former B 2 4 THE LEATHER INDUSTRIES LABORATORY BOOK. are used for delivery, the amount of liquid which clings to the sides must be determined and allowed for. Burettes must not be read till time has been allowed for this to run down, and pipettes must be drained against the side of the vessel, but not blown out. The capacity of a vessel may be determined by weighing the amount of water at 15 C. which it contains (or delivers), when each gram corresponds to a cubic centimeter (c.c.).* Both flasks and pipettes are frequently inaccurately marked, and should be tested before use. The capacity of the lower divisions of burettes is often less than that of the upper ones. Pipettes are filled by sucking till the liquid rises above the mark, quickly placing the forefinger on the top, and allowing" to run down till the mark is reached. Ordinary pipettes should not be used in the mouth for very poisonous or corrosive liquids, but may be sucked through a bulb tube connected with rubber tubing which is pinched when the pipette is full. Marking Vessels. Glass is best marked with a writing dia- mond, or by etching with hydrofluoric acid. The latter is most easily done by covering with a thin film of beeswax or solid parafrln, and etching with a needle. The etched part is then covered for a few minutes with a piece of blotting paper moistened with liquid hydrofluoric acid. The latter must be handled with great care as it is extremely corrosive. Inks con- taining fluorides are sold for writing on glass. Porcelain basins and crucibles may be marked in common ink, dried and heated, first gently and then strongly in the oxidising flame of the blowpipe. The Metric System. As the metric system is now almost universally employed by chemists from its great convenience for purposes of calculation, and is likely ultimately to be also used by all civilised nations for commercial purposes, no apology is needed for its adoption in this book ; but a few words on its relations with the antiquated weights and measures still used in England are perhaps necessary. Metric weights and measures form a connected system related to the meter, a measure of length equal to 39*3/0113 English * Laboratory vessels are generally graduated at this temperature, but the true liter is I kilogram of water at 4 C. Compare p. 5. Glass-tappe'd burettes are generally to be preferred to those with india-rubber tube and pinch-cock, but the tap should be lubricated with a little vaseline, and removed and washed after use with caustic alkalies, or it will stick. GENERAL METHODS. 5 inches, which is approximately 1/10,000,000 of the earth's quad- rant measured from pole to equator. The meter is divided into 10 parts, or decimeters, 100 parts, or centimeters, and 1000 parts or millimeters. The standard of capacity is the cubic decimeter, or " liter," which of course contains 1000 cubic centimeters, or c.c. The standard of weight is that of I c.c. of water at 4 C. (the temperature of the greatest density of water), the weight of which is called I gram. Multiples of this are denoted by Greek numerals, thus : I dekagram = 10 grams. (" Deka ") I hectogram = 100 grams. I kilogram = 1000 grams, the weight of I liter of water. In the same way fractions of I gram are denoted by Latin numerals, thus : i decigram = o * I gram, i centigram = 0*01 gram, i milligram = 0*001 gram. For purposes of reduction the following approximate numbers may generally be used : i gram = 15*43 grains. I Ib. av. = 453* 6 grams, i liter = 0*22 gallon i gallon = 4*546 liters. * i cubic meter of water weighs 1000 kilograms, or 2200 Ib., or approximately I ton. I liter = 61*02 cubic inches, i fluid oz. = 28*34 c.c. i meter = 39*37 inches. I inch = 25 * 4 millimeters. Actual reductions may, however, in most cases be avoided by treating a question as one of proportion. Thus, I grm. or c.c. per liter is equivalent to i Ib. or i pint per 100 gallons, or very approximately to I oz. per cubic foot ; 100 gallons of water weighing 1000 Ib., and I cubic foot 997 oz. For laboratory use, " Mohr's liter," which is the volume of 1000 grm. of water at 15 C. (59 F.), or ordinary laboratory temperature, is usually employed in place of the true liter, which is the volume of 1000 grm. at 4 C. * The small difference between I gallon measured in liters, and 10 Ib. avoirdupois measured in grams, is due to the different standard temperatures of the two systems. Consequently i gallon would only contain 4*536 " Mohr's liters." THE LEATHER INDUSTRIES LABORATORY BOOK. Thermometer Scales. The Centigrade, or Celsius thermo- meter, in which the freezing point of water is taken at O and the interval between freezing and boiling point is divided into IOO parts or degrees, is to be preferred to that of Fahrenheit, which starts from an arbitrary zero, and in which freezing point is represented by 3? and boiling by 212. Centigrade degrees may be converted into Fahrenheit by multiplying by f and adding 32. The following table gives the points at which they agree without fractions : Degrees Cent. Degrees Fahr. Degrees Cent. Degrees Fahr. 20 - 4 45 "3 15 5 50 122 10 H 55 131 23 60 140 32 65 149 5 4i 70 I 5 8 10 50 75 I6 7 15 59 80 I 7 6 20 68 85 I8 5 25 77 90 194 30 86 95 20 3 35 95 IOO 212 40 104 105 221 Reaumur's thermometer, which divides the interval between the boiling and freezing points of water into 80 is still a good deal used in Germany. Specific Gravity is the weight of any body as compared to that of an equal volume of water. Thus, I c.c. of sulphuric acid weighs I 84 grm., while I c.c. of water only weighs I grm. ; the specific gravity of the acid is therefore I 84. As bodies ex- pand by heat, a uniform temperature must be adopted, which is generally 1 5 or 15- 5 C. (59 or 60 F.). Liquids are brought to the required temperature by immersing the containing vessel in warm or cold water. Liquids may be weighed in a flask or bottle carefully filled to a mark in the neck, or more readily in one provided with a perforated stopper (pycnometer). The flask is weighed first empty GENERAL METHODS. and dry, then filled with water, and finally with the liquid of which the specific gravity is to be taken. The weight of the liquid divided by that of the water (deducting the weight of the flask in each case), gives the sp. gr. To fill, place the flask on a saucer, fill quite full, insert stopper firmly and wash with cold water, or in the case of oils and fats, with petroleum spirit or benzoline, or other fat solvent, finally drying with a clean cloth. With thick liquids, such as extracts and viscid oils, a wide-necked flask, and a short wide funnel must be used in filling, and great care taken to avoid air-bubbles. Air-bubbles may also be removed by keeping the flask for some time in an exhausted air-pump receiver. Immediately after weighing the stopper must be with- drawn, and after returning as much of the extract to the sample as possible, the flask must be washed out with hot water, or, in the case of fats, with a suitable fat-solvent, and dried ready for use when wanted. For small quantities of liquid, the Sprengel tube, Fig. 3, is useful. It is filled through A by suction, and the liquid adjusted to mark M by applying blotting-paper to A to remove, or a glass rod with a drop to replace liquid. A and B should not exceed about I mm. in internal diameter. Solids, or very pasty extracts, may be determined by placing any convenient quantity in the flask, weighing, filling up with water, and weighing again. Deducting the weight of water so added from the total the empty flask would contain, the re- mainder is the weight of water which would occupy the same space as the solid, and which divided into the weight of the latter gives its sp. gr. Solid fats may be melted and allowed to solidify in the flask, but 1 8 to 24 hours should be allowed before filling with water and weighing, as fats only slowly regain their FIG. 3. SPRENGEL TUBE. 8 THE LEATHER INDUSTRIES- LABORATORY BOOK. original density after fusion. On the use of sp. gr. in water for determining dry solids in wet hide, see p. 99. Solids may also be determined by suspending by a fine wire from the hook above the balance pan, and weighing first in air, and then in water con- tained in a beaker supported on a bridge over the pan, resting on the bottom of the balance case. The last weight deducted from the first gives that of the water which has been displaced. If a solid of known weight in air, such as a glass plummet, be weighed first in water, and then in some other liquid, supported on the bridge, the loss of weight in the last case, divided by that in the first, will give the sp. gr. of the liquid. Special balances on the principle of the steelyard are often used for the purpose. Specific gravities of liquids are rapidly taken with tolerable accuracy by hydrometers, which sink till they displace a volume of liquid equal to their own weight. Many of these are graduated for special purposes or have quite arbitrary scales. The degrees of the barkometer maybe reduced to sp. gr. by dividing by 1000 and adding I, but are rarely quite accurate. Degrees Twaddell may be similarly reduced by dividing by 200 and adding I (hence i Twaddell equals 5 barkometer). Baume degrees are quite arbitrary, and can only be reduced by tables, of which several discrepant ones exist. The following so-called " Rational " scale is one very frequently used ; but the use of this and other arbitrary and unscientific scales is to be avoided as far as possible : BAUME HYDROMETER TABLE. Baume, 12- 5 C. j Baume, 12-5 C. Sp. gr. o Sp. gr. o I'oooo 40 "3834 5 1*0358 45 '453 1 10 15 20 25 30 35 0745 50 -5301 1160 55 '6158 1608 60 "jn6 2095 65 -8195 2624 70 1-909 3202 72 -960 In taking a specific gravity with a hydrometer, the jar should be placed in a saucer and filled to the brim. Froth is then swept off with a glass rod, and the hydrometer put in, overflowing liquid being caught by the saucer. The hydrometer is read from the point where the curved surface of the water touches the stem. GENERAL METHODS. 9 Evaporation is a general method of separating water and other volatile matter from non-volatile. It is most safely done on the water or steam-bath, but where large quantities have to be evaporated it is sometimes best to use a small flame. Liquids evaporate rapidly at or near boiling-point in broad shallow basins freely exposed to the air, or may be boiled in a flask with a funnel or glass bulb in the neck, or in a deep beaker covered with a clock-glass to avoid spurting, or in many cases, simply in a long-necked (Jena) flask. Tanning liquors may be concentrated in a flask as described without very material loss of tannin, but, like other materials which are injured by heat, are best evaporated in vacuo, the boiling point being much lower under reduced pressure. Many substances apparently dry, as bark, leather, and the residues after evaporation on the water- bath, may be thoroughly dried in the water-oven, which, being surrounded by boiling water and steam, is at nearly 100 C. Some substances part with water of crystallisation or hydration at 100 C., but may often be dried without decomposition at ordinary temperatures over sulphuric acid, and, preferably, in vacuo. In many cases an air-oven may be used with advantage, heated by gas to 105 or 110 C, in which the drying is much more rapid ; but care must be taken that the temperature inside the oven is uniform and steady. To estimate water, a portion of the sub- stance must be weighed in a basin or crucible, dried as above, and re-weighed, when the loss may in most cases be calculated as water. As most substances when thoroughly dried will ab- sorb water from the air, the basin must be cooled in the desic- cator, rapidly weighed, and again dried and weighed till the weight is constant. Inorganic matters can generally be dried at a much higher temperature, or even ignited, though other volatile matters will, of course, be included. Distillation is merely a case of evaporation in which the volatile portion is collected. The liquid is placed in a flask, which is closed by a perforated cork fitted with a tube bent to an angle somewhat more acute than a right-angle, and fitting tightly into an inclined condenser consisting of a rather wide glass tube surrounded by an outer casing, through which cold water is allowed to flow from a tap. The lower end of the condenser tube passes into a flask, with which it should form a pretty close, but not an air-tight, connection, and which, in the io THE LEATHER INDUSTRIES LABORATORY BOOK. case of volatile liquids, should also be immersed in cold water or even in ice. The liquid in the flask is boiled, either by a flame or on a sand-bath, or, in the case of combustible liquids like alcohol or ethers, by immersion in a water * or steam-bath ; and the vapours pass into the condenser, where they are cooled and condensed, and flow down into the receiving flask. This simple arrangement is quite satisfactory where mere purification of a liquid from non-volatile impurities is all that is required ; but where a more volatile liquid must be separated from one of higher boiling-point, as for instance alcohol from water, it does not prove very effective, as a portion of both liquids is carried over, though in simple cases it will act fairly well. It is usually necessary to know the temperature of the vapour which is being distilled, in order to judge when the process has gone far enough, for which purpose a second hole is made in the cork of the distilling flask, and the bulb of a thermometer is introduced into the neck of the flask. A more convenient arrangement is to use a " distilling flask " in which a side tube passing into the condenser is fused into the neck, while the thermometer is fitted through the cork, with its bulb near the opening of the side tube. The most difficult case is that of a liquid like petroleum, which consists of a mixture of a great number of different hydrocarbons of different boiling-points, which it is desired to separate. When such a liquid is boiled, the fractions of higher boiling-point distil first, but the vapour, even from the outset, consists not merely of the most volatile, but of a mixture of these and a smaller proportion of the less volatile. "The separation is greatly facili- tated by the use of a " dephlegmating " or " fractionating " tube, in which the less volatile vapours become condensed, and drop back into the distilling flask, at the same time washing the ascending vapours and giving up to them any more volatile portions. Glynsky's and Le Bel Henniger's are, perhaps, the best forms of this apparatus, but a wide tube filled with large glass beads or short lengths of quill tubing, and fitted with a side tube, is also very effective. The arrangement is shown in Fig. 4. Even with a fractionating tube the separation pf the different liquids is rarely complete ; but by separating into por- tions boiling, say, from 60 to 70, 70 to 80, and so on, and * Higher temperatures may be obtained by the use of calcium 'chloride solution, or by glycerine or oil baths. GENERAL METHODS. ii again fractionating these portions separately, nearly pure pro- ducts may be obtained. Instances of the application of this method may be found in the separation of alcohols and volatile acids of the acetic series, and of petroleum ether. Ignition. Heating to redness in an open vessel drives off chemically combined water, ammonia, and often carbonic acid, etc., and destroys most organic matter. This is generally best done in platinum vessels, but salts of tin, lead, copper and other metals easily reduced to the metallic state, must not be heated FIG. 4. in platinum, or the vessel maybe destroyed. Materials contain- ing chlorides, or other salts fusible at a red heat, must be ignited very gently and slowly, or the particles of carbon will be encased and escape combustion. If this occurs, the addition of a little water when cool, with subsequent drying on water-bath and re- ignition will often succeed in completing combustion. In difficult cases the ash may be moistened with solution of ammonium nitrate, and again dried and ignited, or the soluble salts may be extracted with water from the charred mass, which is collected on a quantitative filter-paper, returned to the crucible and com- 12 THE LEATHER INDUSTRIES LABORATORY BOOK. pletely burnt, and the solution evaporated with the ash after ignition. Precipitation. Many substances soluble in water, etc., can be converted into insoluble compounds, and so separated by precipitation. Thus, on addition ofBaC! 2 , sulphuric acid (free or combined with bases) is precipitated as insoluble BaSO 4 ; and ammonium oxalate precipitates all soluble lime salts from neutral or alkaline solution as calcium oxalate, which may be converted into carbonate by gentle ignition, or into quicklime (CaO) by strong ignition over the blow-pipe. Precipitation is mostly aided by warmth and time, and often delayed or pre- vented by presence of organic matter. It is generally rendered more complete by an excess of the precipitant. Very finely divided precipitates, such as calcium oxalate, usually become coarser by boiling, or standing for some time in the liquid in a warm place, and are then more easily filtered and washed. Filtration. Precipitates are separated from liquids in which they are suspended by settling and filtration. Circles of paper are used, as filters, which are folded in four, one fold opened, and inserted in a funnel, which must be slightly larger than the cone of paper. The filter must generally be wetted with distilled water, except when dry filters are re- quired for solutions for volumetric analyses. This .water must be allowed to run through the filter, so that the upper part of the funnel is not wetted. Special filter paper is used for quantitative work, which leaves no appreciable ash on ignition. With bulky cr gela- tinous precipitates it is often convenient to use the vacuum pump where one is available. When this is not the case, filtration may be much quickened by the use of a piece of small glass tube below the funnel, attached with a piece of indiarubber tube as in Fig. 5. The moistened filter must be closely fitted to the funnel by gentle pressure round the edge, especially at the fold. FIG Washing Precipitates. Before filtration, it is gene- rally best to let the precipitate settle, and pour first the clear liquid on the filter ; and if the precipitate is heavy and settles well, water may be added, and the settling and decantation re- peated several times. The liquid should be poured down a glass GENERAL METHODS. 13. rod on to the side of the filter, avoiding spurting, and the last traces of precipitate removed from the beaker by rubbing with a cap of rubber tube on the end of the glass rod, and rinsing with a stream of water from the fine jet of the wash-bottle. A short glass rod, bent at somewhat more than a right angle, and hung on the edge of the beaker, is often convenient for pouring and does not disturb the precipitate. The precipitate must be well washed on the filter, which is generally best done with hot water blown on it in a thin stream from the wash-bottle. The filter should be allowed to run dry each time before adding more water. If the precipitate is subsequently to be removed from the paper, it is desirable to wash it as competely down to the bottom, as possible. Drying Precipitates. Organic precipitates must be collected on filters previously dried till constant in the oven, and weighed in a closed tube or weighing bottle. The filter with, precipitate is similarly dried, weighed, and the weight of the filter deducted. Instead of weighing the dried filter, it may be balanced by a second which is dried with the precipitate and used to ascertain the tare. Mineral precipitates may generally be ignited. A crucible previously dried and weighed is placed on glazed black paper, and most of the precipitate is separated from the filter-paper and transferred to the crucible. The filter-paper is now folded up small, twisted in a platinum wire, and carefully burnt in a Bunsen flame over the crucible, into which the ash is shaken, and the whole then ignited, cooled in the desiccator and weighed. A deduction must be made for the filter ash if appreciable, but with modern papers it may generally be neglected. Where no reduction is to be feared, precipitates may be ignited with the paper in the crucible. Calculation of Results. In many cases the substance actually weighed is not that of which we require the weight, but some other compound, from which the weight must be calculated. Thus sulphuric acid is generally weighed as barium sulphate, the reaction being as follows : H 2 SO 4 + BaCl 2 = BaSO 4 + 2HC1. The atomic weights are 98 + 208 =233 + 73, and are obtained by adding the atomic weights of the constituents of each body. 14 THE LEATHER INDUSTRIES LABORATORY BOOK. Thus in this case we reckon only 98 parts of sulphuric acid for every 233 parts of barium sulphate weighed. Where many analyses are made, it is convenient to use factors to give the actual weight by simple multiplication. Tables of such factors and their logarithms are given in the Chemiker Kalender and many text-books.* Volumetric Analysis. It is clear that if we know the weight of BaCl 2 required for the above reaction, we can calculate that of all the other substances, and vice versa. If we have a solution of HaCl 2 containing exactly 104 grm. per liter, or 0*104 grm. per c.c., and if for instance it requires 20 c.c. to precipitate all the sulphuric acid (free or as sulphates) in a liter of water, we can calculate that the water contains 20 x 0^049 01*0*98 gi"m. of sulphuric acid per liter. Such solutions of known strength as that of barium chloride are known as " standard solutions " and are usually made to contain I combining weight of the substance in grams per liter, or some definite fraction of it. Referring to the equation, it will be noted that I eq. of H 2 SO 4 liberates 2 eq. of HC1, and that the I atom of Ba in the barium sulphate takes the place of 2 atoms of hydrogen. Such elements are called " divalent," and acids which, like sulphuric, have 2 atoms of hydrogen which can be displaced by one divalent atom are called " dibasic." In making " standard " solutions of such sub- stances, it is found convenient, in order to make them correspond with those of monovalent compounds like hydrochloric acid or sodium hydrate, to take only half their atomic weight in grams per liter, or in general terms, such a quantity as will react with I monovalent atom such as H, Na, or Cl. Such solutions are called normal or N/i solutions. Solutions of one-tenth this strength are called decinormal or N/io. TABLE GIVING THE QUANTITY OF THE FOLLOWING SUBSTANCES CONTAINED IN, OR EQUIVALENT TO, i LITER OF N/i OR 10 LITERS OF N/io SOLUTION. Sulphuric acid . . H 2 SO 4 mol . 49*0 grm. (= 48 grm. SO 4 ). Hydrochloric acid . HC1 I = 36-5 ,, (= 35-5 grm. Cl). Nitric acid . . . HNO 3 i ,, = 63-0 ,, Sulphurous acid . . H 2 SO 3 ,, = 41*0 ,, Oxalic acid . . . H 2 C 2 O 4 . 2Aq & = 63-0 * Kuster's " Logarithmische Rechentafeln fur Chemiker," Bejt and Co., Leipzig, is very complete. The use of the slide-rule saves much time in calculation, and is mostly sufficiently exact. GENERAL METHODS. Formic acid . . . HCOOH Acetic acid'. . . CH 3 COOH Lactic acid . . .C 2 H 4 OH.COOH Sodium hydrate . . NaOH Sodium carbonate Na 2 CO 3 (anhydrous). Sodium carbonate Na 2 CO 3 . loAq (crystallised). Potassium hydrate . KOH Magnesia . . . MgO Baryta . .... * / BaO Barium hydrate . . Ba(OH) 2 . 8Aq Barium chloride . BaCl 2 . 2Aq Calcium oxide (lime) CaO Calcium hydrate . Ca(OH) 2 Calcium carbonate . CaCO 3 Ammonia . . . NH 3 Zinc Zn Silver nitrate . . AgNO 3 Potassium permanga- KMnO 4 nate. Potassium bichromate Chromic "acid (anhydride) Iodine .... Bromine .... Sodium thiosulphate i mol. = K 2 Cr 2 O 7 Cr0 3 I Br Na 2 S 2 3 . 5 Aq 46*0 grm. 60*0 ,, 90-0 ,, 40-0 53'0 56-1 40-2 76-7 I57-7 122' I 28-1 37'I 5' 17-0 32-7 170*0 31-6 33'4 126-9 80-0 248-2 (insoluble). (sol.2-89p.c.ati5). (sol. O'i2p.c.). (insoluble). (= (=35'5 Cl). As oxidising agents. Do. Do J(CrO 3 = loo- 1) Uo -i(Cr,0,= 152-2) Do. Do. in reducing I and Br. These figures are corrected to the International Table of atomic weights, taking O = 1 6, but decimals less than 0*3 may for most practical purposes be disregarded, and in some cases the atomic weights themselves are uncertain to this extent. SECTION II. ALKALIMETRY. ONE of the most important applications of volumetric analysis is the estimation of acids and bases which unite together to form salts. Thus nitric acid and sodium hydrate react to form the neutral salt sodium nitrate, according to the following equa- tion : NaOH + HNO 3 = NaNO 3 + OH 2 40 + 63 = 85 -f 18 It is obvious that if the quantity of nitric acid necessary to combine with the whole of the soda is known, the amount of the 1 6 THE LEATHER INDUSTRIES LABORATORY BOOK. latter can be calculated, and vice versa. Since both free acids and free bases have marked properties which are not shared by their salts, the completion of the reaction can be determined with great accuracy, usually by the aid of " indicators," which are generally substances which show a marked colour-change on passing from the alkaline to the acid condition, or vice versa. According to the modern view, the acids owe their distinctive character as a class to the fact that in dilute solution they become partially dissociated into H and another " ion," differing with the acid, but in the case of nitric acid NO 3 . Similarly, the alkalies dissociate into the ions OH and the alkali-metal (or some group such as NH 4 which takes the place of a metal). Their " strength " or chemical activity depends on the extent to which they are dissociated or ionised in solution, the stronger acids and bases being so dissociated to the extent of about 90 per cent, in decinormal solution, while the weaker ones are dis- sociated very little, the extent of dissociation in all cases in- creasing with increasing dilution. Thus the acids owe their common properties of sourness, etc. to the presence of the ion H in their solutions, and the bases to that of OH, and the most important point in the reaction of acids and bases is the combination of these two ions to form water, as shown in the equation, since the salts really remain dissociated in the solution, and only combine to form a single substance as the water is evaporated.* Indicators are mostly weakly acid organic colouring matters, the salts of which have a markedly different colour from the free acid. (Weak colour-bases may also be employed, but are not so commonly useful.) In the presence of a stronger free acid, the salt of the acid colouring matter is decomposed, and the colour- acid liberated ; and similarly in the case of basic indicators, the free base is liberated by a stronger base. Methyl Orange (Helianthine, Tropaeolin D) is the sodium or ammonium salt of dimethylamidoazobenzenesulphonic acid C 6 H 4 . N (CH 3 ) 2 . N : N C 6 H 4 . SO 3 H. The free colour-acid is red (or in the solid form violet), and its salts are yellow, and their solutions change from yellow to red in the presence of an * Those who are interested in the modern theory of salt-formation are referred to Ostwald's excellent little book, " Die Wissenschaftliche Gruhdlage der Analysis- chen Chemie" (Leipzig), or its English translation. ALKALIMETRY. 17 acid sufficiently strong to set the colour-acid free. As this acid is of medium strength, this change is only pronounced in the presence of the stronger mineral acids, and '5 is scarcely affected by carbonic, boric, sulphydric, arsenious, silicic, oleic, stearic and many other weak inorganic and organic acids. Thus methyl orange is quite useless for the titration of even the stronger organic acids, such as acetic and lactic, with which it gives no definite end-reaction, and their influence is still further diminished in alcoholic solution, probably because of their diminished ionisation in this medium, and it is said that it may be used for titrating sulphuric or hydrochloric acids in the presence of free acetic, if a sufficiency of alcohol is added.* Except for such special purposes the presence of alcohol is to be avoided, as diminishing its sensitiveness. On the other hand, it is very suitable for titrating weak bases, giving sharp reactions with ammonia, and with most of the alkaloids. The presence of neutral salts, and even of chlorides, sulphates and nitrates diminish the sharpness of the colour-change, and should be avoided as far as possible. Many salts of dibasic and tribasic acids are alkaline to methyl orange which are neutral to phenol- phthalein ; and thus by the use of both indicators, bicarbonates, normal carbonates, and indirectly caustic alkalies can be titrated in the same solution (p. 28), and sulphurous acid, bisulphites and sulphites in the presence of each other (p. 27). It is generally best used in the cold, and does not answer well in boiling solutions. It is destroyed by free nitrous acid. An aqueous solution of I grm. per liter is most suitable for general use, of which not more than four or five drops should be used per 100 c.c. of solution titrated, as the end reaction is less sharp with larger quantities. When decinormal solutions are used, the volume required to set free even this quantity of colour-acid is quite appreciable, and for' very exact work such as water analysis, an equal number of drops should be added to a similar volume of distilled water, and the amount of acid required to produce the colour-change noted, and deducted from that used in the actual titration, when the colour produced in the distilled water is matched by that in the liquid titrated. Phenolphthalein is a much weaker acid than methyl orange, * In the Author's hands this method of titration has proved quite unreliable. C 18 LEATHER INDUSTRIES LABORATORY BOOK. and is itself colourless, while its alkaline salts are intensely red. Hence it is extremely sensitive to weak acids, but only sharply affected by the strong bases, such as the alkalies and alkaline earths, and even ammonia gives with it no sharp end-reaction. Thus it is reddened by alkalies and alkaline carbonates, but not by hydric carbonates (bicarbonates), nor by the insoluble carbon- ates of the alkaline earths, nor by neutral sulphites, or salts of the fatty acids (soaps), and is useless with ammonia. In suitable cases, it is one of the most sensitive of indicators. It is best used in I per cent, alcoholic solution, of which two or three drops are required for titration in aqueous, and more in alcoholic solutions. Litmus is less generally useful than the above, except as a convenient test in the form of red and blue papers. It is red- dened by acids, blued by alkalies and alkaline carbonates, and turned an uncertain purple by carbonic acid, but can be used for the titration of carbonates in a boiling solution. Lacmoid changes colour like litmus, and may be employed either as a I per cent, alcoholic solution or as red or blue papers. The papers may in almost all cases be used in place of methyl orange, being practically insensitive to carbonic and boric acids. The solution is more affected by these acids, but may be used with carbonates in boiling solution. Congo Red is turned blue by acids. Its range of sensitive- ness is very similar to methyl orange, though the change is less sudden. Notwithstanding this, it is sometimes useful by artifi- cial light, or in liquids of which the colour makes methyl orange inapplicable. It may also be used in the form of test-papers. Alizarin is a useful indicator in certain special cases, such as the estimation of hardness in water (p. 33), and yields a violet solution in presence of alkalies, turning very sharply to yellow on addition of acid. J per cent, of the commercial paste diluted in water is a suitable strength for use. Standard Solutions. It is obvious from what has been already said that if we have a solution of any suitable acid of exactly known strength, we can determine that of any base by measuring the quantity required to neutralise it, or similarly we may start with a known solution of a base, and with it determine that of any acid ; and, of course, by using the acid solution of which the strength is thus ascertained, we can determine that of ALKALIMETRY. 19 any other base. Evidently, the substance selected as our starting point must be one not liable to change, and which can be exactly weighed, and which is of perfectly constant composi- tion, and it is also desirable that it should be either a strong base, or a strong acid, since with these the indicators react most sharply. Many substances have been proposed for the purpose ; but of these, pure normal dry sodium carbonate is one of the most suitable, and it will be adopted in the following pages. Great pains should be taken to insure its absolute purity, since upon this depends the accuracy of all work done with the standard solutions prepared by its aid. Solutions of any strength may be used in alkalimetry, so long as that strength is accurately known, and it was formerly customary to make special solutions for particular purposes, so that for example, I c.c. corresponded to I per cent, of the sub- stance to be determined. Reasons have been given on p. 14 why it is generally more convenient that they should contain one combining equivalent of the substance weighed in grams in I liter of solution, or some definite fraction of this strength ; and calculation can usually be avoided where many analyses have to be done, by weighing out a suitable quantity of the material of which the strength is to be determined. Starting with sodium carbonate, we may either proceed to dissolve such a quantity as will make exactly I liter of " normal " strength, or we may use the dry carbonate in weighed portions to determine the exact strength of an acid solution, from which we can afterwards standardise any alkaline solutions required. The former is the simpler in theory, and shall be first described, but the latter will usually be found easier and more accurate in practice, and less liable to error. Normal Sodium Carbonate Solution direct. About 60 grm. of purest anhydrous sodic carbonate is heated over a Bunsen burner till the bottom of the basin shows faint redness, and is cooled in the desiccator.* A weighed platinum or porcelain basin may be used. Weights to 53 grm. and the weight of the basin are now placed on the balance, and the basin is rapidly brought to within O'Oi grm. of the same weight by removing as * A still more accurate method is to dry for some hours at 160 C. in an air- oven, but no appreciable quantity of carbonic acid is expelled by the method given in the text. C 2 20 LEATHER INDUSTRIES LABORATORY BOOK. much of the sodic carbonate as is in excess.* A beaker with about half a liter of water is placed on a sheet of glazed black paper, and the sodic carbonate added little by little, constantly stirring with a glass rod till dissolved. The solution is poured into a liter flask, and the beaker repeatedly rinsed with distilled water, which is added to the flask till it is filled nearly to the mark. The temperature is then tested with a thermometer, and brought as accurately as possible to 15 C, and the solution is made up exactly to the mark by water added from a pipette, and mixed in the flask as completely as possible, and finally poured into a clean dry stoppered bottle of somewhat more than one liter capacity, and shaken vigorously for five minutes. Normal Hydrochloric Acid. no to 120 c.c. of the strongest hydrochloric acid is diluted to I liter, and well shaken in a stop- pered bottle. A burette is filled with this dilute acid, 10 c.c. of the normal sodic carbonate is placed in a beaker, 2 drops of methyl orange solution is added, and then acid from the burette drop by drop till the solution just changes to orange ; the solution being constantly stirred with a glass rod during the process. This is repeated several times, the acid being added rapidly to near the required quantity. The average of the readings is taken, and the number obtained multiplied by 100 gives the number of c.c. of the diluted acid required to make I liter. The quantity of distilled water required to make this up to 1000 c.c. is then accu- rately measured into a dry liter-flask, which is very carefully filled up to the mark with the dilute acid. After thorough mixing in a dry bottle, this should again be tested against the normal sodic carbonate, of which it should now neutralise an exactly equal quantity. The solutions should be kept at a temperature of 1 5 C. during the process. In place of using an accurate sodium carbonate solution, an alternative method of fixing the strength of the standard acid is to fill a well-stoppered weighing bottle with hot and carefully * To do this rapidly it must be done systematically. A portion of the material is removed on a spatula, and the remaining weight tested by releasing the balance. If still in excess, the portion on the spatula is rejected, and the process repeated till the weight on the balance proves less than that required. We now know that the true weight lies within the portion on the spatula, which is gradually shaken back till the weight is again in excess. The process is now repeated by removing a much smaller portion, with which a still closer approximation is reached ; and so on until sufficient accuracy is obtained. ALKALIMETRY. 21 dried pure sodium carbonate, and after cooling in a desiccator, to shake out into a beaker a portion conveniently of about half a gram, of which the weight is determined by weighing the bottle before and after. This portion is dissolved in water, and titrated with the dilute acid of which the strength is to be determined ; of which the number of c.c. required to make I liter of normal acid is found by multiplying the c.c. of acid used by 53, and dividing by weight of sodium carbonate taken. The average of several determinations which should agree very closely, is taken as correct, and is made up to a liter as above described. See also note, p. 40, for determination by silver nitrate. Normal Sodium Carbonate from Hydrochloric Acid. A solu- tion of sodium carbonate somewhat over normal strength (say, 60 grm. per liter of ordinary air-dry anhydrous sodium carbonate or 1 60 grm. of pure soda crystals) is placed in the burette, and titrated upon 10 c.c. of the normal acid. The calculation is made precisely as in the case of standardising hydrochloric acid, the average of several determinations, multiplied by 100, giving the volume of the soda solution in cubic centimetres, which must be made up to I liter for a normal solution. Normal Sodium or Potassium Hydrate, is made precisely as above by standardising against normal hydrochloric or sulphuric acid ; 45 grm. of the purest NaOH, or 60 grm. of KOH being suitable quantities for I liter of the rough solution. To exclude carbonic acid, caustic standard solutions must be kept under petroleum oil (which should be previously well washed with soda solution, and afterwards with water) ; and when required for use, drawn off by a siphon passed through the cork of the bottle and fitted with a piece of rubber tube and a pinch-cock ; or the cork may be tightly fitted with a tube filled with soda-lime, or absorp- tion-bulbs with sodium hydrate solution to exclude carbonic acid. Barium Hydrate, Calcium Hydrate. Where absolutely caustic solutions are important, these may be usefully employed as their carbonates are insoluble, but they absorb carbonic acid and lose strength rapidly on exposure to the air. A saturated lime solu- tion is only of about N/2O strength (cp. p. 54), and is often useful, as at a regular temperature it is practically of constant strength. Stronger solutions up to nearly normal strength may be made by the use of sugar solution (cp. pp. 54, 232). Barium 22 LEATHER INDUSTRIES LABORATORY BOOK. hydrate is much more soluble than calcium hydrate, a saturated solution being over N/3, so that any lower strength, such as N/5 can be easily prepared by dilution. If saturated solutions are employed, as they vary in strength with temperature, the quan- tity required should be filtered off immediately before use and the strength determined against standard acid. Supposing that 10 c.c. of N/ 1 o acid require 12 c.c. of the barium hydrate solution and 10 c.c. of the unknown acid titrated requires 15 c.c., its value in N/io acid is given at once by the proportion 12 : 15 = IO c.c. : 12* 5 c.c. Normal Sulphuric Acid is made exactly as hydrochloric acid, but using only about 35 c.c. per liter of the concentrated acid (sp. gr. I 840). The acid should be poured into the water, and not vice versa. It has the advantage that its solution can be boiled without loss by volatilisation. Other Normal or Decinormal Acids can be similarly prepared, but in the case of organic acids, or the weaker mineral acids, phenolphthalein and caustic alkaline solutions must be used for standardising, instead of methyl orange and sodium carbonate. N/i and N/io oxalic acid may be made with considerable accu- racy by dissolving 63 grm. or 6*3 grm. respectively of the purest recrystallised acid to I liter. Before weighing, the acid should be powdered, and allowed to dry for an hour or so exposed to the air at ordinary temperature. The solutions do not keep very well, and the normal solution will deposit crystals if cooled much below laboratory temperature. Oxalic acid may be very approximately titrated with methyl orange, if a few drops of neutral CaCl 2 solution be added at the end of the titration to liberate hydrochloric acid by precipitation of calcium oxalate. Decinormal Solutions may be made from the normal, by measuring 100 c.c. into a liter flask, and filling up with distilled water. It is well to verify after dilution against a solution of known accuracy, or a known weight of substance Testing of Commercial Materials. The following examples will give an idea of the use of alkalimetrical methods in practical testing, but many other applications will suggest them- selves to the chemist. Sulphuric Acid. IO grm. may be made up to 100 c.c. and ALKALIMETRY. well mixed, and of this 10 c.c. may be tested with normal sodic hydrate or carbonate in presence of methyl orange. Each I c.c. of soda solution used corresponds to 0-049 grm. or 4*9 per cent, of H 2 SO 4 . For most purposes, the strength may be ascer- tained from the specific gravity, as measured by a hydrometer or weighed in a specific gravity bottle by the use of the following table : SULPHURIC ACID. (Lunge and Isler.) Specific Gravity z ^ in vacuo. H 2 S0 4 per cent. H 2 SO 4 kilos, per liter. Specific Gravity in vacuo. H 2 S0 4 per cent. H 2 SO 4 kilos, per liter. I 'GOO O'C>9 CTOOI I-40 50-11 0*702 I '01 I'57 "016 i'45 55-03 79 8 I -O2 3'03 031 1-50 59-70 8 9 6 I-0 3 4'49 046 i'55 64-26 99 6 I '04 5-96 062 i 60 68-51 1-096 1-05 7*37 077 1-65 72-82 I '2O2 I -06 8-77 093 1-70 77-17 I-3I2 1-07 10* 19 109 1-75 81-56 I-427 I -08 1 1 -60 125 1-80 86-90 I-564 I-0 9 I-IO 12-99 I4-35 142 1-81 158 1-82 88-30 90-05 I*598 1*639 fiS 20*91 239 1-83 92-10 1-685 I'20 27-32 328 1-84 95'o i*759 1-25 33*43 418 1-8415 97'7o 1-799 I- 3 39'i9 510 1-8385 99*95 1-838 i*35 44-82 605 Note. The table is reckoned for the gravity of acid at 15 C., compared to the (maximum) gravity of water at 4. To calculate to water at 15 multiply by 1-000841 the difference may generally be neglected. It will be noted that the gravity varies very slowly with increasing strength as the acid approaches loo per cent., and -actually falls over 97-7 per cent., so that for very strong acid, gravity determinatiou is of little value. In order to calculate values for figures between those given in this and the following tables, it is generally sufficiently accurate to assume it as proportional to the difference between the next lower and the next higher value. Thus, if it be desired to know the strength of an acid of S.G. I -525, the difference between that for I -50 and I -55 is 4-56, and the value required is 59-70 + 4-56 x ^? or 61-98 per 0^050 cent., the complete table giving 62' 06. Full and accurate tables for most acids and alkalies and many salts are given in the Chemiker Kalender published annually by Springer, Berlin. Degrees of Twaddell's hydrometer may be reduced to specific gravity by multi- plying by 0-005 ancl adding I -, thus 10 Tw. = I -050 sp. gr. LEATHER INDUSTRIES LABORATORY BOOK. As tested by alkalimetry, it is quite possible for an acid to show over 100 per cent. H 2 SO 4 from the presence of anhydride (SO 3 ) as in " fuming " or " Nordhausen " sulphuric acid, now often made by direct addition of anhydride. The impurities of sulphuric acid most common and injurious for tanning purposes are iron and nitrous acid. . Iron is detected on neutralising with soda or ammonia, when it falls as a yellowish precipitate, which may be recognised by the ordinary tests (p. 40). Nitric and nitrous acids are detected by pouring a strong solution of ferrous sulphate cautiously on the top of the strong cold acid in a test-tube held slanting, when a dark ring is formed at the junction of the two liquids. Sulphuric acid fre- quently contains arsenic in considerable quantities and sometimes other metals. Hydrochloric Acid may be tested with soda solution like sulphuric. I c.c. of normal soda = 0*0365 grm. or 3*65 per cent. HC1. Its strength may also be calculated from the sp. gr. by the following table. It may be noted that degrees Twaddell correspond very approximately to the percentage of actual HC1. A yellow coloration of the acid generally indicates the presence of iron, which may be confirmed by the usual tests. HYDROCHLORIC ACID. (Lunge and Afarchlewski.) Specific Gravity i^. in vacuo. HCI per cent. HCI kilos, per liter. Specific Gravity IC . - in vacuo. 4 HCI per cent. HCI kilos, per liter. 1-000 O'l6 0-0016 I *II 21-92 O'O24 I'OI 2-14 022 1-12 23-82 267 I '02 4*13 042 I-I 3 25-75 291 I-0 3 6-15 064 1-14 27-66 315 1-04 8-16 085 1-15 2 9 '57 340 I 05 10*17 107 i"i6 31*52 366 1-06 12-19 129 1-17 33 '46 392 1-07 14-17 152 1-18 35*39 418 I -08 16-15 174 1-19 37'23 '443 1-09 18-11 197 i '195 38-16 456 I -10 2O'OI 22O 2'OOO 39-11 469 !COOH ) COOH 2 A [ shou1 ^ be pure white ALKALIMETRY. 25 and entirely soluble in distilled water. 6*3 grm. may be weighed out, and made up to 100 c.c. If 10 c.c. of the solution for a test be used, each c.c. of normal soda solution equals 10 per cent, or 0*063 grm. of pure crystallised acid, C 2 O 4 H 2 + 2Aq. The end- reaction with methyl orange is rendered sharper by the addition of a few drops of neutral calcium chloride towards the end of the titration ; but it is preferable to use caustic alkali and phenol- phthalein. Acetic Acid, CH 3 . COOH, maybe similarly determined, each c.c. of normal alkali being equivalent to 0-06 grm. of C 2 H 4 O 2 , but caustic soda, or lime-water, and phenolphthalein should be used. Brown pyroligneous acid is difficult to test because of the dark compounds formed with soda, but may be indirectly deter- mined by the quantity of marble, baric carbonate, or magnesia which it will dissolve. Mineral acids in acetic acid, vinegar or other organic acid solutions may be detected and determined by adding excess of normal soda or sodium carbonate solution, igniting till the sodium salt of the organic acid is decomposed, and titrating back with N/i acid, and methyl orange. Any loss of alkalinity is due to mineral acid since salts of all organic acids of the RCOOH type leave carbonates on ignition (Hehner). Methyl orange is not reddened by acetic acid in presence of a sufficient quantity of alcohol, nor by small quantities of free acid in presence of acetates. The addition of alcohol does not prevent the action of the stronger acids, so that they can be titrated independently of acetic in a partially alcoholic solution.* Formic Acid, HCOOH, is determined like acetic acid ; each c.c. of N/i NaOH corresponding to o * 046 grm. of acid. It is now made synthetically by passing CO into soda-lime or concentrated KOH solution at 100 C. and subsequent decomposition of the formate by sulphuric acid. The commercial acid should contain about 65 per cent. Lactic Acid, C 2 H 4 OH . COOH, made by bacterial fermenta- tion of starch or glucose, may be titrated in the cold with N/i (or N/io) NaOH and phenolphthalein after dilution with water to about 10 per cent, 20 grm. of the ordinary commercial acid being made up to 100 c.c. and 50 c.c. titrated. The operation should be conducted as rapidly as possible, and the first perma- * See note, p. 17. 26 LEATHER INDUSTRIES LABORATORY BOOK. nent pink shade taken as the end-point, to avoid acting on the lactic anhydride which is usually present. I c.c. of normal alkali corresponds to 0*090 grm. of lactic acid. To determine the lactic anhydride, 25 c.c. of N/i NaOH is added to the solution, neutralised as above, and the mixture is boiled for twenty minutes in a flask with reflux condenser or a long tube fitted in the cork to prevent loss of water. It is then titrated back with N/i acid. Each c.c. of soda neutralised (or of acid less than 25 c.c. used) represents 0*08 1 grin, of lactic an- hydride, or 0*090 of lactic acid existing in this form. Lactic anhydride is practically useless for the removal of lime or the swelling of hides. The boiling should not be prolonged beyond twenty minutes or other organic matters present may be acted on by the caustic alkali. The only serious impurity for tanning purposes which is likely to be present is iron, which tends to injure the colour of the leather. Mineral ash may be determined by evaporating a known weight of the acid, first on the water-bath and then over a small flame with a cone of filter-paper inserted in the acid, and subsequently raising the temperature till the ash is free from carbon. The ash is then taken up with hydrochloric acid, evaporated (in porcelain basin) with a little nitric acid to small bulk, diluted with water to a known volume, and the iron is determined colorimetrically as described on p. 40. Sulphuric, oxalic and hydrochloric acids and their salts may be detected in the original solution by the usual reagents, and free mineral acids by evaporation to dryness with a measured excess of Na 2 CO 3 solution, cautious ignition and subsequent titration of the sodium carbonate formed. Any deficiency will be due to the presence of free mineral acids, which are often present. Compare p. 367. Boric {Boracic) Acid and B orates. See p. 74. Caustic alkalies may be estimated with N/i or N/io, HC1 or H 2 SO 4 , using either methyl orange or phenolphthalein as indica- tors, i c.c. of N/i acid corresponds to 0*040 grm. of NaOH or 0*0561 of KOH. Ammonia may be similarly estimated, but methyl orange and not phenolphthalein must be used as indicator. Cp. p. 64. Sodium and potassium may be estimated in carbonates, sulphides, borates and the salts of some other, weak acids by the use of standard hydrochloric or sulphuric acid and methyl ALKALIMETRY. 27 orange, which is not at all affected by these weak acids. I c.c. of normal acid corresponds to 0-023 grm. Na and 0*0391 grm. K. Sometimes by taking advantage of the different sensitiveness of methyl orange and phenolphthalein to acids and bases, it is possible to estimate a free acid or base and a salt at once in the same solution, as in the following instances. Estimation of Sulphurous Acid and Sulphites, It has been noted (p. 1 7) 'that bisulphites (hydric sulphites) are neutral to methyl orange and acid to phenolphthalein. If, therefore, both indicators be added to a solution containing free sulphurous acid and a bisulphite, and lime-water or N/io caustic soda be added from a burette till the methyl orange becomes yellow, the whole of the free sulphurous acid will have been converted into bi- sulphite (half saturated with base), and each c.c. or N/io alkali will equal one-tenth of a milligram-molecule of SO 2 , or 6-4 mgr., thus : H 2 SO 3 + NaOH = NaHSO 3 + OH 2 . If the standard solution be again added till the phenol- phthalein becomes pink, the whole of the bisulphite, both that originally present and that formed from the free sulphurous acid, will be converted into neutral sulphite SO(NaO) 2 , and again each c.c. of N/io soda used will equal 6-4 mgr., or one-tenth milligram- molecule of SO 2 present as bisulphite. Deducting from this an amount equal to that estimated as free acid, the bisulphite originally present is obtained. If the solution of sulphurous acid is at all a strong one, the titration must be done very rapidly in a narrow-necked flask to avoid loss by vaporisation, or the acid solution may be put in the burette and run into a known measure of caustic soda solution, when of course the order of the indica- tions will be reversed. If instead of a mixture of free acid and bisulphite, a mixture of hydric and neutral sulphite is to be estimated, the hydric sulphite may be first titrated with caustic soda as described, and then the resultant neutral sulphite titrated with methyl orange and N/io hydrochloric acid till the latter is reddened. Each c.c. of N/io acid will then be equivalent to 6'4 mgr. of the SO 2 present as neutral sulphite, from which that found as bisulphite must be deducted. 28 LEATHER INDUSTRIES LABORATORY BOOK. As an example, the following analysis of a sulphurous acid solution, which has been used for deliming, may be given, lime- water being used for titration. 10 c.c. of N/io HC1 = 21 3 c.c. lime-water. IO c.c. sulphite liquor with) ..u i } = 9*3 methyl orange. ( 10 c.c. sulphite liquor with) _ 2Q>2 phenolphthalein. J ~ Then 9'3 X 6-4 = 2 . and 97 grm. per liter of SO 2 free - 2 *97 =4-10 grm. per liter of SO 2 as bisulphite. For other methods of estimation of sulphurous acid and sulphites, compare p. 73. Hydrates, Carbonates and Bicarbonates of the alkalies may be estimated in a similar manner. Carbonates of the alkalies are alkaline, and bicarbonates neutral to phenolphthalein, and both are alkaline to methyl orange. Thus a mixture of alkaline carbonate and bicarbonate may be titrated in dilute solution by cautious addition of N/io HC1 or H 2 SO 4 till the colour of phenolphthalein is discharged, when each equivalent of acid will equal two equivalents of base converted as under Na a CO s + HC1 = NaHCO 3 + NaCl. Acid is then added till methyl orange is reddened, and the whole acid consumed in both operations corresponds to the total base originally present in both forms. The result, however, is only accurate if the titration is done somewhat slowly and in dilute solution to prevent escape of carbonic acid. If a deep vessel is used, and a capillary tube attached to the burette is carried well below the surface, greater accuracy is obtained. Another method is as follows. Alkaline earthy carbonates are alkaline to methyl orange, and neutral, especially after boiling, to phenolphthalein. This is applied to the analysis of lime, p. 54, and may also be utilised to determine caustic alkali in presence of carbonate in soda ash, etc. The'. solution to be tested is boiled with excess of neutral calcium (or barium) ALKALIMETRY. 29 chloride in a flask, converting the alkaline carbonate into chloride with production of insoluble calcium or barium carbonate, and is first titrated with HC1 and phenolphthalein to determine the caustic alkali, and then with methyl orange for the car- bonate. In the latter titration it is convenient to add slight excess of acid, heat or digest for a short time to dissolve the carbonate, and titrate back with normal soda or sodic carbonate. Alkaline Chromates are neutral to phenolphthalein,* while bichromates are acid. The latter may, therefore, be titrated with caustic soda in presence of the former. For details see p. 259. In conclusion of this section, it may be pointed out that the possession of an accurate chemical balance is by no means neces- sary for many of the estimations just described. Standard solu- tions can be bought ready prepared, and from one accurate normal acid or alkali a great variety of other solutions can be adjusted by the use of measuring vessels only. 100 grm., or even J Ib. (= 113*4 g rm -) of tne material may be dissolved and made up to I liter, wellj mixed, and a measured quantity of the solution such as will bring it to the approximate strength required again diluted to I liter. Liquids, especially concen- trated acids, which are awkward to weigh without proper ap- pliances, may be measured, and either their strength per liter determined, or their weight calculated from their specific gravity. Thus 100 c.c. of sulphuric acid of 188 Tw. (S. G. 1-84) will weigh 184 grm., and, made up to I liter, will be about 3j times normal strength. * The purest potassium chromate slightly reddens phenolphthalein in dilute solu- tion, owing to hydrolytic dissociation ; but this immediately disappears on the addition of a trace of acid or of bichromate, and its amount is too small to practically affect the accuracy of the titration. Similar effects are common with salts, either of weak acids or weak bases, though they are frequently overlooked. 30 LEATHER INDUSTRIES LABORATORY BOOK. SECTION III. TECHNICAL WATER ANALYSIS. THE analysis of water for technical purposes usually only extends to its mineral constituents, the traces of organic impurity which are important in potable waters, having but little effect on most chemical and technical uses. Bacteriological examination has received too little attention from the technical chemist, since there are many manufactures in which the presence of organised ferments is of an importance only second to that of their exist- ence in drinking water ; and some hints on this point are given in the section on bacteriology, p. 427. The "hardness " of water is, on the whole, of the greatest manufacturing moment, but has been considered too exclusively from the point of view of the soap user, which often differs widely from that of other manufacturers ; and although as an empirical method for determining the quan- tity of soap likely to be destroyed by a given water, the soap- test has its practical value, there are other and easier ways of learning with much greater exactitude, what the actual qualitative and quantitative character of the hardness is. If no other fault existed in the soap-test than that it became practically useless in the presence of much magnesia, it would be sufficient to condemn it for general purposes, but when the somewhat indefinite char- acter of the reagent, and the still more indefinite character of the reaction is considered, it becomes unnecessary to labour the subject. It is now possible, by simple volumetric methods, and with a high degree of accuracy, to determine not merely tempor- ary and permanent hardness, but the actual quantities of lime, magnesia, iron, carbonic and sulphuric acids, and chlorine con- tained in a water, and with this information, the effect of the water on any chemical reaction may in most cases be completely foretold, while the labour involved is much less than that of a complete gravimetric analysis. It must, however, 'be pointed out that a water with " total hardness " equal to 50 parts of calcium TECHNICAL WATER ANALYSIS. 31 carbonate per 100,000 is a very hard one, but looked at as a volu- metric solution it is only centinormal, and must be determined with an accuracy which should reach l/5Oth of its strength, and to do this successfully involves many precautions which are quite unnecessary in the use of ordinary normal solutions. Temporary Hardness. The credit of having first introduced a practical and scientific method of hardness determination be- longs to Hehner, who titrates the temporary or bicarbonate hard- ness with N/io HC1, using methyl orange as an indicator, which is practically insensitive to carbonic acid. The method gives very exact results if certain precautions are taken. Methyl orange is the sodium salt of a colour acid of moderate strength, and the change from the yellow salt condition to the red colour of the free acid * marks the end point, which is sharp and exact when working with strong mineral acids, and with normal solu- tions. Even in this case it is desirable to use the smallest pos- sible amount of the indicator, but, in working with N/io solutions,, the amount of acid required to completely decompose the colour salt becomes very perceptible, and the change from yellow to red is not instantaneous, but passes through orange to pink with the consumption of an appreciable amount of acid. Thus it was found that, using a 10 grm. per liter solution of the indicator in 25 c.c. of water freed from carbonic acid by previous boiling, the following quantities of N/io HC1 were required to produce a clear pink : 8 drops of methyl orange solution = 1*5 c.c. ; 4 drops = o'Sc.c. ; 2 drops =0*5 c.c. As even O'5 c.c. in titrat- ing 100 c.c. of water would correspond to 2 5 parts of hardness per 100,000 and there is always a question as to what particular colour corresponds to the neutral point, the following procedure may be recommended. To 100 c.c. of distilled water, one drop or some other definite quantity of the indicator is added, and titrated, with N/io HC1 or H 2 SO 4 , to orange, or to the tint to the change of which the eye of the individual operator is most sensitive. The water of which the hardness is to be determined is similarly titrated with the same quantity of indicator, and in a similar beaker, until it exactly matches the distilled water, and from the amount of acid so used the quantity is deducted as a * A recent view is that the colour substance acts as an acid in presence of bases, and a base in presence of strong acids ; but this point is only of theoretical import- ance. 32 LEATHER INDUSTRIES LABORATORY BOOK. correction which was required to produce the same colour change with distilled water only. The results so obtained accurately correspond with those got by using alizarin as an indicator in boiling solution, though in the latter method the end reaction is sharper. It may be noted that methyl orange is not absolutely unaffected by carbonic acid, a somewhat crocus-yellow being attained instead of the lemon-yellow reached with pure boiled water, but the difference is insufficient to interfere with its satis- factory use as an indicator. Each c.c. of acid used per 100 c.c. of water represents 5 parts of CaCO 3 , or 2 8 parts CaO, 2 parts Ca, and 3 parts combined CO 3 per 100,000. Hardness is often expressed in " degrees." French degrees are parts of CaCO 3 , German of CaO, in 100,000, while English were formerly grains of CaCO 3 per gallon (70,000 grs.). It is best, therefore, to avoid the use of the term and state simply as parts CaCO 3 per 100,000. In place of methyl orange, alizarin paste may be used, with the precautions detailed in the following paragraph. Permanent Hardness. Hehner's method for the determination of permanent hardness is less satisfactory than that of temporary. It consists in evaporating 100 c.c. of the water to dryness with a known excess, say 20 c.c., of N/io sodium carbonate solution, taking up the soluble matter with cold distilled water, filtering off the precipitated calcium carbonate and magnesia on a small filter, washing the precipitate with cold water and titrating back the excess of sodium carbonate in the filtrate with methyl orange or rosolic acid as indicator. With lime-hardness only, and with the precautions above described, the method may be pronounced fairly satisfactory ; with magnesia, it is well not merely to eva- porate to dryness but to slightly heat the residue to thoroughly decompose any magnesium carbonate present, and even then, the washing should not be excessive, as calcium carbonate is soluble to the extent of 3 parts per 100,000, and magnesia to about 2*5 parts. A more accurate, as well as a more rapid method is to employ a fair excess of sodium carbonate, and to make up the solution to a known volume, say 100 c.c., and pipette off an aliquot part for titration, as the presence of excess of sodium carbonate materially reduces the solubility both of calcium and magnesium carbonates. Both these methods, how- ever, should be superseded, where really accurate work is re- TECHNICAL WATER ANALYSIS. 33 quired, by those introduced by Pfeifer and Wartha.* That for the determination of temporary hardness is identical with that of Hehner, except that, in place of methyl orange, a drop of a mixture of about I grm. of the purest alizarin paste in 200 c.c. of distilled water is employed. This indicator is surprisingly sensitive ; even more so than phenolphthalein, but as it is unfortunately affected by carbon dioxide it is necessary to complete the titration at a boiling temperature. The change is from violet in alkaline solution (perhaps slightly varying in shade with the nature of the particular base present) to- a perfectly clear pale lemon-yellow when neutral or acid. The titration of the water should be done with N/io HC1 or H 2 SO 4 in a silver, platinum, or hard porcelain basin. The acid should be added in the cold till the violet shade gives place to a clean yellow, and the liquid then brought to a boil, when, with the escape of carbonic acid, the violet colour will return, and should at once be destroyed by the addition of another drop of the acid, and so on, until no further change of colour takes place. It is undesirable to boil the indicator long, especially in an alkaline condition, as a violet deposit is formed on the sides of the basin, presumably of calcium and magnesium alizarates, which can only be dissolved by excess of acid, and is thus apt to cause percept- ible errors. In place of titrating to exact neutrality, the acid may be added in very small excess, and the whole of the liberated carbon dioxide boiled off at once, and the solution then brought back to neutrality by N/io NaOH, the solution boiled for a moment, and the titration completed. The results in either case are exact, a fraction of a drop of alkali changing the clear lemon-colour to a dirty yellow. If 100 c.c. of water are used, multiplication of the c.c. of acid by 5 gives the temporary hardness in parts of CaCO 3 per 100,000. The boiling must in on case take place in an ordinary glass beaker or flask, as an amount of alkali is dissolved which may lead to serious inaccuracy. Even hard Jena glass is not free from this effect, though the amount dissolved is so small that for most practical purposes it may be neglected. The following experiment will illustrate the point. 100 c.c. of distilled water * Zeits. angew. Chem., 1902, 198. Prof. Lunge states, in a private letter, that the use of alizarin for this purpose was originated in the laboratory of the Zurich Polytechnicum by Weithi. 34 LEATHER INDUSTRIES LABORATORY BOOK. boiled for an hour (with additions to maintain the volume) in a Berlin porcelain basin, showed no alkalinity or colour-change with alizarin ; in a Jena flask a perceptible change of colour was visible, but pure yellow was restored with one drop of N/io acid, while when boiled in an ordinary Bohemian flask, 0*4 c.c. of acid was consumed, and if the neutralised liquid were boiled further, it again became alkaline, and further additions of acid were re- quired, so that no coincident results could be obtained. With the precautions named, the results with a known solution of hydric calcic carbonate, containing only 5 * 5 parts of temporary hardness, and whether titrated alone or with additions of mag- nesium sulphate, were accurate within one part in 1 00,000, and experiments with other quantities were equally satisfactory. The manipulation is, however, considerably more difficult than in Hehner's method with methyl orange, so that for temporary hardness the latter is usually preferable. In the determination of permanent hardness, Pfeifer and Wartha, in addition to the use of alizarin as indicator, have in- troduced the important improvement of replacing the sodium carbonate of Hehner's method by a mixture of equal parts of N/io sodium carbonate and hydroxide solutions. While, as has been already explained, sodium carbonate perfectly precipitates calcium salts as carbonates on merely boiling, it becomes neces- sary to evaporate to dryness and to heat whenever any magne- sium salt is present, in order to convert magnesium carbonate into oxide, since magnesium carbonate is not sufficiently insolu- ble. In presence of sodium hydroxide, however, the magnesium carbonate is at once converted into magnesium hydroxide, and perfectly efficient precipitation is obtained by merely boiling for some time with sufficient excess of the reagent. A good excess, say 50 per cent, or more, is essential, not only because it is im- possible to say before analysis what proportion of sodium car- bonate and what of caustic will be required, but because the presence of the CO 3 ions of the sodium carbonate in the solution greatly lessens the solubility of the calcium carbonate, and similarly the OH ions of the sodium hydroxide lessen that of the magnesium hydroxide. Unless the water is extremely hard, 50 c.c. of the N/io mixed solution to 200 c.c. of water is a con- venient and sufficient quantity. The mixture, may be boiled until reduced within 200 c.c. in a platinum or porcelain basin, or TECHNICAL WATER ANALYSIS. 35 more conveniently, and with no material loss of accuracy, in a 300 c.c. Jena flask, but on no account in ordinary Bohemian glass. Even the Jena flask will become perceptibly etched at the water line if used repeatedly. The solution after cooling, is made up to 200 c.c. with distilled water in a gauged flask, and .allowed to stand till the precipitated bases have settled, and IOO c.c. is pipetted or siphoned off and titrated. As the quantity named corresponds to 100 c.c. of the original water, and 25 c.c. of N/io alkali, the difference between the acid actually used and 25 c.c. will correspond to the amount of alkali neutralised by the .acids of the permanent hardness, and multiplied by 5 will give the latter in terms of milligrammes per 100,000 calculated as calcium carbonate. The temporary hardness will also be pre- cipitated ; but, containing no fixed acids, will not interfere. Methyl orange may be used with advantage in place of alizarin, where the extremest accuracy is not required, but the precau- tions mentioned on p. 17 must be rigorously observed, and the standard solutions must be adjusted with the same indicator as is subsequently used in titration. In place of allowing the pre- cipitate to settle, the solution may be filtered through a small filter, which is carefully washed with the solution, of which the -first 50 c.c. or so is rejected, as filters are rarely absolutely free from acidity or alkalinity, and even if at first perfectly neutral, easily absorb acids or ammonia from the laboratory air, unless very carefully protected. Many irregularities occurred in the determinations until this source of error was detected. 15 cm. filters of three different makes were macerated with distilled hot water, and proved in all cases alkaline to methyl orange and acid to phenolphthalein, the difference between the two in- dicators, + or , amounting in each case to about 0*75 c.c. of N/io solution. Alkaline Carbonates. It occasionally happens that in the determination of permanent hardness, a larger quantity of acid is required to neutralise the mixture than corresponds to the volume of N/io alkali which has been added, and that therefore the permanent hardness would appear as a minus quantity. This somewhat puzzling result is due to the presence of sodium carbonate in the original water, which in this case can have no permanent hardness other than that due to the solubility of calcium carbonate, which cannot be removed by softening. D 2 36 LEATHER INDUSTRIES LABORATORY BOOK. Where sodium carbonate is thus found, a proportionate amount must be deducted from the temporary hardness, each part of hardness corresponding to 1*06 part of sodium carbonate, and each c.c. of N/io acid to 5 3 mgr. of Na 2 CO 3 . Determination of Magnesia. Since in the ordinary methods of water softening, lime is precipitated as carbonate, but mag- nesia as oxide, with the consumption of a double quantity of caustic lime or caustic alkali, it is impossible from hardness- determinations alone to calculate the materials required for softening, or the actual weights of the bases titrated, so long as it is uncertain whether or in what proportion magnesia is present. Pfeifer determines this in the following manner. 100 c.c. of the water is neutralised with N/io acid in presence of alizarin, in boiling solution, exactly as in the determination of temporary hardness, which may precede that of magnesia. A known quantity of clear limewater (25 or 50 c.c.) which should be at least 50 per cent, in excess of that required for pre- cipitating the magnesia present, is measured into a 200 c.c. flask, the hot neutralised solution is rinsed in with boiling distilled water free from carbonic acid, and made up with the latter to 5 c.c. above the mark to allow of contraction in cooling ; the flask is tightly corked or stoppered, and well shaken to mix (for which purpose the neck above the mark must be a long one), and set aside to cool and settle. Though not essential, it probably increases the completeness of the precipitation if the corked flask is heated for half an hour or so on the water-bath. I prefer to allow sufficient time for the liquid to completely clear, and to siphon or pipette off 100 c.c. to titrate back with N/io acid, which may be done cold with phenolphthalein, or hot with alizarin with equal accuracy. If a pipette is used it should be sucked with a rubber tube which can be pinched with the fingers, and the slight excess discharged into another vessel to avoid disturbance of the precipitate, while care is needed to avoid carbonation by the breath. Pfeifer filters, but in this case the strength of the limewater must be determined by a blank experiment conducted in exactly the same way with distilled water ; and it is better to reject the first 50 c.c. in each case to avoid error from want of neutrality of the filter-paper, and great care must be taken to filter rapidly, and to avoid possibilities of carbonation by the atmosphere, for which TECHNICAL WATER ANALYSIS. 37 purpose a suction filter with a perforated porcelain disc, covered with a neatly-fitted disc of filter-paper, answers well. If, on the other hand, the liquid is settled and pipetted, the risk of carbona- tion is so small that an equal quantity of the same limewater may be measured direct and titrated, using the same indicator as has been employed for the water, phenolphthalein in the cold being on the whole preferable. Deducting the N/io acid re- quired for the mixture of limewater and water from that employed for the limewater alone, and multiplying the difference by 5, gives the hardness due to magnesia in terms of milligrammes of calcium carbonate per 100,000, from which actual Mg may be reckoned by multiplying by 0*24; or MgO multiplying by 0*4. Carefully conducted, the method is extremely exact, its accuracy being quite equal to that of the determination of hardness, and probably superior to that of any gravimetric method for such minute quantities. The theory of the process is that, while calcium hydrate will precipitate magnesia, it has no action on lime salts ; and a good excess of lime serves not only to quicken the reaction, but to diminish the solubility of the magnesia. If iron is present it will /of course be reckoned with the magnesia, and should be determined colorimetrically with thiocyanate {p. 40), and deducted. It may be assumed that it is present in the ferric state, and that therefore 0*24 of Mg corresponds to '3733 f Fe. Aluminium, if present, would behave like iron, any traces of alumina dissolved by lime having no effect (on phenolphthalein at least), but it is rare that more than traces of alumina exist in natural waters, though it would have to be reckoned with in river-waters receiving manufacturing effluents, and its estimation would not be particularly easy. Possibly a colorimetric method with alcoholic extract of logwood or some other mordant dyestuff might be devised where the water was required for dyeing, but it is not likely that it would introduce any material error into water-softening calculations, and it would be removed with the other impurities. Calculation of Chemicals Required for Softening. Having determined the magnesia, or, more strictly, the acid with which it and any other bases are combined which are precipitable by lime, it becomes possible to calculate the calcium present in the water, by deducting the magnesia hardness from the total hardness, and calculating the remainder into Ca by multiplica- 38 LEATHER INDUSTRIES LABORATORY BOOK. tion by 0*4. When the proportion of hardness due to magnesia is known, it is possible to calculate the quantities of lime and sodium carbonate required for softening, since magnesium salts, as has been stated, cannot be satisfactorily removed as carbo- nates, but must be converted into hydroxides by lime or some other caustic alkali ; and this applies to the permanent hardness which is converted into carbonate by sodium carbonate as well as to the bicarbonate reduced to carbonate by lime. Thus each equivalent of magnesia present requires an additional equivalent of lime beyond that required by the corresponding calcium salt. Pfeifer gives a formula for this purpose, calculated for German degrees of hardness, which are reckoned in parts per 100,000 of CaO instead of parts of CaCO 3 , as is customary in France and England. I have, therefore, taken the liberty of transposing it into terms of parts of CaCO 3 per 100,000, and slightly modifying some points where it appears incorrect. Ht in the formula signifies temporary, and Hp permanent hardness, and Hm hardness due to magnesia, whether temporary or permanent. The quantities given are in milligrammes per liter, grammes per cubic meter, or Ib. per 100,000 gallons of the water to be treated. For complete softening 5 6 (Ht -f Hm} lime (CaO) and IO-6 Hp dry sodium carbonate ; or 28 '6 Hp = soda crystals (Na 2 CO 3 . ioH 2 O) are required. If hardness is to be softened by liming alone, the quantity required is 5-6 (Ht -f Hm). This will remove temporary hardness and permanent hardness due to magnesia, but not that due to lime. Finally, for softening with sodium hydroxide and sodium carbonate only, which is sometimes convenient for small boiler installations, we have 8 (Ht + Hm) = NaOH required ; 10*6 (Hp - Ht - Hm) = Na 2 CO 3 required. Consequently, if the water has less permanent hardness than the sum of the total and magnesia temporary hardness, it cannot be softened completely in this way without leaving excess of sodium carbonate in the water. In such cases barium hydrate may be used with advantage. If the quantity of CaO for lime-softening only is required, including that consumed by the free CO 2 , it may be obtained experimentally by treating the unneutralised and unboiled water exactly as has been described for the determination of magnesia, and in this case the lime actually consumed in milligrammes per IOO c.c. of water will represent the lime in Ib. required to soften TECHNICAL WATER ANALYSIS. 39 10,000 gallons. It must not be forgotten, however, that com- mercial limes are far from pure, and the " available lime " of the sample used should be determined as on p. 54. Commercial barium hydrate may be employed instead of lime to remove temporary hardness, reacting at the same time with the sulphates which are removed as insoluble barium sulphate. Its price and the poisonous character of barium salts are, however, serious obstacles in many cases to its employment. Free Carbonic Acid. Some waters contain large quantities of dissolved free carbon dioxide in addition to the "half com- bined " present as temporary hardness ; and though this is not included in any hardness determination, it, of course, combines with and renders useless an equivalent quantity of the lime added for softening, and must, therefore, be taken into account in reckoning the lime required. Estimation of Free Carbonic Acid (Arc/ibutt). N/io sodic carbonate is added slowly from a burette to 100 c.c. of the water, to which a little phenolphthalein has been added, till a faint permanent pink is produced. Each c.c. of N/io sodic carbonate corresponds to 2 2 mgr. of CO 2 . The reaction (Na 2 CO 3 4- CO 2 + OH 2 2 NaHCO 3 ), is accurate but somewhat slow, sodic carbonate being alkaline to phenolphthalein, while hydric sodic carbonate is neutral. The sample of water must be taken and preserved in a tightly closed and full bottle, as the carbonic acid readily escapes. For each c.c. of N/io sodic carbonate used on 100 c.c. of the water, 0*028 grm. of CaO per liter (or 0*028 Ib. per 100 gallons) will be required for softening in addition to that required for the hardness (see p. 38). Estimation of Chlorine. 100 c.c. of the water is titrated with N/io silver nitrate with potassic chromate as indicator. Each c.c. of silver nitrate corresponds to 3*55 mgr. of chlorine or 5*85 of sodic chloride. Decinormal Silver Nitrate is prepared by dissolving 17 grm. of pure recrystallised silver nitrate (one-tenth of a gram-equivalent of AgNO 3 ) in one liter of distilled water free from organic matter. The Potassic Chromate Indicator is a saturated solution of neutral chromate free from chlorine in distilled water. Its freedom from Cl is tested by slightly acidifying with nitric acid, and adding 40 LEATHER INDUSTRIES LABORATORY BOOK. a few drops of silver nitrate solution, when no precipitate should be produced. A few drops of the chromate solution are added to the water, and then silver solution from the burette till the pure primrose- yellow of the chromate is faintly tinged with brick-red, and remains so after vigorous stirring. This method is applicable to the determination of chlorine in all soluble chlorides. In the case of BaCl 2 the barium must previously be converted into sulphate by addition of sodium sulphate, or it will precipitate the chromic acid as barium chromate.* The solution in which Cl is determined by this method must in all cases be strictly neutral. Acid solutions are easily neu- tralised by shaking with excess of calcium carbonate or magnesia (free from chlorides) or a solution of sodium acetate may be added, the acetic acid liberated being under these conditions too feebly ionised to interfere. If, however, large amounts of free acid are present, it is wise first to approximately neutralise with soda solution. Color imetric Estimation of Traces of Iron. Evaporate 100 c.c. of water to dryness in a porcelain basin with one or two drops of dilute sulphuric acid, ignite gently, moisten with nitric acid, and evaporate to dryness on the steam bath. Redissolve in water (with a little nitric acid if necessary), and make up to 25 c.c. Mix 10 c.c. of dilute nitric acid (containing I c.c. concen- trated, sp. gr. i "35 to I '40) with 5 c.c. of a 5 per cent, solution of potassium thiocyanate (sulphocyanide) in a Nessler glass, add as many c.c. of the water-residue solution as will produce a convenient colour for comparison, and make up to 50 c.c. with distilled water. A second glass is made up with distilled water, and the same quantities of nitric acid and thiocyanide to about 40 c.c., and a solution containing O'OI grm. of ferric iron per liter is added from a burette till, after thoroughly mixing by stirring, the colour matches the first. The two glasses are placed side by side on white tile or paper. The comparison should be repeated several times, conveniently by adding say I c.c. more of the water solution and re-matching. The volume of standard solu- * The method is conveniently employed as a check on the accuracy of decinormal hydrochloric acid, and soda or sodium carbonate solutions. '. 10 c.c. of the acid, exactly neutralised by 10 c.c. of the alkaline solution, should require exactly 10 c.c. of the N/ 10 silver nitrate to produce the red tinge. \ TECHNICAL WATER ANALYSIS. 41 tion divided by (original) volume of water used gives parts of Fe per 100,000. The standard solution may be conveniently made by dissolving 0*496 grm. of pure crystallised ferrous sulphate in a little nitric acid in a flask, heating till no more red fumes are given off, making up to I liter with distilled water, and diluting the solution I : 10 before use. Considerable care is required to obtain nitric acid free both from nitric oxide and traces of iron. It is best diluted and distilled in a glass retort, and the pure dilute acid should be kept in a white glass bottle. As, however, the same quantity of nitric acid is used in both glasses, the presence of a trace of iron in it does not vitiate the result. Organic Impurities. While organic impurities even in traces may be of the greatest importance in potable waters, they are rarely of moment to the tanner, except so far as they are con- nected with the actual presence of injurious bacteria. The bacteriological examination is of too special a character to be usually attempted in the ordinary tannery laboratory, but if the necessary appliances are available it will be well to make plate- cultures (p. 445) on a gelatine bouillon or gelatine peptone medium, and especially to note any organisms which liquefy the jelly, and which are, therefore, likely to attack hide-substance. Actual pelt may be sterilised by treatment with chloroform or ether vapour, with subsequent free washing with sterilised water, or perhaps better by treatment with 2 per cent, caustic soda solution and subsequent neutralisation with a calculated quantity of HC1, and may then be used to test the effect of cultures of special bacteria. No bacteria or yeasts are known to destroy tannin, though they produce acid fermentations in liquors (p. 433). Such acid-forming bacteria are most likely to grow on wort or glucose gelatines of faintly acid reaction. While the determination of unorganised organic matter is rarely required in the analysis of tannery waters, it is not unfrequently important in connection with cases of alleged river pollution by tannery effluents. Apart from the colour, smell and putrescibility of the water, it is desirable to note its reaction with iron-salts, since traces of tannin of no practical importance give objectionable colorations with the traces of iron naturally present in many waters. A solution of iron-alum is the most suitable reagent. While the total amount of organic matter is not easily determined, apart from the rough results obtained by ignition of 42 LEATHER INDUSTRIES LABORATORY BOOK. the total solids, useful comparative figures are given by the amount of oxygen absorbed from a solution of potassium per- manganate. Many different methods of making this determin- ation have been proposed, yielding different results according to the concentration of the permanganate, and the temperature and time of its action. The following method of Kubel-Tiemann is one of the best, but to obtain comparable results, the directions given must be strictly adhered to. Pure distilled water is necessary for making up and diluting the solutions, as ordinary distilled water often reduces perman- ganate because of the volatile organic matters which it contains. To obtain a suitable water, a few crystals of potassium perman- ganate, and a little pure caustic soda (preferably made from metallic sodium and freshly fused) are added, and the distillation is carried out in glass apparatus, the first fourth which comes over being rejected. A solution of potassium permanganate is required yielding cr i mgr. of oxygen per c.c., and one of pure oxalic acid consum- ing the same amount. These are best made by diluting in each case 125 c.c. of accurately decinormal solution to I liter, the solutions containing 0*395 grm. permanganate, and 0-7875 grm, of pure oxalic acid respectively. The determination is carried out as follows : 100 c.c. of the pure distilled water, and 5 c.c. of the purest dilute sulphuric acid (1:3) are placed in a flask of about 300 c.c. capacity, and the temperature raised to a boil ; 2-3 c.c. of the permanganate solution is now added from a glass-tapped burette, the solution is boiled for ten minutes, and 10 c.c. of the oxalic acid is added and well shaken, and the titration with permanganate continued till a faint but permanent pink can be seen on holding the flask over white paper. Very approximately 10 c.c. of the permanga- nate should be required. An exactly similar determination is now made, using 100 c.c. of the water to be tested in place of the distilled water. The additional amount of permanganate consumed represents that employed in oxidising the impurities of the water, and each c.c. represents o* I part of oxygen consumed per 100,000, or i mgr. per liter. If the solutions of permanganate and oxalic acid did not accurately correspond when tested with distilled water, the value obtained must be multiplied by the permanganate required TECHNICAL WATER ANALYSIS. 43- for 10 cc. -of oxalic acid, and divided by 10. If more than 6-7 c.c. of permanganate are required to oxidise the impurities of the water tested, a less quantity than 100 c.c. must be taken, and the volume made up with distilled water, the result being, of course, multiplied by 100, and divided by the quantity taken. The presence of ferrous iron or nitrites, or great quantities of am- monium salts in the water may introduce errors, but a very short exposure to the air suffices to oxidise and precipitate iron present as ferrous carbonate, and the other impurities will rarely seriously interfere. More important from the sanitary standpoint than contamin- ation with tan-liquors, is that from old lime-liquors, puers, bates,, and drenches containing putrescible organic matter. In rivers containing fish, the presence of caustic lime is also specially objectionable, and is easily tested for with phenolphthalein, and if necessary estimated by titration. Calcium carbonate, which does not redden phenolphthalein, is practically harmless, soon settling as mud. Nitrogenous matter is generally estimated by the ammonia which it will yield, and medical officers of health are very fond of the old " albuminoid ammonia " test, from its apparent ease of execution. However useful this may be as a first test for nitro- genous impurity in potable waters, it is totally illusory and un- reliable for ordinary factory effluents, from the very small quantity of water which can be used and the consequent multi- plication of errors, as well as from the great variations arising from small differences in manipulation. A much more reliable method, in these cases, is that of Kjeldahl as modified for the purpose by Brown and Martin and others. For the outlines of this method, see p. 64. Complete Gravimetric Analysis. The volumetric methods already given are so accurate, if carefully carried out, that it is rarely necessary for the purposes of the tanner, to resort to- gravimetric analysis, but occasionally reasons arise which make it necessary to know all the mineral constituents and their quan- tity, and the following details are therefore given. For the general methods of manipulations required, Section I. must be consulted ; but unless considerable chemical experience has been gained, the fuller accounts of the regular analytical works should be studied. 44 LEATHER INDUSTRIES LABORATORY BOOK. The water, if turbid, must be well shaken before samples are drawn for any determinations, and 500 c.c. or I liter measured into a flask, and then filtered through a dried and tared quanti- tative filter paper (p. 1 3). The filter is washed with a little dis- tilled water, which is added to the filtrate, and is then dried in the water oven (or air oven at iO5-no C.) till constant, and weighed : the gain of weight is " mud and suspended matter." The filter is now ignited, and the residue weighed and entered as " inorganic suspended matter," and the difference from the weight of residue at 100 C. as " organic suspended matter." This is not strictly accurate, as some part of the loss may be due to combined water. The filtrate, or 500 to 1000 c.c. of the original water, if quite clear, is now evaporated to small bulk in a weighed platinum {or porcelain) basin over a small flame applied at one side of the basin to avoid spurting, the water being added in small quan- tities as it evaporates. When reduced to perhaps 50 c.c. the basin is placed on the steam bath and evaporated to dryness, and then dried at IOO to 1 10 C. till constant, cooled and weighed. The weight of the residue is " total solids." It is gently ignited and the loss entered as " organic matter and combined water," and the residue as " inorganic constituents." If the ignition has been at all strong, it is best to moisten with solution of ammo- nium carbonate to restore CO 2 , and again ignite very gently. Magnesium chloride loses its Cl on ignition. The residue is moistened with concentrated hydrochloric acid, and evaporated to dryness on the steam bath, and, after drying at 120 C. for two hours to render silica insoluble, is taken up with distilled water, more hydrochloric acid being used if any dark-coloured residue (ferric oxide) remains undissolved. The solution is filtered, and the residue, after thorough washing, is ignited and weighed as silica, SiO 2 . It should be colourless, and on boiling with dilute hydrochloric acid the solution should give no precipitate with ammonia and ammonium oxalate (calcium sulphate).* The filtrate is boiled with a little nitric acid to oxidise ferrous salts, and ammonia is added till the solution smells pretty strongly, and the boiling continued a short time. If a precipitate (ferric hydrate, alumina, phosphates) is formed, * If any precipitate of calcium oxalate is obtained, it must be estimated and added to that found in the ordinary course of analysis. TECHNICAL WATER ANALYSIS. 45 it is filtered off and washed and ignited and weighed as ferric oxide, Fe 2 O 3 , from which Fe may be calculated as 112 : 160, or by multiplication by 0*7. If the precipitate is considerable it is well to wash by decantation, re-diss>olve in a little HC1, and precipitate again with ammonia, as otherwise lime and magnesia are apt to be carried down with the iron. It will contain any alumina or phosphates present, for the separation of which chemical text-books must be consulted, but the whole precipitate is usually so small in quantity as to render this negligible. (Cp. also p. 365.) To the filtrate (or clear solution) ammonium chloride is added if little hydrochloric acid has been used, and then excess of ammonium carbonate and a little ammonium oxalate, and the solution is boiled, and, if possible, set aside for twelve hours in a warm place. It is then filtered, and the preci- pitate, after washing, is strongly ignited, first over the Bunsen burner, and then before the blow-pipe till the weight remains constant. It is weighed as lime, CaO, from which Ca may be calculated as 40 : 56. If the quantity is large, ^ it is better to ignite very gently and weigh as calcium carbonate, CaCO 3 . It should then be moistened with ammonium carbonate, dried in the water oven, and again very gently ignited, when no gain of weight will take place unless the first ignition has been too strong. The atomic weights are : Ca = 40, CaO = 56, CaCO 3 . = 100. The filtrate contains salts of Mg and the alkalies from the water, and ammoniacal salts which have been added. The Mg may be very accurately determined by the volumetric method with lime water, given on p. 36. To determine directly, the filtrate is evaporated to small bulk, strongly acidified with nitric acid and evaporated to dryness in a porcelain basin under an inverted funnel to destroy excess of ammonium chloride, or the evapora- tion may be carried to dryness and the ammonium chloride expelled by simple heating till no further fumes are given off.. The funnel and basin are rinsed with distilled water into a beaker, excess of ammonia and sodium phosphate and a little ammonium chloride added, and the solution covered up and allowed to remain undisturbed for 12 to 24 hours. The crystalline precipitate is collected on a filter and washed with a mixture of I part strong ammonia to 5 parts water, dried and detached as completely as possible from the filter, and transferred to a crucible, which is 46 LEATHER INDUSTRIES LABORATORY BOOK. heated at first gently with the cover on, and then .ignited. The filter is burnt separately, rolled in platinum wire, and the ash .{free from carbon) is added to the precipitate and strongly ignited for 15 minutes. The residue is magnesium pyrophos- phate, Mg 2 P 2 O r = 222, from which Mg 2 = 48 is calculated. The alkalies are most conveniently determined in the filtrate from the sulphuric acid determination (see below). Determination of Sulphuric Acid. At least 250 c.c. of the original water is rendered distinctly acid with hydrochloric acid, .and is concentrated by evaporation to about 100 c.c. ; and to the boiling liquid excess of barium chloride solution is added {5 to 10 c.c. of a 10 per cent, solution). The mixture is allowed to stand for some hours in a warm place, and the clear liquid is poured through a quantitative filter, and the precipitate washed several times with hot water by decantation, the wash-waters being poured through the filter, into which the precipitate is finally transferred and well washed, dried, ignited and weighed. It consists of barium sulphate, BaSO 4 = 233, from which the SO 4 = 96 is calculated. In place of gravimetric estimation with BaCl 2 , sulphuric acid, free or as alkaline sulphates, may be conveniently and accurately estimated by precipitation as benzidine sulphate according to the elegant method of Wolf Miiller * or its modification by Raschig.f For the estimation of sulphates in water and for many other technical purposes, Wolf Muller's method is perfectly satisfactory, and will be here described, but in presence of ferric salts and some other substances fails to give satisfactory results ; and to overcome these special difficulties Raschig's method has been devised, for details of which the original papers must be consulted. Wolf Muller's method depends on the fact that while sulphuric acid, either free or as sulphate, is very completely precipitated by benzidine chloride, the base benzidine is so weak that any strong acid in combination with it can be completely estimated by titration with NaOH and phenolphthalein, as if free. If, then, an excess of an acid solution of benzidine chloride, of which the acidity as compared to N/io NaOH is exactly known, is run into a solution of an alkaline sulphate carefully neutralised to phenolphthalein, the sulphuric acid combines with the benzidine * Chem. Berichte, 1902, p. 1587. t Zeit. angew. Ch., 1903, Hefte 26, 34; Collegium, 1904, p. 170 et seq. TECHNICAL WATER ANALYSIS. 47 and is precipitated and filtered off, while an equivalent portion of the hydrochloric acid is neutralised by the base, and on titrating back with NaOH, this deficiency is determined and calculated as SO 4 . For the practical execution of the method. I grm. of com- mercial benzidine base is rubbed up with a little water and dissolved as completely as possible by warming in 100 c.c. of N/i HC1, made up to I liter and filtered. Such a solution will be approximately N/io as regards acid, and about 30 c.c. will precipitate 0*1 grm. H 2 SO 4 , leaving sufficient excess of benzidine in solution. It is desirable that the requisite quantity of the benzidine solution should be measured, and the sulphate solution added to it, so that the benzidine is always in excess, and after well mixing, allowed to stand at least five minutes, while for very dilute solutions, or those containing traces of organic matter, some hours are not too much, and where much organic matter is present, precipitation is frequently imperfect. In the estimation of the sulphates in waters, it will usually be sufficient to add 200 c.c. of the water to 50 c.c. of the solution, and after filtration to twice titrate 100 c.c. corresponding to 20 c.c. of benzidine solution and 80 c.c. of the water. If 20 c.c. of the benzidine solution requires 19*3 c.c. of N/io soda, and 100 c.c. of the filtrate 6*7 c.c., we have (19- 3- 6-7) X 0-0048 X IQQ =O . O726 grm . SOi per oO 100 c.c. or 72*6 parts per 100,000. The water must have been previously boiled to expel CO 2 and remove temporary hardness, or an amount corresponding to the temporary hardness of the water must be added to the number of c.c. of soda required for neutralising the filtrate, and in any case it is well to raise the liquid to a boil to expel CO 2 before completing the titration, if the greatest accuracy is required. In applying the method to sulphates of bases which do not redden phenolphthalein, it is necessary to convert them into alkaline sulphates by treatment with sodium carbonate, and subsequent neutralisation and boiling to expel CO 2 , or to adopt Dr. Raschig's modification (loc. Y.) in which the benzidine sulphate is filtered off, washed and titrated. In all cases of doubt the absence of sulphates in the filtrate should be ascer- attained by testing with barium chloride, which at most should show a faint turbidity after a considerable time. 48 LEATHER INDUSTRIES LABORATORY BOOK. !C H NH ) p 6 TT 4 -NTTT 2 [. It has a molecular weight of 184, and, being a di-amido compound, combines with I molecule of sulphuric or 2 of hydrochloric acid. Determination of Alkalies. Boil the filtrate from barium sulphate, p. 46, with excess of baryta water or pure milk of lime to remove magnesia, iron and phosphoric acid. Filter, concentrate filtrate, add ammonia, ammonium carbonate and a few drops of ammonium oxalate, evaporate to dryness, ignite very gently, and wash residue with distilled water, filter, and evaporate filtrate to dryness in a weighed basin with a few drops of hydrochloric acid. It is safer to repeat the treatment with ammonium carbonate and oxalate, as traces of lime or baryta are very liable to escape precipitation. The residue is alkaline chlorides. If potassium is present in appreciable quantities, which is very rarely the case, it may be determined as chloroplatinate, or the chlorine may be determined (p. 39 ) and calculated as sodium chloride, which is deducted from the actual weight of mixed chlorides. The remainder multiplied by -- -- or 0*463 39*i - 23 represents the potassium chloride. Statement of Results. As chemical analysis furnishes no direct information as to the state of combination in which the acids and bases found existed in the original water, it is cus- tomary in the first instance to state them separately, the acids usually being calculated as radicals (ions) which with the base form the salt, thus Ca + SO 4 , Ca -f CO 3 , etc. Afterwards from these figures the salts existing in the water may be arranged with acids and bases in combining proportion and usually on the assumption that CO 3 first combines with Ca and then with Mg, then SO* in the same order. The CO 3 is calculated from the temporary hardness, each part of hardness corresponding to 0*6 part of CO 3 . Probably this will be best illustrated by the calculation of an imaginary somewhat hard water, of temporary hardness 16*0 and permanent 20*8, and total inorganic solids (p. 44), 46*5 parts per 100,000, and of the composition shown in table on next page. Of course it often happens in practical work that the figures do not fit quite so exactly as in the above, but at least the total inorganic solids should very approximately equal the sum of TECHNICAL WATER ANALYSIS. 49 the individual determinations stated as above. The method fur- nishes a Useful check on analytical work, but is otherwise of little practical value, as the original figures together with the hardness Per 100,000. CaCO 3 . CaSO 4 . MgSO 4 . NaCl. SiO 2 . I Ca .. .. 8-9 6-4 2'5 ' '** "t .. .. Mg 3'5 3*5 .. Na .. .. I'2 ..,..' I'2 .. C0 3 9'6 9*6 .. .. S0 4 ... 2CTO 6*0 14*0 .. .. Cl .. .. 1-9 .-. I- 9 .. SiO a i'4 ,. i*4 Total .. 46-5 16-0 8-5 i7'5 3'I I- 4 determinations furnish all the information required by the chemist, and we have no real knowledge as to the combinations actually existing in water, and according to modern theory the statement as uncombined ions probably more nearly represents the actual facts than any arrangement in definite salts. In judging of the suitability of a water for tanning, it maybe generally stated that, in absence of serious organic impurity, the softer it is the better. The amount of temporary hardness is of the first importance, as, when employed for washing goods after unhairing, it fixes both the lime in the hide and that present as temporary hardness in the form of carbonate, rough- ening the grain and causing stains and discoloration in the liquors. The lime thus fixed is not removed by puering or in the ordinary course of bating, nor even by drenching, unless the drench is more than usually acid. Temporary-hard waters are also very objectionable for dissolving extracts and extracting tanning materials, as, in absence of stronger acids, the lime carbonate reacts with the tannins, of which it neutralises and renders useless many times its own weight, forming lime tannates which oxidise with the formation of dark-coloured compounds in the liquors. Permanent hardness is much less injurious in this respect, as the tannins cannot displace the stronger acids directly, but will react in presence of alkaline salts of weak acids, which always exist in tanning liquors, the tannin combining with the lime of E 50 LEATHER INDUSTRIES LABORATORY BOOK. the sulphate or chloride, while the sulphuric or hydrochloric acid combines with the alkali, displacing the weak organic acid ; but of course this can only take place in nearly neutral liquors, and may be prevented by the addition of a calculated quantity of acetic, formic or lactic acids sufficient to decompose the lime salts formed. It is possible also that permanent hardness diminishes the solvent power of the water for the more difficultly soluble tannins. Considerable experimental work has been done on this subject by Nihoul,* and by Paessler and Appelius, f who have arrived at almost opposite conclusions as to its importance ; though both agree that temporary hardness causes some loss, and decided darkening in colour, both of the liquor and the leather tanned with it. The explanation of these differences in result is probably to be sought in the different amounts of acidity, especially including free carbonic acid present, and in the tem- peratures of extraction ; the injurious effect being most serious when heat was employed, and being especially marked with sumach. In the manufacture of the finer leathers, hardness is injurious in several ways beside those already mentioned. In leather- dying, temporary hardness causes precipitation of the basic coal- tar colours, which however can be prevented by the addition of small quantities of acetic acid. Neither temporary nor perma- nent hardness interfere with the ordinary use of acid colours, as acid is usually added in considerable excess (see p. 376), and with the dye-woods they are rather advantageous than otherwise. In fat-liquoring of chrome and other tannages the influence of hard water of either sort is very marked, causing the curdling and separation of the fat-liquors, with the formation of insoluble lime and magnesia soaps ; and in scouring with soaps it is equally objectionable. Theoretically each part of hardness per 100,000 destroys about 12 parts of soap (sodium stearate or oleate), but really the loss is greater, though somewhat variable. For this reason, where destruction of soap is the important factor, * L'influence de la nature de 1'eau sur 1'extraction des matieres tannantes. Collegium, 1902, i. 80. t Einfluss der Harte des Wassers auf die Ausnutzung 627-8 .. "17. Guy's Hosp. 6-0 2'0 i3'9 M 73'8 i'3 *i8. Cawnpore 9'4 nil i'3 10*6 16-7 3'9 *I9. Limerick 3*5 13-0 2'0 nil 22-7 0-7 *20. Stourbridge 18-0 24-0 I3-7 M 49-6 8'2 *2I. 20'0 32-5 16-2 J 97'3 22-5 22. Leicester .. 26-5 10-8 2-8 * '. 53'o 20 '9 TECHNICAL WATER ANALYSIS. DETAILED ANALYSIS OF SOME OF ABOVE WATERS. (Marked *.) 53 Al. | Ca. Mg. Fe. Na. S0 4 . C0 3 . Cl. Si0 2 . C0 2 Free. 0'8 12. Chesterfield.. .. -. 3-i I-I 6-1 4'5 1*4 15. London district .. 6-3 4-3 trace .. 6-0 15-0 1*7 .. 16. ,, .. 17- , 13-760-4 .. .. i 1-5 i-o PI 4I-3 23*7 70-3 15-8 20-5 i3 P 4 63-9 i3'9 330-5 I'l " 1 8. Cawnpore .. .. ! i'8 0-3 O 2 4-6 4*1 u-6 i'3 i -o 19. Limerick .. | 3'I 2-1 2'2 9-0 2'I 2'O .. 20. Stourbridge . . .. '10-9 i'o O'2 1-9 7'6 II'4 6-2 1-3 2-6 21. .. .. .. JI6-5 i'4 I'l 6-0 15-2 12-3 i3'7 0-9 4*4 SECTION IV. DEPILITANTS. Solubility of Limes. The strength of lime-water given by different limes is found to vary slightly. The explanation of these results is not very easy, but pro- bably depends on the presence of some other soluble base in addition to the lime. It harmonises with the old belief of tanners that pure chalk lime is milder in its action on skin than that made from less pure limestones. The solubility of any given lime is easily determined by adding it in excess to water in a stoppered flask, and shaking frequently till a solution of constant strength is obtained, of which a measured quantity is titrated with N/io hydrochloric acid and phenolphthalein. Saturated lime water may be conveniently used as an alkaline standard solution for many purposes, and if kept on excess of lime, is always caustic, and varies little in strength at ordinary labora- tory temperatures. It is better to allow it to settle and to remove the clear solution with a siphon or pipette than to filter, which often reduces the strength. The solution is nearly ^ normal, but for accurate work its strength should be occasionally deter- mined with N/io acid. The solubility of lime diminishes with increasing temperature. 54 v LEATHER INDUSTRIES LABORATORY BOOK. The following table gives the strength of a pure saturated lime solution for various temperatures as [determined by Mr. A. Guthrie in the author's laboratory.* IOO c.c. of saturated lime water at 5 C. contain o* 1350 grm. of CaO. IO 0-1342 , 15 0-1320 20 0-1293 25 0-1254 30 0-1219 . 35 0-1161 , 40 0-1119 i 5o 0-0981 60 0*0879 70 0-0781 80 0-0740 90 0-0696 IOO 0-0597 Determination of "Available" Lime. The practical value of lime for the tanner is easily determinined by drawing a sample by breaking off small pieces from a number of lumps of the bulk, coarsely pulverising them in a mortar, and then grinding a portion as fine as possible, and transferring it at once to a stop- pered bottle for weighing. A portion of this not exceeding I grm. is shaken into a stoppered liter flask, which is filled up with hot and well-boiled distilled water, and allowed to stand for some hours with occasional shaking. When cold it is filled to the mark with recently boiled distilled water, and well shaken again and allowed to settle, or filtered with as little exposure to the air as possible, and 25 or 50 c.c. withdrawn with a pipette and titrated with N/IO hydrochloric or sulphuric acid and phe- nolphthalein. Each cubic centimeter of N/io acid equals O'OO28 grm. CaO. The presence of a small quantity of insoluble sediment in the liquid titrated does not affect the phenol- phthalein. Another method, depending on the solubility of lime in cane sugar solutions, is described by Messrs. Stone and Scheuch.f One grm. of finely powdered quicklime is shaken for twenty minutes with 150 c.c. of a 10 per cent, solution of pure lump sugar, and 10 c.c. of the filtered solution titrated as before with phenolphthalein and N/io hydrochloric acid. Calcium car- * Journ. Soc. Ch. Ind., 1901, p. 224. t Journ. Amer. Chem. Soc., xvi. p. 721 ; Chem. News, 1894, p. 278. DEPILITANTS. 55 bonate, magnesia, alumina and ferric oxide are not dissolved by sugar solution, and do not interfere. This method gives accurate results with careful manipulation, but in the author's experience, it is safer to allow the solution to stand some hours, and it is not on the whole to be preferred to that given above. A solution of lime in cane sugar solution may be used as a standard solution, with phenolphthalein as an indicator where a stronger solution than lime-water is desirable, and the addition of sugar has also been used to shorten the time of liming. A 25 per cent, solution of cane-sugar will dissolve lime to nearly N/i strength. If desired, a complete analysis of a lime can be made by gravimetric methods similar to those employed in determining the mineral constituents of water (p. 43), but this is rarely neces- sary for the purposes of the tanner. The total bases existing as oxides, carbonates and easily decomposable silicates, are readily determined by treating I grm. of the finely powdered sample with 50 c.c. of N/i HC1, and titrating the excess of acid. Sodium Sulphide, crystallised sulphide of sodium, Na 2 S-9 Aq, contains 67*5 per cent, of water, and 32*5 per cent. Na 2 S. A fused salt is also sold, containing a much larger percentage of Na 2 S. It will generally be sufficient to determine the sodium and sulphur, any sodium in excess of that required to form the sulphide being calculated as sodic .carbonate, which is gradually formed by exposure to the air. IO grm. or conveniently 12 grm. of the sample is dissolved in water and made up to I liter. Sodium is determined by titrating with N/io HC1 or H 2 SO 4 in presence of methyl-orange, which is not sensitive to SH 2 . Each c.c. of N/io acid is equal to 2 -3 mgr. of Na or 3*9 mgr. of Na 2 S, or 12 mgr. of the crystal- lised salt, so that if 12 grm. are weighed out, the solution of the pure salt would be decinormal. Determination of Sulphur as SulpJiide. 14*35 g rm - f pure crystallised zinc sulphate (ZnSO 4 . /Aq) is dissolved in water, and ammonia is added till the precipitate which is at first formed is dissolved, and the whole is made up to a liter. An alternative method is to dissolve 3*25 grm. of pure zinc in dilute sulphuric or hydrochloric acid with the aid of a piece of platinum foil, add excess of ammonia and make up to I liter. The solution is decinormal, and each c.c. equals I -6 mgr. sulphur, or 12 mgr. of 56 LEATHER INDUSTRIES LABORATORY BOOK. the crystallised sulphide of sodium. If 12 grm. of the sample is made up to I liter, and 100 c.c. titrated, each c.c. of either N/io solution will correspond to I per cent, of pure crystallised salt. The standard solution is added with constant stirring to, say 25 c.c. of the sodium sulphide solution in a beaker, and after each addition a drop is taken from the beaker and placed on a piece of filter paper on a white plate, side by side with one of a solution of lead acetate, but not actually touching it, so that the two run together by capillarity. (A somewhat more delicate lead indicator may be made by dissolving lead acetate in solution of sodium tartrate, or tartaric acid made strongly alkaline with sodium hydrate and filtering.) So long as any unprecipitated sulphide is present, a black or brown ring is formed where the two liquids come in contact. Care must be taken that the lead solution does not reach the precipitated zinc sulphide, which is always blackened by it. It is best first to make a rough determination, adding, say, 2 c.c. of standard solution at a time ; and having in this way determined the approximate quantity required, nearly the full amount may be added at once. A solu- tion of sodium sulphide may similarly be employed for the determination of zinc. Instead of using a lead salt as indicator, a solution of sodium nitro-prusside may be employed, drops of which are simply spotted on a white tile and give a strong purple reaction with the least trace of alkaline sulphide. The zinc process may be employed for the determination of sulphur in other alkaline sulphides, but where polysulphides are present a yellow zinc sulphide is formed, probably a polysulphide, and the lead indicator gives an orange instead of a black ring. As a consequence the whole of the sulphur is not determined, but only that corresponding to the normal sulphide, and which would be liberated as SH 2 on treatment with an acid. Since it is pro- bable that only this portion of the sulphur is effective in un- hairing, this may probably be reckoned as an advantage rather than the reverse. Sulphides may also be determined by titration with N/io iodine,* but as many other sulphur compounds also reduce iodine, the method becomes complicated and is not to be recom- * Jean, Ann. de Chim. Anal., ii. (18), p. 341; Analyst, 1897, p. 306; Journ. Soc. Ch. Ind., 1897, p. 1041. DEPILITANTS. 57 mended for the analysis of depilitants, though quite suitable for the determination of small quantities of hydrogen sulphide. Realgar, Rusma, Red Arsenic, an impure form of As 2 S 2 made by fusing suitable proportions of arsenious acid and sulphur, and much used in combination with lime as a depilatory, especially in the manufacture of glove-kid and other light leathers. The realgar is usually mixed with excess of hot slaking lime for this purpose, and the decompositions which occur are not well under- stood and probably vary with the temperature. It is said that a mixture made with hot lime has a more powerful effect than a mixture made in the cold, or even with boiling water. Probably in all cases calcium sulphydrate is the active depilitant.* Orpiment or Yellow Arsenic Sidphide, As 2 S 3 , is generally said to be unsuitable for depilation, but in the author's experiments it proved more efficient, weight for weight, than realgar, and no difference could be detected in the character of the result. The yellow arsenic is now much cheaper than the red, as it is a by- product in the purification of arsenical sulphuric acid. Both are, of course, very poisonous, and must be handled with great caution, and lime liquors containing them must not be run into rivers ; and they do not appear to have any real advantage over calcium sulphydrate, or the mixture of equivalent quantities of sodium sulphide and calcium chloride, suggested by the author for use where the powerful alkaline swelling action of sodium sulphide used alone as an addition to limes is objectionable. Analysis of Arsenic Stilphides. The sample must be very finely pulverised in an agate mortar. Non-volatile impurities are determined by igniting a small quantity, not exceeding i grm., in a porcelain crucible over a good Bunsen burner and weighing the residue. As the fumes are very poisonous, the operation must be performed in a good draught-chamber. Both sulphides of arsenic are soluble in the cold in caustic soda or potash solutions. I grm. of the finely powdered sample is digested for some hours with 50 c.c. of 10 per cent, caustic soda solution with frequent shaking, and made up to IOO c.c. with * Stiasny (Gerber, 1906, pp. 272, 288), shows that only a small quantity of the arsenic goes into solution, and that the soluble arsenic compounds have no depilatory effect. Eitner (Gerber, 1907, p. 48), found that they also had no effect as antiseptics,, and that it was impossible to distinguish skins treated with calcium sulphydrate only rom those on which arsenic was used. Stiasny (loc. cit.) shows that the presence of OH as well as SH ions in solution is essential to satisfactory depilation. 58 LEATHER INDUSTRIES LABORATORY BOOK. water, and filtered. The filter is well washed, the wash waters being rejected and not allowed to mix with the filtrate ; the filter is ignited, and the impurities are weighed. The sulphide-sulphur may be determined in the filtrate by the zinc method (p. 55), but in a commercial realgar only amounted to about 12 per cent, while the total sulphur theoretically present in pure As 2 S 2 is 29*9 per cent. It is probable, however, that the sulphur determined with zinc is proportional to the depilatory power of the sample, and may be safely used for comparison of relative value and for determining residual sulphide in lime liquors. Both sulphides are liable to contain excess of either As 2 O 3 or of sulphur, according to their mode of manufacture, and for the purposes of the tanner As 2 O 3 is of no value and free S of very little. Hence the determination of total sulphur or total arsenic is not of much practical importance, except as showing that they are present in the proper proportions. The estimation is best made by oxidising sulphur to sulphuric acid and arsenic to arsenic acid by digestion with nitric acid. Estimation of Sulphur. O'io-*2 grm. of the sample is di- gested with a mixture of 3 volumes of concentrated nitric acid and I volume concentrated hydrochloric acid, and the solution evaporated repeatedly nearly to dryness with addition of con- centrated HCL The solution is diluted and filtered if necessary, and precipitated with excess of BaCl 2 solution and the precipi- tate washed, dried, ignited and weighed with the usual pre- cautions (p. 12). The weight of the precipitate multiplied by or 0*137 1 gives that of sulphur present in the quantity taken. Estimation of Arsenic. cr 1-0-2 grm. of the sample is boiled with concentrated nitric acid till dissolved, diluted with water, rendered strongly alkaline with ammonia, filtered if necessary, and precipitated with magnesia solution (n grm. cryst. magne- sium chloride, 14 grm. ammonium chloride, 70 c.c. strong ammonia and 130 cc. water) and its volume of strong alcohol. The solution is allowed to stand at least 12 (and preferably 48) hours ; the precipitate is filtered off and washed with a mixture of 2 parts strong ammonia, 2 parts water and I part alcohol. After drying the filter in an air or water-oven, trre precipitate is DEPILITANTS. 59 transferred as completely as possible from the filter on to glazed paper ; the filter is replaced in the funnel and washed with a little hot dilute nitric acid, which is evaporated to dryness in a weighed porcelain crucible ; the remainder of the precipitate is added, and the covered crucible is slowly heated, finally to a dull red. The cover is now removed and the open crucible strongly ignited over a powerful flame, placed while still hot in an exsiccator over sulphuric acid, and weighed when cool. The weight of Mg 2 As 2 O 7 X - 1 -- , 01-0-483 As. A shorter but less exact method is, after oxidation with fuming nitric acid, simply to evaporate the solution to dryness in a weighed porcelain crucible, drive off the sulphuric acid by cautious heating, and finally raise the crucible to dull redness, cool in an exsiccator over P 2 O 5 and weigh as As 2 O 5 . Any non- volatile impurities will be included in the result. After weighing, the residue should be re-dissolved and tested with BaCl 2 for sulphuric acid too high a temperature of ignition will cause loss of arsenic. SECTION V. THE ESTIMATION OF AMMONIA AND THE DETERMINATION OF NITROGEN. THE estimation of ammonia is one of the most important of those required by the leather trades chemist, not only because its presence in lime liquors has a very considerable effect on the unhairing process, and indicates to some extent the destruction of hide-substance, but because in the Kjeldahl process all the nitrogen present in organic matters is converted into sulphate of ammonia by digestion with concentrated sulphuric acid, and consequently the estimation of the ammonia so formed is a measure of the quantity of hide-substance which has been treated, 60 LEATHER INDUSTRIES LABORATORY BOOK. and thus forms a means for its estimation both in solution and in the form of leather. Some information is given on pp. 26 and 7 1 with regard to the ordinary testing of ammonia solutions and ammonium salts, but where these occur together with organic matter further pre- cautions are required. In the case of lime liquors, all the ammonia is free, together with an excess of caustic lime, which, if the solution is heated for any length of time, acts upon the nitrogenous matter present, decomposing it and producing further quantities of ammonia. It is therefore desirable to avoid heating as much as possible, and Schloessing's method has been commonly used. It has, how- ever, the disadvantage of being very slow, and in very weak ammonia solutions it is certain that the whole of the ammonia is not really estimated. Experiments have been made in the Author's laboratory * to devise a more rapid method of ammonia estimation in the cold by placing the lime liquor in an absorption apparatus connected with a second containing standard acid, and gently sucking a current of (ammonia-free) air through the apparatus so as to carry the ammonia into the acid. Good comparative results were obtained, but it is difficult without the use of heat, absolutely to carry over the whole of the ammonia. With a suitable apparatus and a temperature of about 65 C, however, the estimation is very rapid and exact, and no appreciable evolution of ammonia from the nitrogenous compounds present in a lime liquor occurs, so that where many estimations are required the process is much to be recommended, as once the apparatus is arranged, very little attention is required. The arrangement employed is shown in Fig. 6, the air, previously freed from carbonic acid and ammonia, is passed into a wide test- tube partially filled with broken glass and containing 50 c.c. of the lime liquor, through a tube so arranged as continuously to raise the liquor along with the air and allow it to trickle down over the broken glass, while the air is drawn through an absorption tube containing 30-50 c.c. of N/5O sulphuric acid, or of a stronger standard acid if the quantity of ammonia is excessive. The * Procter and McCandlish, "The Estimation of Ammonia in used Lime Liquors," Journ. Soc. Ch. Ind., 1906, p. 254; Colleg., 1906, p; 270. THE ESTIMATION OF AMMONIA. 61 test tube is surrounded with water at about 90 C., sufficient to raise the interior of the tube to about 65 C. The detail of the bubbling tube is shown on a larger scale on the left of the figure. A few drops of turpentine may be added to the lime liquor to prevent frothing. With a steady current of air, the operation is complete in thirty minutes. Dr. Nihoul has devised a somewhat similar but continuous apparatus, in which the lime liquor is allowed to trickle in a slow continuous stream over the broken glass, through which the air is aspirated. In this way it is To Vacuum Pump Air FIG. 6. possible to work with much larger quantities of lime liquor, and hence to titrate larger quantities of ammonia. If suitable indicators could be found, a rapid titration of ammonia in lime liquors might be made by testing their total alkalinity, then boiling the solution for a short time to drive off the free ammonia, and titrating again. Unfortunately the presence of amido-acids and other weakly acid bodies renders the end reaction for the total alkalinity an uncertain one with methyl orange, and phenolphthalein is not sufficiently sensitive to the action of ammonia to be employed. 62 LEATHER INDUSTRIES LABORATORY BOOK. Determination of Ammonia by Schloessings Method. 25 c.c. of the liquor which has been quickly filtered into a flask from a funnel covered with a clock-glass to avoid loss of ammonia and carbonation, is placed in the lower basin of a Schloessing's ap- paratus (Fig. 7), and 25 c.c. decinormal sulphuric acid in the upper one, and the apparatus is closed for at least 48 (preferably 72) hours. The ammonia passes into the acid, which is titrated back with N/io sodium hydrate or barium hydrate solution in presence of methyl orange, litmus or lacmoid FIG. 7. (Kathreiner), or with lime-water. A modified apparatus may be constructed by supporting a small basin or porcelain crucible in a flat-bottomed glass crystallising dish. The lime liquor is placed in the crucible, while the acid, diluted if necessary with water, surrounds it in the bottom of the crystallising dish, and forms a water-seal for a second dish, or beaker, which is inverted over the crucible. If this be placed on the top of an air bath or other warm place, the reaction will be complete in 24 hours. It is very important that both the vessels should expose as much surface as possible, and that while there should .be ample space THE ESTIMATION OF AMMONIA. 63 for the ammoniacal vapour to diffuse into the acid, the capacity of the containing vessel should not be uselessly great. The closure must be accurately airtight to prevent escape of am- monia, and if ground joints are used, they must be accurate,, and lubricated with vaseline. Corks, or rubber stoppers, often give off traces of ammonia, and are therefore inadmissible. A good form for the outer vessel is that of a shallow exsiccator or glass jar, with a cover ground to fit. The ammoniacal liquid is placed in the bottom of the jar, and the acid in a shallow basin, raised just above the surface of the liquid by a glass tripod. By Distillation. Much more rapid determinations may be made by using a conical flask or a retort of at least 500 c.c. capacity, connected with an absorption flask or U tube contain- ing standard acid. The apparatus used in the determination of ammonia in Kjeldahl's process (p. 67) is very suitable, but it is well to introduce a wide tube filled with beads or broken glass in the limb above the boiling flask, as liquids containing organic matter are apt to froth excessively. The liquid condensing in and above this constantly washes the frothing substance back into the boiling flask. The frothing may be greatly diminished by the addition of a few drops of oil of turpentine. A suitable quantity of standard acid is placed in the absorption apparatus, and 100 c.c. of the liquor in the retort, which is briskly boiled for 1 5 minutes and the absorption apparatus rinsed out, and the liquid titrated back, usually with caustic alkali in presence of methyl orange. If the boiling of alkaline liquors containing nitrogenous organic matter is long continued, traces of ammonia are evolved by the action of the alkali on the organic matter present. This may be prevented or at least much lessened, if the residue is not required for the determination of caustic lime, by adding two or three c.c, of a 10 per cent, solution of MgSO 4 .7OH 2 , which converts the alkali or alkaline earth into sulphate, setting free the MgO ; but is rarely of importance for technical purposes, if the boiling is not long continued. In liquids containing ammoniacal salts, the am- monia is best liberated by the addition of calcined magnesia, which has little or no action on organic matter, and does not affect methyl orange if carried over as spray. In comparing cold methods with distillation methods of estimating ammonia, it must be remembered that amines, if present, will be more com- 64 LEATHER INDUSTRIES LABORATORY BOOK. pletely estimated in the latter case on account of their lesser volatility. In estimating ammonia by titration, indicators more sensitive to weak acids than weak alkalies, such as phenolphthalein, are unsuitable. Congo red has been a good deal used by continental chemists, but its end-reaction is not sharp. Methyl orange is better by daylight and for the trained eye, but probably carminic acid (i per cent, aqueous solution) is to be preferred, not only on account of the sharpness of the colour-change, but because it is easily seen by artificial light. The Kjeldahl Method of Nitrogen Determination. The principles of this method have already been briefly alluded to. Few processes have been subject to more varied suggestion and real or supposed improvement, but, except as regards estimation of nitrogen in the form of nitric acid and nitro-compounds, which can scarcely concern the leather chemist, it is doubtful if any real advance has been made on the process as first proposed by the inventor. The essential idea of the method is the digestion of the nitrogenous organic matter with an excess of concentrated sulphuric acid, which acts not only as a dehydrating but an oxi- dising agent, destroying all carbonaceous matter, and transform- ing all organic nitrogen into ammonium sulphate. With many proteids indeed, and with hide itself, accurate results may be obtained by this simple method if the digestion is carried on sufficiently long to destroy all organic matter, and produce a perfectly colourless solution. Where, however, matters contain much carbon in proportion to nitrogen, as is the case with leather, this digestion becomes very tedious, two or three hours being often required to complete the reaction, and many suggestions have been made to shorten the time required, and render com- plete oxidation more certain. This may be accomplished either by the addition of more powerful oxidisers, such as the potassium permanganate employed by Kjeldahl, of oxygen carriers, such as mercury and copper, or of salts, such as potassium sulphate, cal- culated to raise the temperature of the boiling sulphuric acid. It must be particularly emphasised that by all these methods, carefully carried to complete oxidation and decolorisation of the solution, accurate and concordant results are obtained, and that the question to be decided is rather one of rapidity and conveni- ence of manipulation than of exactness. There is at least a THE ESTIMATION OF AMMONIA. 65 theoretical possibility that very strong oxidising agents may cause loss by liberation of free nitrogen, and this danger becomes probably a real one where chlorides are present from which free chlorine may be evolved. The use of metallic salts, such as the sulphates of mercury or copper, besides somewhat complicating the process, causes precipitates in the solution when rendered alkaline with soda, and increases the very annoying tendency to " bumping," always present during the distillation of the am- monia ; and even the addition of a soluble salt like potassium sulphate is somewhat troublesome in the latter respect. For the analysis of leather, Dr. Nihoul* recommends the following method as rapid and accurate: 0*6-0-7 of grm. the leather, freed from fats and water-soluble matter, is boiled for half an hour with 10 c.c. of pure concentrated sulphuric acid. The flask is allowed to cool to a moderate temperature, and a dozen large crystals of potassium permanganate are added and shaken till dissolved. The flask is then again heated till the liquid is completely decolorised, and allowed to cool. The solu- tion is made up to 250 c'.c. (including washings if it is transferred to a separate distilling flask), 1 50 c.c. of 30 per cent, soda solu- tion is added, and a known quantity of zinc dust, and the dis- tillation is continued slowly so that in f hour about 200 c.c. have passed over into the receiver, which contains 25 c.c. of N/5 sul- phuric acid, and the solution is titrated back in presence of carminic acid or methyl orange. If this method is applied to actual hide-substance, the quan- tity of dry matter used must not exceed 0*4 grm. In place of the permanganate recommended by Nihoul, l grm. of potassium persulphate, or more if required for decolorisation, may be used, and has the advantage of causing no change in the colour of the solution, and no precipitate on the addition of caustic soda.f If the use of oxidising agents is considered objectionable, good results are obtained by the use of 1 5 c.c. of concentrated sulphuric acid to which, when the substance is dissolved, 10 grm. of dried and powdered potassium sulphate is added ; but con- * Bourse aux Cuirs de Liege, Sept. 1901, p. 14 ; Abst. Journ. Soc. Chem., 1901, p. 1249. f Compare Law, Journ. Soc. Ch. Ind., 1902, p. 847, and Dakin, ibid. p. 848. Potassium persulphate, however, usually contains traces of ammonia, which must be determined and allowed for. F 66 LEATHER INDUSTRIES LABORATORY BOOK. siderably longer boiling is required before the liquid becomes colourless. The time is somewhat reduced by the addition of 2 grm. of dry copper sulphate * or of about I grm. of mercury ; but in the last case, sodium sulphide must be added with the caustic soda on distillation to decompose the ammonia-mercury compound formed. Dr. O. Bottcherf recommends in the analysis of artificial manures, the use of a mixture of 200 grm. of phosphoric anhy- dride with I liter of concentrated sulphuric acid to which, for 2O c.c. of acid, I grm. of mercury and 10 to 15 grm. of potassium sulphate may be added. This quantity of mercury will require about 25 c.c. of a 4 per cent, solution of potassium polysulphide (liver of sulphur) for precipitation ; but Dr. Bottcher states that if ij grm. of granulated zinc (free from arsenic) is added to the distilling flask, no sulphide is needed, and the boiling proceeds without bumping. If zinc is used for this purpose, the acid, after dilution with water, should be nearly neutralised with the soda before adding the zinc. For this purpose, it is convenient to ascertain by a rough titration what quantity of the soda solution is needed to neutralise the volume of acid employed. As regards the apparatus employed, long-necked spherical Jena glass flasks of 100-200 c.c. capacity are most suitable for the digestion with sulphuric acid, which is done over a small flame, under a good draught, the flask being held in a clamp or on a suitable support, with its neck inclined, and closed with a glass bulb with a short sealed stem, or with a small funnel, to avoid the possibility of loss by spirting. Jena glass flasks of about 400 c.c. capacity should be used for the distillation, and hard glass, which does not yield any alkali to ammoniacal vapours, must also be selected for tubes and receivers. Care should be taken that wherever rubber tube is used for jointing, the ends of the glass tubes butt closely together so as to expose the rubber as little as possible. In place of glass flasks, copper flasks are advantageously used for distillation, and entirely avoid trouble from bumping, but are difficult to obtain without a soldered seam, which is liable to attack by the caustic alkali. If glass flasks are used, it is advisable to carry a tube drawn to a narrow point through the cork down to the bottom of the flask, * Cp. Procter and Turnbull, Journ. Soc. Ch. Ind., igoo, p. 130. t Lunge's Chemisch-Technische Untersuchung-Methode, ii. 379. THE ESTIMATION OF AMMONIA. 67 and carry in a current of steam from ammonia-free water from a boiling flask to prevent bumping. In this case a condenser should also be employed. In place of steam, a current of ammonia-free air may be sucked through the apparatus. 1 5 to 20 minutes boiling is sufficient. Almost every chemist has his favourite form of distilling apparatus ; but the really important point, on which the accuracy of the analysis depends, is the complete separation of the spray FIG. 8. of the strongly alkaline soda-solution from the ammoniacal steam which passes into the receiver. The condensation of the steam after it has carried the ammonia into the absorbing standard acid seems unimportant, and the writer has always obtained good results with the simple apparatus shown in Fig. 8, in which the long rising tube of not less than 5 mm. bore serves at once to condense part of the water which runs back into the boiling flask, and to act as a spray-trap in which particles of spray striking the wet sides of the tube are caught by the condensed F 2 68 LEATHER INDUSTRIES LABORATORY BOOK. water and washed back into the boiling flask. The bottom of this tube may be simply cut off obliquely as shown, but a better form very easy to make has a side-hole blown near the end of the tube, through which the steam and spray enter, while the end is drawn out to a small opening, through which the condensed water drips back into the flask. The stoppage of spray would be rendered still more perfect by widening the lower part of the tube and filling it with beads or broken glass, but the apparatus as figured answers every practical purpose so long as steam only has to be dealt with. If, however, hydrogen is evolved from granulated zinc, the spray is of a much more finely divided type, and many traps have been devised to remove it ; but most of these are beyond the scope of the amateur glass-blower, and must be obtained of the chemical instrument maker. It is not necessary to cool the absorption-flask, if standard sulphuric acid is used to absorb the ammonia, as the latter is quite efficiently retained even by hot acid, if the point of the pipette dips very slightly below its surface. The object of the pipette, or bulbed tube, is to prevent the standard acid being drawn back into the boiling flask by any slight condensation of steam. Many chemists, however, prefer to use a condenser, the inner tube of which maybe either of glass or of pure tin, which is a better conductor of heat and is unaffected by ammonia. A small guard-tube in the cork of the absorption-flask, as shown in the figure, filled with broken glass and wet with a portion of the standard acid coloured with methyl orange, prevents danger of escape of ammonia un- absorbed. It is, of course, rinsed into the flask before titration. To sum up, the simplest form of process, that of digestion with sulphuric acid alone, is to be preferred for substances which it will oxidise within a reasonable time. It is sometimes neces- sary if the digestion is long continued, to make up for loss of sulphuric acid by a further addition. The use of potassium sulphate to raise the temperature is quite unobjectionable if the solution be freely diluted before distillation, and if it be necessary to further shorten the time by the use of an oxidising agent, potassium persulphate is ideal if it can be obtained free from ammonia ; and potassium permanganate, though less convenient, is also satisfactory, but in no case should an oxidising agent be added till solution is complete, and the digestion well advanced. Where many analyses are made, copper flasks are strongly to be THE ESTIMATION OF AMMONIA. 69 advised for distillation, in which case a condenser is unnecessary, but if glass flasks are employed the use of a current of steam and a condenser is advisable. As the calculation of Kjeldahl estimations is slightly com- plicated, it may be well to give an example. 0*7 grm. of leather has been taken, and 5 c.c. of N/5 caustic soda has been required to titrate back the 25 c.c. of N/5 acid placed in the absorption- flask. Each c.c. of N/5 acid corresponds to 0*0034 grm. NH 3 ; to O'OO28 grm. N; or to 0*01573 grm. of hide-substance reckoned to contain 17*8 per cent, of nitrogen.* Taking the latter figure, we have ?o_x^y_7!X_K> = 44 . 95 percent Qf hide . substance; and, of course, the nitrogen or the ammonia can be obtained in a similar way by using the corresponding factors. SECTION VI. THE MATERIALS USED IN CURING, BATING, ETC. Common Salt (NaCl = 58-5). This material is so cheap that it is not likely that it will be purposely adulterated, but may contain accidental impurities. Moisture may be determined by heating a few grams to 110 C. or over, in a crucible or basin which should be at first covered to avoid loss by decrepitation. Insoluble impurities are detected by dissolving say 10 grm. in water ; and may be collected on a tared filter, washed, dried at IIO, and weighed, and afterwards ignited to determine organic matter by loss. The residue, if brown or yellow, probably con- tains iron, which may cause stains on leather, and both residue and solution should be tested for ferric salts with thiocyanate, if neces- sary after solution in HC1 and oxidation with a little nitric acid. The iron may be estimated colorimetrically (p. 40) as in water. Salt not unfrequently contains magnesic and sometimes calcic * Von Schroeder for hide-substance, the following percentages of N. Ox, calf, horse, camel, pig, rhinoceros, 17-8 ; goat and deer, 17*4;', sheep and dog, 17*0 ; cat, 17-1. 70 LEATHER INDUSTRIES LABORATORY BOOK. chlorides as impurities, which tend to make it absorb moisture which for some purposes is objectionable. These salts also lose hydrochloric acid on ignition. To determine these impurities, a known quantity, say 10 grm., is dissolved in 100 c.c. of water; insoluble impurities if present are filtered off as above described, and the filter is washed, the solution boiled with a few drops of nitric acid, ammonium chloride and ammonia are added, and ferric oxide and alumina filtered off, and estimated if present (p. 44). Ammonium carbonate and a little ammonium oxalate are now added, and the mixture boiled, and, if any precipitate forms, allowed to stand 24 hours, the lime estimated as on p. 45 ; and the magnesia is then precipitated from the filtrate with sodium phosphate, and estimated as pyrophosphate (p. 45). A simpler method where magnesia only is to be estimated, is to precipitate the MgO with excess of lime-water as in the case of water analysis (p. 36). To 100 c.c. of solution containing 10 grm. of salt, made with water well boiled to expel CO 2 , and quite neutral to phenolphthalein, 100 c.c. of lime-water are added, the strength of which has been accurately determined with N/io HC1 and phenolphthalein. The mixture is allowed to stand overnight or heated in the water-bath for half an hour and allowed to cool and settle quite clear, when 50 c.c. are drawn off repeatedly and titrated. If 100 c.c. of the lime-water require 48 c.c. of N/io acid, and the 200 c.c. of mixture, as calculated by multiplying the average result of 50 c.c. by 4, require 18 c.c., the result is (48 18) x 10 X o 002 = 0*6 percent., the N/ioacid corresponding to 0*002 grm. of MgO or 0*00475 of MgCl 2 per c.c. The quantity of lime-water used must be such that the solution remains deeply coloured with the phenolphthalein after standing. The chlorine may of course be determined with argentic nitrate as on p. 39, but this is rarely necessary except for the estimation of small quantities of salt in wash-waters and the like. The specific gravity of salt solutions of different strength is given in the following Table. Salt is only soluble in water at 15 to the extent of 26*4 per cent, by weight, or 318 grm. per liter, and the solubility only increases slowly with temperature, rising to 28^9 per cent, at 109*7, the temperature of boiling salt solution, and falling to 24*5 per cent, at 14, its freezing point. THE MATERIALS USED IN CURING, BATING, ETC. 71 Between o and 25 it remains almost constant at about 26^4 pel- cent, arid 1000 parts of water dissolve 369 parts of salt. Baume degrees correspond very approximately to the percentage of salt. SPECIFIC GRAVITY OF SODIUM CHLORIDE SOLUTIONS AT 15 C. (59 FAH.). (Gerlach.) Sp. Gr. Percentage Salt. Sp. Gr. Percentage Salt. Sp. Gr. Percentage Salt. I-00725 I 1-07335 10 . I'I43I5 19 I*OI45O 2 1-08097 II I-I5I07 20 I '02174 3 1-08859 12 I'I593I 21 I-02899 4 1-09622 13 I-I6755 22 1-03624 . 5 I ' 10384 14 I'I758o 23 1-04366 6 I'll 146 15 1-18404 24 I-05I08 7 I '11938 16 I * 19228 25 I-0585I 8 I'I2730 J 7 1-20098 26 I'06593 9 I-I3523 18 1-20433 26-395 Ammonium Salts and Ammonia. So-called "ammonia'* or " liquid ammonia " is a solution of NH 3 (or NH 4 OH)in water. Real liquid ammonia is NH 3 gas liquefied by pressure, and is now largely used Jn ice machines, and for other purposes. Beside am- monia, the most important compounds are sulphate and chloride. The strength of ammonia solution is usually calculated from its specific gravity, which may be taken with the pycnometer or by means of a suitable hydrometer. A sealed glass bulb, loaded with mercury so as just to float in a solution of the required density, is a convenient and delicate means of making a solution of constant strength. Strong solutions lose strength rapidly and serious damage has been done to leather by not allowing for the greater strength of a fresh solution. The strongest com- mercial ammonia contains about 33 per cent, of NH 3 , corre- sponding to a density of 0*886, and may be conveniently diluted with twice its bulk of water, which will give a gravity of about 0-956. It should be kept in carboys as closely stoppered as possible. The following table, abridged from Lunge and Wiernik, gives the weight of NH 3 contained in 100 grm. and 100 c.c. of liquid ammonia at different specific gravities taken at i 5 C. LEATHER INDUSTRIES LABORATORY BOOK Specific Gravity at 15 C. Grms. per 100 grms. Grms. per 100 c.c. 1-000 0*00 o-oo 0-990 2-31 2-29 0*980 4-80 4-70 0-970 7-31 7-09 0*960 9-91 9-51 0-950 12-74 12-10 0-940 15-63 14-69 0-930 18-64 I7-34 0-920 21-75 ' 20-01 0-910 24-99 22-74 0-900 28-33 25-50 0-890 31-75 28-26 0-882 34-95 30-83 The strength of ammonia (previously diluted with water if strong) is also determined easily and exactly by titration with standard hydrochloric or sulphuric acid, with carminic acid, methyl orange or litmus (not phenolphthalein) as indicator. Each c.c. of N/i acid corresponds to 0-017 grm. of NH 3 . The ammonia of ammonium salts is similarly determined by liberating with caustic alkali or lime, and distilling into a known volume of standard acid, which is afterwards titrated back with standard sodium carbonate. Each c.c. of N/i acid corresponds to 0*0535 grm. of N H 4 C1, or o 066 of (N H 4 ) 2 SO 4 . The apparatus described for ammonia determination by distillation in Kjeldahl's method (p. 67), or a simple flask connected with a capacious U tube may be employed. Excess of soda solution may be run in from a tapped separating funnel, as shown in Fig. 8, or stick soda or potash suffi- cient to render the solution strongly alkaline may be dropped in wrapped in filter paper, and the apparatus rapidly closed to pre- vent loss. Liquids containing ammonia with organic matters should be treated like lime liquors, and if necessary free ammonia may be determined by simple distillation, and alkali then added to decompose the salts, and the distillation continued. Where lime or caustic alkalies are objectionable, magnesia may be used to liberate ammonia. Ammonia in waters and extremely dilute solutions may be estimated by Nesslerising instead of titration. THE MATERIALS USED IN CURING, BATING, ETC. 73 Sulphurous Acid and Sulphites. For ordinary alkalimetrical estimation, see p. 27. They may also be estimated by oxidation with iodine as follows : * About 0*5 grrn. of powdered sodium sulphite, or an equiva- lent portion of other salts or liquids, is introduced at once into a beaker containing excess (4Oc.c.) of N/io iodine solution, stirred till dissolved, and the solution is at once diluted with water and titrated back with N/io sodium thiosulphate, as described on pp. 258 and 306. Under these circumstances the SO 2 is oxidised completely to sulphuric acid, while if the iodine solution is added gradually to the sulphurous acid or sulphite, the reaction is irregular and incomplete. If it is found that considerable excess of iodine solution has been used, it is desirable to repeat the experiment with such proportions of the sulphite and iodine solution as will leave but slight excess of the latter. One c.c. of N/io iodine solution is equivalent to 0*0032 grm. of SO 2 ; 0*0126 grm. of cryst. sodium sulphite, Na 2 SO 3 + 7 Aq ; and 0*00475 grm. of sodium anhydrosulphite (Boake's " metabisulphite "), Na 2 S 2 O 5 . Giles and Shearer f have also investigated the specific gravity of sulphurous acid solutions, with the result that it is practically directly proportional to the strength, each per cent, of SO 2 corresponding to an addition of 0^005 to the gravity ; or in other words, that each degree Tw. corresponds to I per cent, of SO 2 with very considerable exactness. In the original article a table is given. The similarity of sulphurous to hydrochloric acid (p. 24) in this respect is worth noting. A saturated solution at ordinary temperatures contains about 10 per cent, of SO 2 . Sulphurous acid, while somewhat costly to buy, from the dilute character of its aqueous solution and consequent cost of transport, is one of the cheapest of acids to make by combustion of sulphur.f For use in small quantities, the compressed liquid acid in siphons is very convenient, and on a larger scale Boake's " metabisulphite " (see above) may often be advantageously em- ployed, as it contains 67 * 3 per cent, of its weight of SO 2 which can be liberated by the addition of i mol. of sulphuric acid (98), or 2 of hydrochloric (73), to I mol. of the salt (190), practically I part of sulphuric acid sp. gr. I -84 (164 Tw.) or 2 parts of hydro- * Cp. Giles and Shearer, J. S. C c I., 1884, p. 197 ; 1885, p. 303. t Loc. cit. \ Principles of Leather Manufacture, p. 24. 74 LEATHER INDUSTRIES LABORATORY BOOK. chloric acid sp.gr. 1*17 (34 Tw.) will decompose 2 parts of metabisulphite, leaving only sodium sulphate or chloride. A solution of i per cent, of sodium metabisulphite and 0*5 per cent, of sulphuric or i per cent, of hydrochloric acid answers well for bleaching and removing iron stains from leather, though its action is rather slow, about 12 hours' soaking being necessary. With the addition of 10 per cent, of salt, the same solution may be used for removing iron-stains (so-called " salt-stains ") from un- tanned pelt, but some salt must be added to the early liquors. The operation is best carried out in a closed drum. Boric (Boracic) Acid and Borates. These substances are coming into extended use in leather manufacture, not only as antiseptics but as deliming agents. Boric acid forms compounds with lime, soluble in dilute solutions, and is one of the safest chemical " bates." * It also exercises a peculiar and little under- stood effect on skins in tanning liquors, giving good colour and preventing " drawn grain " in the early stages of tanning, even when green goods are introduced into comparatively strong liquors ; but if continued throughout the process producing soft and light-weighing leather. This property is probably con- nected with the power of boric acid to form conjugated acids with tannins and their derivatives, as well as with other phenolic compounds. Boron, B = n, is tribasic, forming the oxide B 2 O 3 and the very weak acid B(OH) 3 = 62, known as boric or boracic acid. This acid, which appears to be dibasic, but forms no stable salts, is converted into metaboric acid HBO 2 by heating to 100 C, at which temperature it is slowly volatile. On heating to 140 C., it very slowly parts with more water and becomes pyroboric acid, H 2 B 4 O, a glassy mass which on ignition to redness is converted into B 2 O 3 . Both metaboric and pyroboric acid form salts, the former being monobasic and the latter dibasic, but both acids are weaker than carbonic. Borax is a pyroborate, and a calcium metaborate is known, Ca(BO 2 ) 2 , which is somewhat soluble in water, like all other known metallic borates. Detection. Free boric acid in either aqueous or alcoholic solution turns turmeric paper a reddish brown somewhat different * In the use of boric acid as a deliming agent, it is necessary to avoid th accumulation of lime borates, with which the solution soon becomes saturated, and is then, in spite of excess of boric acid, unable to remdve lime, but fixes it in the pelt. THE MATERIALS USED IN CURING, BATING, ETC. 75 to that produced by alkalies, and unlike the latter, blackened by acids instead of being discharged. Free boric acid is soluble in alcohol, and can consequently be separated from sulphates and many other matters by digestion with alcohol, to which the turmeric test is applied. Boric acid, though capable of with- standing the heat of the blowpipe in the dry state, is decidedly volatile when solutions in water are evaporated, and still more so with methyl and ethyl alcohols, with which it forms volatile methyl and ethyl borates or esters. Solutions containing it must therefore be neutralised or rendered slightly alkaline before evaporation to avoid loss. In presence of soda or sodium car- bonate it may be safely evaporated and even ignited. Small traces of the acid can be detected by taking the ignited residue and slightly acidifying with sulphuric acid in a test-tube. A little alcohol is now added, and the mixture is boiled and the escaping vapours are ignited. In presence of boric acid the edges of the flame are coloured green. Somewhat similar coloration may be produced by the presence of copper salts, or by ethylic or methylic chlorides produced by the action of concentrated hydrochloric acid on alcohols ; but in the latter case the flame is a bluer green than that of boric ester. If the traces of boric acid are too minute to be detected even in this way, they may still be separated by distillation as described below, and then tested for with turmeric or the alcohol flame. Quantitative Determination. It has been noted that boric acid, like carbonic acid, does not redden methyl orange, and the whole of the base in alkaline borates, such as borax, may be estimated with standard sulphuric or hydrochloric acid by the aid of this indicator, or of lacmoid paper, precisely as in the case of carbonates (p. 28). Borax (crystallised sodium pyro- borate Na 2 B 4 O 7 + ioAq = 382) often called Cl biborate," has been proposed as a basis for alkalimetrical solutions in place of sodium carbonate, but is only soluble in cold water to the extent of about 60 grm. per liter. 19*1 grm. per liter forms a solution N/io to acids with methyl orange as indicator. It has been stated by Dr. L. Spiegel* that it is occasionally, though rarely, contaminated with triborate, but this is somewhat doubtful. Boric acid gives no defined end-reaction with phenolphthalein,. but on adding neutral glycerin, a conjugated acid is formed * Chem. Zeit., 1904, p. 750. 76 LEATHER INDUSTRIES LABORATORY BOOK. to which the indicator is very sensitive, and advantage is taken of this fact in estimation. One curious result is due to this property. If glycerin is added to a solution of borax, which is distinctly alkaline to litmus, the reaction becomes acid. The formation of a similar compound with phenol is probably the cause of the very satisfactory deliming action of a solution con- taining 2 per cent, of boric acid and I per cent, of phenol, noted by Parker and Procter.* A saturated aqueous solution of boric acid and phenol is a useful antiseptic, which may be used for washing hands or applied to cuts and abrasions where septic poisoning may be feared. For the determination, 50 c.c. of the boric solution, which may contain 3040 grm. borax or 10-15 grm. boric acid per liter, is carefully neutralised in presence of methyl orange, with sul- phuric or hydrochloric acid if alkaline, or with caustic soda free from CO 2 if acid, and is then boiled to expel carbonic acid if its presence is suspected. 30 c.c. of glycerin (previously neutra- lised, if necessary, with caustic soda) is now added, and the solution is titrated in presence of phenolphthalein with N / 2 or N / 10 NaOH (perfectly free from CO 2 ), till a pink tint is pro- duced. 10 c.c. more glycerin is now added, when in some cases the pink colour will disappear, and more soda must be added to bring it back, and the addition of soda and glycerin alternately must be continued till the pink remains permanent. Under these cc nditions the end-reaction is quite sharp, and the alkali consumed corresponds to the equation B 2 O 3 = ,2NaOH ; or each c c. of N/i alkali corresponds to 0*035 grm. B 2 O 3 , or 0*062 grm. B(OH) 3 , or 0*0955 grm. of borax. If borax is titrated in this way withouc previous neutralisation to methyl orange, only half this quantity of acid will be consumed, or i c.c. will correspond to o* 191 grm., as in the case of alkalimetrical determination.f In the titration of boric acid solutions which have been used in deliming hides, only approximate results can be obtained from the presence of lime-salts of weak organic acids which render the end-point with methyl orange indistinct. These may be destroyed by evaporation to dryness and ignition. In many cases it is desirable to separate boric acid from other * Journ. Soc. Chem. Ind., 1895, p. 124. This book, p. 239. t Cp. Honig and Spitz, Zeits. f. ang. Chem., 1896, p. 549 ; Journ. Soc. Chem. Ind., 1896, p. 742 ; Thomson, ibid., 1895, p. 1070. THE MATERIALS USED IN CURING, BATING, ETC. 77 matters before determining it by titration or otherwise, and this is most conveniently achieved by taking advantage of its volatility when boiled with alcohols. The substance is freed from organic matter by rendering alkaline with sodium hydrate or carbonate, evaporating to dryness, and igniting. It is then powdered, and rendered slightly acid with sulphuric acid, and digested with loo c.c. of absolute alcohol in a flask with reflux condenser. The solution is then distilled into a flask containing 20 c.c. of N/i sodic hydrate free from CO 2 . The last drops of the dis- tillate are tested with turmeric paper for boric acid (alcohol alone gives a darkened yellow edge to the spot, but no red coloration), and if any trace is observed, 25 c.c. more alcohol are added to the residue, and the distillation .continued till all boric acid has passed over. The boric acid flask should be immersed in the water or steam bath up to its neck, so as to prevent conden- sation before the boric acid vapour reaches the condenser. The quantities given should suffice if the residue treated does not contain more than 0*5 grm. of boric acid. Where the quantity of foreign matter is considerable, it is better to filter off and distil the alcoholic extract only, after washing the residue with hot alcohol. The alcohol is now distilled off the receiving flask, leaving the boric acid in combination with the soda, which can be estimated by titration as above described. A less quantity of alcohol may be used, if it is distilled backwards and forwards between the two flasks by changing their positions and heating them alternately, but in this case it is desirable to use an equivalent quantity of a concentrated solution of soda, instead of the normal solution, to avoid dilution of the alcohol. By heating both the flasks, and connecting them with two condensers sloping in opposite directions, and somewhat contracted at their lower ends so that a drop of condensed liquid hangs in the end of the tube and checks entrance of the vapour in the wrong direction, both distillations can be conducted simultaneously ; and in this case the substitution of 25-50 c.c. of methyl-alcohol for the ethyl- alcohol is advantageous. The connection with the return-flow condenser should not be quite air-tight, so as to avoid risk of explosion. One hour and a half is said to be enough for the distillation of 0*3 grm. of boric acid.* * Cp. Jay and Dupasquier, Comp. Rend., cxxi. p. 260 ; Journ. Soc. Chem. Ind., 1895, p. 889 ; 1896, p. 136 ; and Dr. Schneider, Ch. Zeit., 1896, p. 822. 78 LEATHER INDUSTRIES LABORATORY BOOK, Carbolic Acids and Creolins. Pure phenol, " pure crystallised carbolic acid," is hydroxy-benzene C 6 H 5 (OH), but the crude forms which are generally employed contain cresols and other higher members of the series in which one or more atoms of hydrogen are substituted by CH 3 groups. These are oily bodies sparingly soluble in water, and even pure phenol is only soluble to the extent of about 7 per cent. Crude carbolic acid should not be employed in the tannery, since the insoluble oily particles stain the hide and render it insusceptible of tanning. Suitable carbolic acid should be of a pale yellow colour when fresh (though it will darken on exposure to air and light), and it should be wholly soluble in a sufficient quantity of water. Its specific gravity should be I '050 to I "065. The actual analysis of crude carbolic acids is beyond the scope of this book, but the following tests, described by Allen,* will give an approximate idea of their value. They consist mainly of phenol and its higher homologues, but frequently contain water and neutral tar-oils of little direct value as antiseptics. Determination of Water. 20 c.c. of the carbolic acid are agitated in a stoppered graduated cylinder or corked tap-burette with 10 c.c. of saturated salt solution. The water will pass into the salt solution and the volume may be read off. The whole of the water is not actually removed : pure anhydrous phenol actually absorbs a small portion of the water, precipitating salt. The volume of 10 grm. of pure crystallised phenol, after shaking with salt solution, is about 9*7 c.c., from which the actual weight of the anhydrous phenols may be inferred, though their specific gravity will be somewhat greater than that of pure phenol. Determination of Neutral Oils. 10 or 20 c.c. of the carbolic acid are agitated with twice its volume of a 9 per cent, solution of caustic soda free from alumina. The phenol and cresols will dissolve, and the neutral oils will remain floating on the aqueous layer if " light," and sinking below it if " heavy." Their volume may now be noted. The addition of 10 c.c. of petroleum spirit facilitates their separation, and causes them to float. The volume of spirit added must of course be deducted from the reading. Creolin (Jeyes' Fluid) and similar disinfectants, which are * Com. Org. Analysis, ii., p. 545. THE MATERIALS USED IN CURING, BATING, ETC. 79 emulsions of the heavier homologues of phenol with soap or an alkali, may be similarly examined by adding sufficient normal acid to redden methyl orange to 20 c.c. in a burette, arid then saturating with salt, when the phenols and cresols will separate in oily drops. The acid is run off by the tap, and the cresols washed with saturated salt-solution, which is added to the acid and titrated back with normal soda, when the loss will correspond to the alkali originally present. The neutral oils may be deter- mined as in carbolic acid, but often separate with great difficulty. Formaldehyde (H.CHO = 30). The simplest member of the aldehyde series, obtained commercially by the oxidation of methyl-alcohol, CH 3 OH, and now used largely both as a disin- fectant, and as a tanning and hardening agent for skin and other gelatinous substances, which it renders insoluble even in hot water. It is usually sold as an aqueous solution of approxi- mately 40 per cent, strength under the names of u formaline," " formol," etc., the pure substance being gaseous at ordinary temperatures, but readily polymerising or combining with itself to form a white crystalline body, tri-formol or paraform, C 3 H 6 O 3 , which sublimes at iooC, and is decomposed again into formalde- hyde at somewhat higher temperatures. This body is frequently formed as a white sediment in formaldehyde solutions. The specific gravity of formaline varies from I 08 to I * 09, not only from the percentage of formaldehyde, but from that of methyl- alcohol, which is frequently present to the extent of 15-20 per cent., either as an impurity of manufacture or to prevent pre- cipitation of tri-formol. Both as a bactericide, and in its hardening action on gelatine and animal tissues, formaldehyde is extremely powerful. It is stated that I part of formaldehyde or 2j parts of formaline in I2,OOO parts of water is sufficient to sterilise, while in Payne and Pullman's tanning process* a solution of 2\ parts of formaline in 1000 of water rendered alkaline with sodium carbonate will convert hides into leather in a few hours. In acid solution it has the power of fixing hide in a swollen condition and allowing of very rapid tannage with vegetable materials. This hardening and tanning action seems to be common to all aldehydes in a greater or less degree. Detection. Aldehydes in general are usually easily detected * Principles of Leather Manufacture, p. 380. 8o LEATHER INDUSTRIES LABORATORY BOOK. by their power, even when present only in traces, of reddening a solution of rosaniline previously acidified, and nearly decolorised by addition of sodium sulphite (Schiff's reaction). This property is common to all aldehydes, including acetaldehyde, chloral (but not chloral-hydrate), benzaldehyde, furfurol, acrolein, etc., and is slightly produced by some alcohols and by acetone, but not by most other ketones. It is, therefore, not absolute proof of the presence of formaldehyde, and it also seems incapable of detecting it in actual combination with gelatinous matter, though most leathers which have been treated with formaldehyde retain enough in a free state to give the reaction readily. It must be remem- bered that oil- and chamoised leathers may react from the pre- sence of acrolein (acrylic aldehyde), which is probably the active agent in oil-dressing. Nierenstein has also shown* that most tanning infusions give a red coloration with it on standing, from the reduction to alde- hydes of the tannin-acids by the sulphurous acid present. The colour of the reagent also reappears by oxidation and loss of SO 2 on exposure to the air. The reagent is in the author's experience most conveniently prepared by passing SO 2 very slowly into a 0*025 per cent, solution of magenta till it is just decolorised, and may be kept for a considerable time. At the author's request, Dr. Nierenstein has examined the most important qualitative methods for the detection of aldehydes, and gives the preference \ to Lebbin's, J but with the modification of employing phloroglucol instead of resorcinol. This test is carried out by adding to a few c.c. of the liquid to be examined 2-3 c.c. of a per cent, solution of phloroglucol, or a small trace of the dry substance on the point of a penknife, rendering alkaline with 5 or 10 drops of caustic soda or potash solution, and boiling (rapidly in order to prevent darkening by oxidation). Formal- dehyde gives a brownish-red, and acrylaldehyde a blue-green coloration, so that they are readily distinguished. The reaction is considerably less sensitive than that of Trillat, but is much less tedious in execution. Trillat's reaction is obtained by adding 0-5 c.c. of the purest dimethylaniline (boiling between 190 and * Collegium, 1905, p. 159. t Ibid. 1905, p. 158. .i^.- i Lebbin, Chem. Centralblatt, 1896, ii. 930. Comp. Rend., 1898, 127, 132. THE MATERIALS USED IN CURING, BATING, ETC. 81 191) to 10 c.c. of the suspected liquid, acidifying with sulphuric .acid and shaking thoroughly. The liquid is now rendered alka- line with sodium hydrate, boiled till free from smell of dimethyl- aniline, and the precipitated tetramethyldiamidodiphenylmethane is collected on a filter, washed, and treated in a small porcelain basin, first with acetic acid and then with a little finely powdered lead peroxide, when a blue colour is produced. Quantitative Estimation of Formaldehyde. For general use -where accurate results are required, and in absence of bodies which reduce iodine, such as thiosulphates, sulphides and sul- phites, tannins and the like, Romijn's method * is one of the most satisfactory and rapid, especially for dilute solutions. It depends on the oxidation of the formaldehyde to formic acid, accord- ing to the equation : H.CHO + 2NaOH + 2! = HCOOH + 2NaI + OH 2 ; and hence each molecule of formaldehyde re- quires 2 molecules of iodine for oxidation, and each c.c. of N/5 iodine, or indirectly of thiosulphate, is equal to 0*003 grm. (i molecule) of H.CHO. The formaldehyde solution must not exceed about 2 per cent in strength, and therefore 5 c.c. of 40 per cent. " formaline " must be diluted to ioo c.c. with distilled water, and well mixed by shaking. N/5 iodine solution is made by dissolving 12*685 grm. of pure iodine (if necessary obtained from the commercial by sublimation with a little potassium iodide) with 1 8 grm. of pure potassium iodide in about 250 c.c. water and making up to 500 c.c. The iodide must be free from iodate, which may be tested for by dissolving a little in water and acidifying with hydrochloric acid, when the solution will become yellow and give the usual blue reaction with starch if iodate be present. A N/io solution of sodium thiosulphate is also made by dissolving 24-83 grm. of the pure salt, and making up to I liter, which must exactly correspond to the N/io iodine solution. The -accuracy of the thiosulphate solution, and indirectly that of the iodine, may be tested against a standard solution of potassium bichromate (see p. 259). The thiosulphate is best kept in the dark, and is liable to deposit sulphur and diminish in strength if kept more than a few weeks. Instead of thiosulphate, a N/io solution of sodium arsenite may be used, which is not liable to change, but is very poisonous. * Zeit. Anal. Ch., 1897, No. 36, pp. 18, 21. G 82 LEATHER INDUSTRIES LABORATORY BOOK. To make the estimation, 5 c.c. of the diluted formaldehyde solution and 30 c.c. of N/i NaOH solution are measured into a bottle of about 500 c.c. capacity, and with a well-fitting stopper, and 70 c.c. N/5 iodine solution is added from a burette with constant shaking. The bottle is closed and again well shaken- and" allowed to stand for 10 minutes, when it is acidified with 40 c.c. of N/i HC1, and titrated back with N/io thiosulphate solution, a little starch-paste being added as the yellow ceases to be distinct, and the end-point being that at which the blue of the starch first completely disappears. 70 c.c. of the N/5 iodine solution corresponding to 140 c.c. N/io thiosulphate have been used and supposing 73 '3 c.c. of the latter are required for titration back, iodine equal to 66*7 c.c. of the N/io thiosulphate or 33*4 c.c. of N/5 iodine have been con- sumed, equal to 33*4 X 0*003 = O'l grm. of formaldehyde, or if 5 c.c. of a solution of formaldehyde diluted 5 : 100 have been employed It is necessary to employ considerable excess of iodine to insure complete oxidation of the formaldehyde. It is safest to compare the iodine solution direct with that of thiosulphate or arsenite at the time of making the estimation, and take any little inaccuracy into calculation, and the addition of the NaOH and HC1 should not affect this result. If many estimations have to be made, a N/5 solution of thiosulphate may be conveniently sub- stituted for the N/io, which, however, is more generally useful. Starch paste is best made as required, and only a minute quantity is added to the solution towards the end of the reaction. In cases where the use of iodine is impracticable from the presence of other reducing substances, the method of Blank and Finkenbeiner* dependent on the oxidation of formaldehyde to formic acid in alkaline solution by hydrogen peroxide may some- times be employed, and is of equal accuracy to the preceding ; but the presence of other substances giving acid products under oxidation must be guarded against, and it is less adapted than Romijn's for the determination of very minute quantities. 3 grm. * Berichte 1898, 2979. THE MATERIALS USED IN CURING, BATING, ETC. 83 of the (approximately 40 per cent.) solution * are weighed into a tall conical (Erlenmeyer) flask with 25 c.c. of double-normal NaOH solution, and 50 c.c. of 2*5-3*0 per cent, hydrogen peroxide previously neutralised with NaOH are slowly added during about three minutes ; and the mixture is allowed to stand at least half an hour, and is then titrated back with double-normal H 2 SO 4 or HC1 in presence of phenolphthalein. If 30 c.c. of the 2N sulphuric acid are required to neutralise the soda, 20 c.c. of the latter have been consumed by the formic acid formed and,, multiplied by 2, give 40 per cent, as the strength of the formalde- hyde solution, the calculation being (50 - 30) X 2 X 0*03 X IQO 3 Of course, if a larger or smaller volume or weight than 3 grin, or c.c. have been employed, it must be substituted for 3 in the divisor. It must be borne in mind that in all cases of analysis,, if a liquid has been measured instead of weighed, the result will be obtained in grm. per 100 c.c. and must be multiplied by the specific gravity to get percentage by weight. Another method, probably less accurate than the preceding r but very convenient as a test for commercial formaline, was first published by Lemme,f but had been long previously used by E. M. Payne as a general method for the estimation of aldehydes. It depends on the power of aldehydes to combine with bisul- phites, which in the case of formaldehyde, acetaldehyde, and many others, is sufficiently strong to decompose neutral sodium sulphite, setting free sodium hydrate, according to the equation,. H.CHO + Na 2 SO 3 + OH 2 = CH 2 OHNaSO 3 4- NaOH. To 60 c.c. of a solution of pure Na 2 SO 3 . 7 Aq of 250 grm. per liter, a few drops of phenolphthalein solution are added and the red colour produced is just barely discharged by the careful addition of H 2 SO 4 or HC1 ; 3 c.c. or 3 grm. of the formaline are now added, when the red colour will return, and must be again discharged with N/ 1 acid. As each c.c. of the latter corresponds to o 03 * Fresenius and Grunhut, Zeits. f. Anal. Ch., 44, 1905, 13-24, recommend weighing in a tube which is introduced upright into the Erlenmeyer flask which already contains 25-30 c.c. of double-normal NaOH, and there upset and mixed, immediately adding the H 2 O 2 as above described. They recommend litmus or azolitmin as indicator. t Ch. Zeit., 1903, 896. G 2 84 LEATHER INDUSTRIES LABORATORY BOOK. grm. of H.CHO, the percentage either on volume or weight will correspond directly to the number of c.c. used. Where formaldehyde cannot be estimated or detected in the original mixture in which it exists it may be separated by dis- tillation, preferably with a little dilute sulphuric acid to decom- pose proteid compounds. The cooling must be good and the distillation carried nearly to dryness, if the results are to be at all quantitative, though a good deal of the formaldehyde comes over in the early stages. Formaldehyde leathers should be digested for 1-3 hours with I c.c. of N/io sulphuric acid, and 20 c.c. of water on the water-bath, either in a closed flask, or connected with the condenser. A good criticism on the various methods of estimating formaldehyde by Wallnitz is contained in Collegium, 1903, pp. 105, 115. " Coal-Tar" Bates. Several derivatives of coal-tar products have been used for deliming purposes, for which, but for cost, they are well adapted. Carbolic acid (phenol) has considerable effect in removing lime, and has been used for that purpose (p. 239) by Parker and Procter, especially in connection with the colour-testing of liquors. Sulphonated derivatives of naphtha- lene, phenanthrene, and other hydrocarbons, probably obtained by treating the heavy, or " dead," oils of coal-tar with concen- trated sulphuric acid, have been put on the market by the Martin Dennis Co., and by Hauff, of Feuerbach. The latter firm also introduced " cresotinic acid," a mixture of impure homologues of salicylic acid ; and also some allied derivatives of the naphthalene series. Most of these compounds possess veiy pronounced anti- septic powers, and hence may be usefully employed for deliming sole leather hides in hot weather, when otherwise putrefactive changes are liable to interfere with swelling in the liquors. Owing to the varied character of these materials it is obvious that no general directions can be given for their chemical ex- amination. SECTION VII. CONTROL METHODS FOR LIMING, BATING, ETC. IT is becoming recognised that no processes of the tanner are more important and more in need of exact scientific control than those which precede the tanning proper, for which they prepare the skin by removal of hair and certain portions of the cementing substance of the fibres, and by bringing the latter into suitable conditions of swelling and hydration for the par- ticular sorts of leather which are required. Unfortunately the scientific methods of control with which we are acquainted still leave much to be desired, but an effort will be made in the following section to point out what can already be accomplished. Soaks and Soaking. Estimation of dissolved organic matter, and ammonia may be made as in limes, but if total ammonia is to be estimated, it must be set free from its salts by addi- tion of calcined magnesia. Peptones may be estimated by precipitation (see pp. 89, 90). Mineral salts will be found which have been used in curing. Sodium (and possibly calcium and magnesium) chlorides and sulphates, sodium hydrate, sulphide, and carbonate, boric acid or borates, and arsenic are all probable. Arsenic may be looked for by the ordinary Marsh test. Many of the salts named have a solvent action on hide-substance, especially in medium dilutions, and their concentrated solutions interfere with proper swelling. Limes and Liming. The examination of lime itself has been described on p. 54. The total caustic lime present in a lime- liquor or sludge is easily estimated by titration with standard hydrochloric acid with phenolphthalein as indicator, the ammonia being generally negligible in comparison with the lime, or expelled by a short boiling. Very exact results are not to be expected in the case of old lime-liquors, which contain weak organic acids and bases which render the change of colour of the indicator less sharp. Where much carbonate is present, as in the case of lime sludge, the material must be freely diluted with water, and the titration done with sufficient stirring and somewhat slowly. *6 LEATHER INDUSTRIES LABORATORY BOOK. Where soda or sodium sulphide is used in the limes, caustic soda may be present in the lime-liquors, and will be estimated as lime. Before it could be accurately separated from lime, it would probably be necessary to destroy organic matter by evaporation to dryness and ignition, but a filtered lime-liquor may generally be assumed to be a saturated solution of caustic lime, the strength of which is given on p. 54, and therefore anything over about I * 3 grm. per liter may be reckoned as caustic soda, ammonia, or some other alkali more soluble than lime. Accumulations of caustic soda are liable to occur in limes which are used for reliming skins or hides which have been depilated with sodium sulphide, and if not watched, may lead to disastrous results. Such cases are usually best dealt with by making new limes at frequent intervals, but the addition of calcium chloride will remove caustic soda by converting it into chloride, and regenerating caustic lime, as in the Payne-Pullman liming process. For this reason it is sometimes useful to add calcium chloride to limes in which sodium sulphide is used if it is desired to avoid swelling, and by doing so, the effects of arsenic-limes may be closely imitated by the much cheaper and safer sodium sulphide. Considerable quantities of lime are contained in lime-liquors in clear solution neutral to phenol phthalein in combination with organic matter. The total lime present is easily estimated by evaporating a measured quantity of the filtered liquor to dry- ness, weighing the total solids, igniting, and weighing the ash as calcium carbonate (see p. 45), after moistening with ammonium carbonate, and again gently igniting. If other mineral matters are present the residue of carbonates from the ignition of organic matter may be dissolved in hydrochloric acid, and the lime pre- cipitated as oxalate will give the total lime present, both caustic and as salts of organic acids ; and by deducting the former, the weight of the latter may be obtained. If soda is present from the use of sulphide of sodium or soda ash in the limes, it may be first separated by washing the ignited and carbonated residue on a filter and titrating the solution with N/io HC1 and methyl orange. The insoluble calcium carbonate on the filter is then determined by dissolving in a known quantity of standard acid, and titrating back with methyl orange and HC1. From it must be deducted the amount found as caustic lime. Strictly, the CONTROL METHODS FOR LIMING, BATING, ETC. 87 lime and alkalies should be calculated as oxides, and the corre- sponding amount of CO 2 added to the organic matter. As it is not easy to keep a milk of lime so thoroughly mixed that equal quantities of the suspended lime can be pipetted off, it will generally be best to dilute with such a quantity of distilled water (free from CO 2 ) as will keep all the actual caustic lime in solution, as in the estimation of "available lime." Methyl orange is generally useless as an indicator in the case of lime- liquors, as it gives no sharp end-reaction from the presence of weak organic acids and bases. Ammonia has an important influence on liming, swelling less but dissolving more hide-substance than lime alone. Stiasny * has shown, however, that on mixture with lime, this effect is pre- vented, or greatly diminished, probably by the formation of complex lime-ammonia salts. It is usually present in greater or less quantity, being evolved from the hide-substance itself by the action of lime. It is best determined either by Procter and McCandlish's method (p. 60) or, preferably, by distillation (p. 63). In the latter case it must be borne in mind that old lime liquors are apt to froth excessively in boiling. Care is necessary in drawing samples, to take them from the freshly plunged lime, and to avoid loss of ammonia. Samples should be analysed without delay,as ammonia maybe evolved from the organic matter present. Amines (substituted ammonias) are contained in most lime liquors, and are less alkaline but otherwise similar in their effects to ammonia itself. By the distillation-method they will be mostly determined with the ammonia, but as they are less vola- tile, will probably be little estimated by Schloessing's process. Amido-acids are also present, but no convenient method is known for their determination. \ Alkaline Sulphides or sulphydrates derived from the addition of sodium or arsenic sulphide to the limes may be estimated by the zinc method given on p. 55. Not much is known either of their chemical action on the hair, or of what becomes of them in the process, but they usually disappear somewhat rapidly. Further investigation is very desirable. Arsenic may be detected by the Marsh test ; for estimation see p. 58, remembering the presence of traces of organic phosphorus. * Gerber, 1906, xxxii., pp. 200 o mgr. of acetic or 90 mgr. of lactic acid, the calculation becomes one of simple rule-of-three. While there is no doubt that the principal effect of the bran- drench is due to the organic acids produced by bacterial fermenta- tion, it has been noticed that a mere unfermented infusion is not without action in removing lime, and a part of its effect has been attributed by Mr. E. M. Payne in conversation with the author to the presence of a peculiar organo-phosphoric acid, which has been shown by Patten and Hart* to exist in bran in combination with magnesium and small quantities of other bases, and which is probably identical with the anhydro-oxymethylene-diphosphoric acid extracted by Posternakf from beans and other seeds, to * Bull. 250 New York Ag. Exp. Station ; Ch. News, 1904, p. 112. t Revue Generate de Botanique, 1900, xii. pp. 5 and 65. Compt. Rend., cxxxvii. :Nos 3, 5, 8. CONTROL METHODS FOR LIMING, BATING, ETC. 97 (CH 2 -O.PO(OH) 2 which he gives the formula O - acid is a thick syrupy liquid, soluble in water. Its alkali salts are also soluble, those of magnesia and lime less so, and those of barium and strontium very sparingly. The magnesium and calcium salts are soluble in acetic acid. It exists in bran to the extent of about 2*75 per cent. It is somewhat analogous to the glycerophosphoric acid obtained by mixing glycerin and meta- phosphoric acid, and if further investigation proves that it is of importance in the drenching process, it would be worth experi- menting with the glycerin compound as a substitute. Many substitutes have been proposed for bran-drenching, usually containing glucose, sugar, or other vegetable matters brought into fermentation, often with little knowledge of the proper fermentation to be produced. Weak glucose solutions with a little mineral nutriment like Cohn's solution (p. 442) set with the milk-ferment give good results, and could probably take the place of bran fermentation with profit in many cases. If fresh milk beset with a little sour milk, and the process repeated from day to day, an almost pure culture of the milk-lactic bac- terium results, as no other organism thrives so well under the conditions. The method has long been used in producing a " starter " in souring milk for butter-making. The true drench bacteria are probably originally derived from the puers and cannot long be kept in a state of activity as a separate culture, without fresh nourishment, though in the drenching process they are sufficiently abundant about the vessels used to secure a continuance of the fermentation without intentionally added ferment. Acids, both mineral and organic, have now in many yards largely taken the place of drenches, and even of puers, as a means of deliming, or to complete and shorten the fermentative process, which is always more or less dangerous to the goods. Lactic, acetic and formic acids in particular are very complete substitutes for drenching. Hydrochloric, and even sulphuric acids are successfully used even on light skins, for reducing the lime to a small residue, the process being usually completed by drenching or the use of organic acids. These stronger acids can also be employed to liberate the weaker organic acids or boric acid from the lime salts formed, and so economise the quantity H 98 LEATHER INDUSTRIES LABORATORY BOOK. required in strengthening for successive parcels of skins. The acid effect of all organic acids is lessened by the presence of their own neutral salts in the solution, and the swelling power of mineral acids may be controlled by additions of salt ; while organic salts added to solutions of mineral acids act as regulators of their action, liberating organic acids as soon as the mineral acid is in excess. For further details * Principles of Leather Manufacture,' pp. 154 et seq., may be consulted. It is obvious that in all these cases accurate determination of the acids, both in the original solution and after use, is of great importance, but the ordinary alkalimetrical methods will mostly suffice, methyl orange being usually the appropriate indicator for the mineral acids, and phenolphthalein with caustic standard solutions or with lime and baryta water for the organic. It is not easy to devise simple methods for the separate estimation of organic and mineral acids in mixture, though in some cases the different sensitiveness of indicators can be taken advantage of. If mineral acids are used to decompose the salts of organic acids it is usually well to be sure that the salts are in liberal excess. Boric acid may be estimated in presence of mineral acids by the methods given on p. 76, but very accurate results are not generally to be expected, since the organic acids from the skins usually affect the sharpness -of the change of colour of methyl orange. The total boric acid can of course be accurately esti- mated by adding excess of base, and destroying the organic matter by ignition. Control of Hides and Skins. Beside the control of the liquors it is very important to adopt scientific methods in the examina- tion of the hides and skins themselves during the preparatory processes, and especially, the tanner constantly wants to ascertain at the earliest moment the actual weight of dry leather-forming matter in his raw material, and consequently, the maximum weight of leather which he can expect to obtain. Attempts have frequently been made to calculate this from the weight of the soaked or limed pelts, but though these figures are useful guides in regular work, they are very variable, owing to the different moisture, amount of hair, and so on, in hides from the soaks, and the variability of the swelling in limed hides. Actual drying and weighing of portions of hide-matejrial itself is im- practicable both from the length of time which it occupies, and CONTROL METHODS FOR LIMING, BATING, ETC. 99 the impossibility of getting really average samples. By far the best method is that of the determination of weight in water, a method probably first suggested by the writer to his students, but afterwards worked out in more detail by Carini,* and tested on a large scale by C. E. Parker and G. H. Russell. f In principle, this depends on the fact that water weighed in water weighs nothing, and that consequently the water-weight of a hide is simply that of the actual dry hide matter, less that of the water actually displaced by the dry matter itself. If a thoroughly soaked sponge be hung in water by a fine wire, from a delicate spring balance, it will be found to weigh exactly the same, whether squeezed out or allowed to expand ; and in the same way, the swelling or flatness of the hide makes no difference to its water-weight unless indeed the actual density of the hide- substance is altered by some chemical change. Parker and Russell prefer to weigh the hide or skin by at- taching it by a thin copper wire to a delicate spring balance (Salter, Birmingham), or, with still greater accuracy, to a beam balance made for a similar purpose by Reimann, of Berlin. As the water- weight is only about 1/16 of that in air, a delicate balance must be used, and care taken not to let the weight come upon it till the hide is really immersed in the water, and it must be hung in a deep pit or vessel in such a way that it does not inclose any air in its folds, and does not touch the sides of the pit. From the water-weight, the weight of dry matter is ob- tained simply by multiplication by a factor C, and deduction of the dry weight so found from the weight weighed wet in air gives the amount of water contained in the hides as weighed. The factor C will probably prove to be a constant when accu- rately ascertained for pure hide-substance, or at least to vary within very small limits. It may be determined either directly by dividing the water-weight of an average sample of hide by its actual dry substance as found by drying at 100 C., weighing, and deducting any lime or other mineral substances present (re- membering that the lime must be calculated as CaO, and not as carbonate), or the sample may be Kjeldahled, and the dry sub- stance calculated from the nitrogen found, which would bring it into harmony with the hide-substance as determined in the * Lancashire Tanners' Federation Year Book, 1905, p. 45. t Collegium, 1905, pp. 161, 171. H 2 ioo LEATHER INDUSTRIES LABORATORY BOOK. leather by the same method. The factor C may also be deter- mined by calculation from the specific gravity S of hide-substance, if the latter is known, as C = S /(S - i). On the whole, the first method is to be preferred as more direct, as it is not certain that the specific gravity of the swollen hide is exactly that of the hide dried at 1 00 C. If the gravity is taken direct, great care should be taken, especially in the more porous sorts of hide material, that no air is contained in the pores. This will pro- bably be best attained either by actually weighing in water after placing the pycnometer for some time under an air-pump receiver to remove air, or by weighing in paraffin oil after pro- longed soaking. Gravity can, of course, be calculated from the factor C, since I/ (C i) = (S i). The presence of fat and other foreign matters, such as hair on the hide, will of course introduce errors, but these in most cases will be of small importance. The gravities are supposed to be taken at 1 5 C. or 60 F., but small differences of temperature are quite negligible ; a temperature of 25 C. only introducing an error of about 0*2 per cent, which is much less than probable errors of weighing. The influence of lime contained in the lime-water absorbed by the hide is some- what greater, but will be allowed for in the factor, if this is determined experimentally under the same conditions. As in any case, rather relative than absolute results are required, the exact accuracy of the factor is not of much importance, so long as the same is always adhered to. Parker and Russell give the gravity of hide-substance in plumped butts as i '425, which gives a factor C of = 3*353, and in plumped shoulders and bellies as i -374, giving C = 3*674. They themselves, however, calculate the factors as C = 3 * 38 and 3 * 754 respectively. They quote the following from Carini as the result of many determinations of pelt dried at iooC. Hides ex soaks with hair on .. .. C = 3*216 Sp. gr. = i'45. ,, limes, unhaired C = 3 353 Sp. gr. = i -425. ,, unhaired with sodium sulphide C = 3*227 Sp. gr. - 1*441. Parker and Russell remark on the importance of the degree of swelling as influencing the tannage ; and in this they are un- doubtedly right, not only as regards sole-leather, but probably to a still larger degree in the case of skins and dressing goods, in which it is well known that the more the swelling is reduced, CONTROL METHODS FOR LIMING, BATING, ETC. 101 the softer is the leather. As a measure of swelling they calculate what they call the " apparent specific volume " as (air- weight water-weight) / (dry hide-substance). This figure does not seem to represent any actual physical quantity, but is ap- parently founded on a misconception, and it is surely better either to calculate the total wet weight of one part of dry sub- stance, which is done by dividing the actual wet weight by the calculated dry substance, or the quantity of water held by one part of dry substance, which is of course one less than the above, since both of these figures are obvious actual quantities. There is no need that this method of determining hide-sub- stance should be confined to limed hides, as it is probably equally applicable in any stage of the wet-work, and should give valuable results as to the effect of bating, puering and drenching in reducing swelling. It is also applicable with some modifications to leather in process of tanning. Lime in Hides. The presence of caustic lime in hides is easily detected by moistening a freshly cut edge with solution of phe- nolphthalein, which will turn red or pink according to the quan- tity of lime present. Caustic soda or sodium sulphide gives a similar reaction. This test is extremely useful in judging of the extent of deliming by bates or drenches. The quantitative deter- mination of caustic lime in the hide is somewhat difficult and troublesome, but is best done by a method devised by Mr. E. M. Payne. The hide or skin is cut into thin shavings, avoiding exposure to air as much as possible, and is weighed into a stoppered bottle with distilled water free from CO 2 and some phenolphthalein. N/io HC1 from a burette is now added till the red colour of the liquid is just discharged, and the bottle shaken and allowed to stand, when the liquid will again be reddened by lime diffusing from the hide. Further additions of acid are made very slowly and over many hours, till the red colour seen in the shavings just disappears, and the total acid used is calculated into caustic lime. Total lime is determined by drying and igniting the sample, dissolving in excess of N/i or N/io HC1 (according to the quantity of hide taken), and titrating back with standard caustic soda and phenolphthalein, or, if it be desired to include sulphate or chloride, by precipitation as oxalate (see p. 45). Payne* * Lancashire Tanners' Year Book, 1905, p. 75 j Collegium, 1905, p. 184. 102 LEATHER INDUSTRIES LABORATORY BOOK. gives the following percentages of lime calculated as calcium hydrate in hide dried at 100 C. Per cent. Hide fully limed 3*766 Hide bated , ..* .. 0-689 Calf limed Y> 2'6oi ,, bated 0*122 Goat limed 5*613 .. ... % .. .. .. .. .* ... .. 5'778 ,, bated .. 1*268 The differences are very striking, and show the need of further investigation. Acid in Hides and Skins. If skins have been delimed or swollen with acid, it is sometimes desired to determine the excess which they contain. Total acid may of course be estimated by reversing Payne's process for caustic lime (vide supra} and titrating with caustic alkali or lime-water till a faint permanent red appears. Another and less tedious method is to place the weighed portion of hide in a bottle with distilled water, and a weighed portion of freshly ignited magnesia, and allow it to stand with occasional shaking till the acid is completely neutralised, which with thick hide may take many hours. The point of neutralisation may be determined by testing a freshly cut surface with litmus paper. The hide is carefully freed from adhering magnesia, and the whole of the undissolved MgO is collected on a filter, ignited and weighed, when the loss may be calculated into total acid. Probably methods could also be devised taking advantage of the fact that borax or sodium bicarbonate does not swell hide, but can be titrated as alkali in presence of methyl orange. Free mineral acids may be estimated in presence of their salts (as in pickled skins) by adding excess of N/i NaCO 3 to a weighed portion, evaporating the whole to dryness, igniting gently, extracting the ash with water, and then completing the ignition, finally adding the extracted solution, evaporating to dryness, and again gently igniting, and titrating back the soda solution, when the loss will represent mineral acids. The error to be guarded against is the reduction of sulphates by the or- ganic matter, and their subsequent conversion into carbonates. For this reason the salts are as far as possible removed from the organic matter by washing, either before drying or after very gentle ignition. The use of most oxidising agents would be THE CHEMISTRY OP TANNINS, ETC. 103 inadmissible, as they would produce variations in the alkalinity of the residue. (Compare p. 369 on the estimation of free acids in leather.) The loss of sodium chloride is of no consequence as it is a neutral salt, and calcium or magnesium chlorides which are dissociated on ignition cannot exist in the material after treatment with sodium carbonate. SECTION VIII. THE CHEMISTRY OF TANNINS AND THEIR DERIVA TIVES. THE essential constituents of tanning materials are various members of a large group of organic compounds called tannins, or tannic acids,* which are widely distributed throughout the vegetable kingdom, and which are said to occur in one insect, the corn weevil. Their function in vegetable physiology is yet uncertain, and in some cases they appear to be waste products of organic changes. These bodies often differ widely both in chemical constitution and reaction, but have the common property of precipitating gelatin from solution,f and forming insoluble compounds with gelatin-yielding tissues. By virtue of this power, they convert animal hide into the insoluble and imputrescible material called " leather." They are mostly uncrystallisable ; and all form blackish-blue or blackish-green compounds with ferric salts, and, in common with many other organic substances, are precipi- tated by lead and copper acetates, stannous chloride and many other metallic salts. In some cases the tannin combines with the base only, liberating the acid ; but frequently the salt as a whole enters into combination. This is the case with the preci- pitates formed with lead and copper acetates. Tannins are also * Ger. Gerbsaure ; in German, the word Tannin denotes usually gallotannic acid only. t Certain bodies have been described as tannins, such as " moringatannic acid," or maclurin, and " caffetannic " and " lupulotannic " acids, which do not precipitate gelatin or tan hide, and, in the writer's opinion, are more closely related to the colouring matters than the tannins. 104 LEATHER INDUSTRIES LABORATORY BOOK. precipitated by many organic bases, such as quinine, morphia, and most of the basic colouring matters. With alkalies, the tannins and many of their derivatives give compounds which oxidise and darken rapidly, usually becoming successively orange, brown, and black. A. H. Allen has shown that these bodies also give instantaneously a deep-red coloration with a solution of potassium ferricyanide and ammonia. The reaction is one of considerable delicacy. Tannins are more or less soluble in water and glycerin ; and freely so in ethyl and methyl alcohol, mixtures of alcohol and ether, acetone and ethyl acetate, but scarcely in dry ether alone, nor in dilute sulphuric acid ; and insoluble in carbon disulphide, petroleum spirit, benzene and chloroform. A. G. Perkin* points out that the colouring matters of plants usually yield the same decomposition products as the tannins associated with them, though this is not invariably the case. The following table shows this in a convenient form : Tanning Material. Tannin Contained. Decomposition Products of the Tannin. Colouring Matter. Decomposition Products of the Colouring Matter- Quebracho Colorado Quebracho- tannin Phloroglucol f and Protocatechuic acid Fisetin. Resorcinol. Protocatechuic acid. Rhus species Gallotannin Gallic acid . M>ricetin . Phloroglucol. Gallic acid. Catechus . Calechin . Phloroglucol and Protocatechuic acid Quercetin . Phloroglucol. Protocatechuic acid. Cape Sumach A catechol- tantiin ' Protocatechuic acid . Quercetin . Phloroglucol and Protoca- techuic acid. Divi-divi . Ellagitannin ^' Ellagic acid | The natural tannins are all compounds of carbon, hydrogen and oxygen only. They all contain the benzene group of carbon atoms, but their ultimate structure is, except in the case of gallotannic acid, very imperfectly understood, and probably differs considerably in type -in different members of the family. They may be regarded as derivatives of the phenols * Dr. Nierenstein (Collegium, 1905, p. 69) shows that Quebracho also contains a tannin which yields resorcinol. t Journ. Chem. Soc. 1898, pp. 374, 665, 1016. THE CHEMISTRY OF TANNINS, ETC. 105 (hydroxybenzenes) and their corresponding acids, of which a general view is given in the following table : C 6 H 6 Benzene. C 6 H 5 OH Phenol. C 6 H 4 (OH) 2 Catechol (i : 2). Quinol (i : 4). Resorcinol (i : 3). C 6 H 3 (OH) 3 Pyrogallol (1:2:3). Phloroglucol ( I : 3 : 5). Hydroxyquinol (1:2:4) C 6 H 5 CO.OH Benzole acid. C 6 H 4 .(OH)CO.OH Salicylic acid (i : 2). 2 Oxybenzoic acids, Meta-(i: 3 ). Para- (1:4). C 6 H 3 (OH) 2 CO.OH Protocatechuic acid * (and 5 other isomeric acids). C 6 H 2 (OH) 3 CO . OH Gallic acid (3:4:5). Pyrogallocarboxylic acid (2:3: 4). Phloroglucocarboxylic acid (2:4: 6). 3 possible acids unknown. The figures in brackets represent the relative position of the O H groups round the benzene ring. In the acids the CO. OH group counts as No. I. It will be noticed that a large proportion of the formulae given above represent several isomeric compounds, identical in composition, but frequently very distinct in their properties. The explanation of these differences lies in the different relative positions of the OH and GO. OH groups round the benzene ring. Thus the following diagram represents the probable rela- tive positions of the dihydroxybenzenes. (Ortho-, i : 2). OH OH Catechol. (Meta-, I : 3). OH OH Resorcinol. (Para-, i : 4)- OH Besides these six-carbon phenols, others exist in which one or more of the H-atoms are substituted by methyl or higher hydrocarbon groups, as in the case of ordinary alcohols. Thus the cresols are C 6 H 4 (CH 3 )OH, and yield acids (cresotinic), * The dihydroxybenzenes yield six possible isomeric acids, two from ortho-, three from meta-, and one from para-dioxybenzene, all of which are known. co6 LEATHER INDUSTRIES LABORATORY BOOK. analogous to salicylic, which have been used as antiseptics and chemical bates. The OH groups fmay also be replaced by methoxyl, OCH 3 , ethoxyl, OC 2 H 5 , and other similar compounds. As these higher phenols may enter into the constitution of tannins, it will be seen that the possibilities of complication are very great. Vanillic and isovanillic acids are the meta- and para-methylprotocatechuic acids. Miller,* who has investigated these acids, remarks : " Of the three phenols C 6 H 4 (OH) 2 , catechol alone gives a precipitate with lead acetate, and of the six acids, C 6 H 3 (OH) 2 CO . OH, none yield precipitates with lead acetate except the two which are derived from catechol. Researches by Berthollet and Werner f on the heat evolved by neutralisation with soda throw some light on the constitution of these polyhydric phenols. Resorcinol and quinol yielded about double the heat given by ordinary phenol, showing that "both their OH groups have a similar function to that of phenol. Catechol, on the other hand, in which the groups occupy the ortho-position, gave only the same heat as phenol, showing that only one of its OH groups behaves like that of a phenol, while the second has apparently alcoholic functions. In pyrogallol -and phloroglucoi, two OH groups have phenolic and one an alcoholic function. The authors suggested that pyrogallol has its OH groups in the 1:2:3 position, phloroglucoi in the I : 2 14, leaving the I : 3 : 5 for the third isomeric phenol, not then dis- covered. Later work has shown that phloroglucoi has the 1:3:5 position. Researches by A. Lambert \ have shown that polyhydric alcohols having a primary alcoholic function, as glycerol, laevulose and dextrose, combine with boric acid to form energetic conjugated acids much stronger than boric acid itself. {See p. 75.) This reaction is given by pyrogallol, catechol and alkaline gallates and tannates, but not by quinol or resorcinol. All natural tannins, when heated to i8o-2OO C., yield either catechol or pyrogallol or both, together with secondary products and frequently phloroglucoi, and may be conveniently classified as catechol and pyrogallol derivatives, as the two groups are marked by considerable difference of character. Artificial pro- * Journ. Chem. Soc. xli. 1882, p. 398. t Compt. Rend., c., p. 586 ; Chem. Soc. Abst., 1885, p. 628. \ Op. cit., cviii., p. 1017 ; Chem. Soc. Abst., 1889, p. 864. THE CHEMISTRY OF TANNINS, ETC. 107 ducts, having many of the reactions of tannins, have, however, been obtained trom other phenols, and most phenols and their derived acids give purplish or greenish blacks with ferric salts. Several classifications of the tannins have been suggested. The division most obvious to the tanner is between those tannins which yield ellagic acid as a whitish deposit in the surface of the leather, called " bloom," and those which do not. Stenhouse, many years since, divided tannins into two classes, one of which gives bluish, and the other greenish blacks with ferric salts. In the main, both these classifications correspond to that just suggested, as most tannins which yield a blue-black with iron acetate also give bloom to the leather, and the green-black is generally characteristic of catechol derivatives, and the blue-black of those of pyrogallol. In some cases, however, the difference of tint is due to accidental impurities, or the presence of colouring matters, and is largely dependent on the acid or alkaline condi- tion of the solution, since even gallotannic acid will give a decided green with strongly acid ferric chloride. The Australian mimosas generally give a dull violet with iron salts, though they unquestion- ably belong to the catechol class. Pure gallotannic acid yields no bloom, because it contains no ellagic acid, though it is of course the representative pyrogallol derivative, while ellagic acid, though closely related to gallic acid, yields no pyrogallol on heating, and belongs rather to the colouring matters (p. 140) than the tannins, though it is a decomposition product of ellagitannic acid. A reaction, which appears to be quite universal with catechol tannins, is that when their dilute solutions are treated with excess of bromine water, insoluble bromine derivatives are pro- duced which cause turbidity or precipitates, while with pyrogallol tannins only soluble derivatives are formed and the solution remains clear. This was first pointed out by the present writer in a paper read before the International Congress of Chemists held at Chicago in 1893.* Nierensteinf has shown that catechol tannins give a precipitate with a dilute solution of diazobenzene chloride (an insoluble azo-compound), while pyrogallol tannins do not. Catechol tannins usually form crimson solutions in concentrated sulphuric acid, while pyrogallic tannins give brown or yellow. * Journ. Soc. Ch. Ind., 1894, p. 488. f Collegium, 1906, p. 376. jo8 LEATHER INDUSTRIES LABORATORY BOOK. Trimble points out * that the tannins may be divided into two groups, one of which contains about 52 per cent, of carbon, and the other about 60 per cent. The first of these includes gallotannic acid, and the allied tannins of sumach, chestnut and pomegranate, all of which are known to be pyrogallol tannins ; and the second the catechol tannins, among which must be reckoned cutch, gambier, oak bark, mangrove, canaigre, etc. A few of the analyses cited by Trimble are given below. GROUP I. PYROGALLOL TANNINS. Carbon. Hydrogen. Oxygen. Sumach tannin (Lowe) 52-42 3-56 Chestnut tannin (Nass) .. .. 52-07 3-97 Chestnut bark tannin ^2' 4.2 4-67 Chestnut wood tannin 52-II 4-40 Pomegranate bark tannin (Culley) .. 50-49 3'99 Gallotannic acid, C 14 H 10 O 9 .. .. ! 52-10 3'52 44-38 GROUP II. CATECHOL TANNINS. Carbon. Hydrogen. Oxygen. Oak bark tannin (avg. 9 species) 59 '79 5-08 Oak bark tannin (Etti) 59-29 4'99 White oak bark tannin (Kraemer) 59-65 4-65 Mangrove tannin SQ'76 4-69 Canaigre tannin . ... .?-7 / w ^8- 10 4J. \jy eM 5 33 Rhatany tannin (Ohmeyer) J w iv/ 59-20 4-72 Kino tannin (Bergholz) 59-65 4-87 Catechu tannin (Lowe) 61-93 4-80 Tormentil tannin (Rembold) 60-75 4'65 If tannins are boiled with dilute sulphuric or hydrochloric acids, or allowed to ferment under the influence of pectose and other natural ferments which are usually present in vegetable tanning materials, a different series of decompositions takes place. Many tannins yield glucose, or starch sugar, as one of * 'The Tannins,' ii. p. 131. THE CHEMISTRY OF TANNINS, ETC. 109 their products, or as that of closely associated impurities ; of which more must be said later. In addition it will be found that the catechol tannins invariably yield insoluble reddish-brown bodies, which have been called phlobaphenes or "reds," and which seem often derived from the original tannins by loss of one or more molecules of water, or, in other words, are anhydrides of their respective tannic acids. The pyrogallol tannins, on the other hand, yield gallic acid, or ellagic acid (the deposit forming bloom), either alone or in mixture. It is possible that the so- called " ellagitannic acid " is merely a soluble glucoside of ellagic acid, which exists ready formed in the plant, though it is more probably a gallic acid derivative. Oak bark and valonia give both bloom and insoluble reds, as well as traces of gallic acid. It is quite possible that this is due to their containing a mixture of catechol and pyrogallol tannins. If the red anhydrides, which are produced from the catechol tannins, or the tannins themselves, or their lead-salts, be fused with caustic potash, or in many cases, if simply boiled with con- centrated potash solution, they are broken up still further, and from the fused mass, protocatechuic acid (which bears the same relation to catechol that gallic acid does to pyrogallol) may always be obtained. This is in many cases accompanied by phloroglucol. Cutch, gambier, mimosa, quebracho, and many other materials contain phloroglucide tannins. The tannins which do not yield phloroglucol frequently give acetic acid and other acids of the acetic series, along with protocatechuic acid, and sometimes other products. The presence of phloroglucol may often be detected without the necessity of a potash-fusion by the reaction with deal shaving described on p. 152. If, how- ever, more than one of its hydroxyls are substituted or linked to other groups, it does not react in the compound, though it may be found among the products of decomposition. The classification of tannins is as yet very incomplete, and there are many of which the decomposition products have not been examined. In order to make the information available for further research, the characteristics and mode of recognition of the different products will be given, and as simple a scheme as possible of treatment of the tannin to be examined will be de- scribed ; but the recognition of such products in a state of mixture presents great practical difficulties, and the tanner will usually i io LEATHER INDUSTRIES LABORATORY BOOK. be compelled to confine his attention to simpler, though less conclusive tests, based on the colour-reactions of the unpurified infusions of tanning materials. Such tests will be described later (p. 149 et seq.}. Preparation and Purification of Tannins. The great obstacle to successful investigation of the tannins is the impossibility of preparing them in a state of actual purity, as none of them can be crystallised or volatilised, so that the use of appropriate sol- vents, and fractional precipitation are almost the only means at our disposal. A further difficulty lies in their very unstable character, and in many cases it is impossible to be sure that the supposed purified tannin is really identical with the original substance, and not merely a product of its decomposition. When, as is frequently the case, two or more tannins occur in the same plant, it is often quite impossible completely to separate them. Owing to their considerable differences in character, no general method of purification can be given, but the following processes will be found in many cases to give good results. For the special methods adopted by different investigators, the original memoirs must be consulted, references to many of which will be found in the following pages. The oldest method of separating tannins from other con- stituents is that applied by Pelouze to the preparation of commercial gallotannic acid from gall-nuts. The finely pulverised material is placed in a percolator and exhausted with commercial ether containing water and alcohol. The liquid separates, on standing, into two layers, of which the lower contains most of the tannin in a tolerably pure form, dissolved in water and alcohol with a little ether, while the upper mainly ethereal layer contains the gallic acid. Gall-nuts thus treated yield 35 to 40 per cent, of tannin. If equal parts of ether and 90 per cent, alcohol are used, a larger yield is obtained, but the liquid does not separate into two layers, and it is questionable if the product is so pure. For Chinese galls, washed ether acts better than ether-alcohoh The tannin may be still further purified by dissolving in a mixture of I part water with 2 of ether, when three layers are formed, of which the lowest contains nearly pure tannin. Com- mercial ether generally contains a considerable percentage of alcohol. These methods are applicable to the dried or highly con- THE CHEMISTRY OF TANNINS, ETC. nr centrated extracts of many tanning materials. Many tannins may be separated from their strong aqueous solution in a state of considerable purity by first agitating with ether to remove gallic acid, and then saturating the aqueous solution with com- mon salt, and shaking with acetic ether, which takes up the tannin. The acetic ether is distilled off nearly to dryness, and the last traces expelled by drying on the water-bath with small repeated additions of ether, which expels the acetic ether and leaves the tannin in a porous mass. Another method is to extract with alcohol, evaporate to a small bulk at as low a tem- perature as possible, and treat at once with a considerable quantity of cold water to precipitate phlobaphenes which are filtered off. The infusion is then precipitated with successive small quantities of lead acetate ; the first and last portions of the precipitate are rejected as contaminated with colouring matters and other impurities, while the middle fraction, after rapid washing, is suspended in water and decomposed with sul- phuretted hydrogen. The solution is filtered, and the excess of SH 2 driven off by heat, and the aqueous portion is either evapor- ated at a low temperature in a partial vacuum to a thin syrup, and the drying completed over sulphuric acid in vacuo ; or, preferably, shaken out with acetic ether. Dr. Nierenstein and the author have observed that in many cases treatment with SH 2 appears to decompose or alter the tannin, and prefer to liberate it with a quantity of dilute sulphuric acid insufficient to decom- pose the whole of the lead salt. Compare also Schiff. Prof. Trimble * has made use of acetone as a solvent for the extraction of tannins with great advantage. The following is the general method he adopted in his researches on the oak tannins. The powdered bark was covered with acetone in a percolator and allowed to macerate for 48 hours. It was then percolated somewhat rapidly till 500 c.c. of percolate per kilo- gram of bark was obtained, which was mostly sufficient for exhaustion ; the acetone retained in the bark was displaced by water, and the acetone recovered from the percolate by distilla- tion on the water-bath, the last portions being removed under reduced pressure. The dry residue was dissolved in a small quantity of warm water, or of alcohol sp. gr. 0*975, filtered and then diluted with water so long as anhydrides and colouring * ' The Tannins,' ii. p. 78. Philadelphia, 1894. 112 LEATHER INDUSTRIES LABORATORY BOOK. matter were precipitated. The solution was again filtered, and shaken out with successive portions of acetic ether. The acetic ether solution was distilled off under reduced pressure, and the tannin further purified by solution in water and shaking out with acetic ether. The acetic ether was finally removed by solution in ether-alcohol, which was distilled off under reduced pressure, and the dry tannin digested with absolute ether to remove traces of resin and other impurities. The resultant tannins were pale coloured and perfectly soluble in water. In the case of common oak bark (Q. robur) it was necessary to add salt before shaking with acetic ether, and in some cases the lead method of purification was employed. It is not always necessary to separate the tannin itself in a state of purity, since much light can be thrown on its constitution by investigation of its metallic salts or organic products. The lead salts have been used both for ultimate analysis, and for fusion with potash, and some experiments have been made in the author's laboratory in investigating the insoluble bromine- derivatives of the catechol tannins, which are obtained in a state of tolerable purity by precipitation of the dilute solutions with bromine-water. Some insoluble diazo-compounds have also been prepared, but without much result. GENERAL METHODS OF EXAMINATION OF TANNINS. Decomposition by Heat. The ordinary method is to distil the tannin or dried extract in a small retort, and examine the distillate for catechol and pyrogallol. Unless the heat be very carefully regulated, much loss is caused by the destruction of the catechol and pyrogallol with formation of metagallic acid, etc., and their detection is greatly complicated by the presence of secondary products. This difficulty is somewhat lessened by passing a stream of carbon dioxide through the retort, which carries the products quickly out of the heated portion. A better method is to heat the tannin in glycerin.* About I grm. of the sample is heated with 5 c.c. of pure glycerin to 1 60 C, and the temperature slowly raised to 2OO-2io C, at which it is maintained for twenty or thirty minutes. After * Thorpe, Chem. Soc. Abst., 1881, p. 663. THE CHEMISTRY OF TANNINS, ETC. 113 cooling about 20 c.c. of water is added, and the liquid is shaken with an equal volume of ether, without previous filtration. As the ether when shaken with the glycerin is apt to form a trouble- some emulsion, Trimble* recommends shaking the glycerin solu- tion with 20 c.c. of ether without previous dilution with water ; or heating the dry tannin in 5 grm. of paraffin, which is removed with petroleum ether and the residue taken up with water and treated as above. The ethereal layer, which may contain pyro- gallol and catechol, and possibly phloroglucol or other phenols (cp. p. 109), is separated from the aqueous portion, evaporated to dryness, and dissolved in 50 c.c. of water. The filtered solution is divided into several portions and tested by the table on p. 1 1 8. By these means it is easy to distinguish between catechol and pyrogallol, and either may be detected in presence of a small portion of the other ; but if in nearly equal quantities their recognition is difficult. Catechol may be derived from catechin, etc., and pyrogallol from gallic acid, and it is therefore necessary in some cases to remove these bodies from the tannin before treatment by shaking its aqueous solution with ether. Phloroglucol is produced from all the colouring matters of the flavone group (p. 141) on fusion with potash (except fisetin which yields resorcin), and in most cases together with protocatechuic acid ; and these colouring matters are frequently associated with the tannins, usually as glucosides. As a general rule, however, catechins and catechol derivatives are only present in any quan- tity with catechol tannins, and the same is true of gallic acid with regard to pyrogallol. Products of the Decomposition of Tannins by Heat. Pyrogallol, C 6 H 3 (OH) 3 (formerly called pyrogallic acid, but really a phenol), has a bitter, but not sour taste, and feebly reddens litmus ; the addition of the smallest trace of alkali gives it an alkaline reaction. It is somewhat poisonous, but less so than some other phenols. It is soluble in less than three parts of cold water, and still more freely in hot. It is also soluble in alcohol, ether, acetone, ethyl acetate and glycerin, but not inabsolute chloroform or petroleum spirit. It fuses at 132 C. out of toluol, but ordinarily at II5C., and sublimes at about 210 C. * ' The Tannins,' i. p. 27. I ii4 LEATHER INDUSTRIES LABORATORY BOOK. With pure ferrous sulphate it gives a white precipitate, which redissolves to a fine blue liquid in presence of the least trace of ferric salt. Mineral acids change this to red, and the blue tint is restored by cautious neutralisation with ammonia, and is not destroyed, but sometimes rendered greenish, by excess of acetic and other organic acids. Any excess of ammonia produces an amethyst-red, and acetic acid restores the blue. Ferric chloride and other acid ferric salts produce a red coloration at once. Pyrogallol solution is turned brown by traces of nitrous acid. With lime-water it produces a beautiful but evanescent purple, rapidly turning brown. In presence of alkalies it absorbs oxygen from the air with great avidity, turning orange, brown, and black ; and is therefore used as an oxygen-absorbent in gas analysis. Pyrogallol does not precipitate gelatin. Its solution rapidly re- duces permanganate, Fehling's solution, and salts of gold, silver,* mercury and platinum. It precipitates copper and lead acetates, and with ammoniacal cupric sulphate it gives an intense purple brown coloration, and, in common with gallic and gallotannic acids, is coloured purple-red by iodine in either aqueous or alcoholic solutions. Gum arabic, saliva, and various other organic matters cause solutions of pyrogallol exposed to the air to absorb oxygen, forming purpurogallin, which separates in small yellow or orange-red needles. If 0*2 per cent, of pyrogallol be added to a I per cent, solution of gum arabic, it becomes yellow in a few hours, and purpurogallin separates in hair-like crystals, which continue to increase for some months. If these crystals are freed from pyrogallol by washing with water, and a trace of alkali is added, they dissolve in absence of oxygen to an orange- red solution, which rapidly oxidises to an intense blue colour, and finally to a brownish yellow. Purpurogallin is also formed by oxidation with silver nitrate, potash permanganate and ferricyanide, nitrous acid, and many other reagents. Potas- sium nitrite, and probably nitrous acid is also capable of pro- ducing purpurogallin from gallic acid, and it is very possible that the colour-reaction of pyrogallol-tannins described on p. 162 may be due to its formation. A. G. Perkin and Stevens,! and A. G. Perkin and F. M. Perking have lately investigated and described * Hence its use as a " developer " in photography. t Journ. Chem. Soc., 1903, Ixxxiii. p. 194. \ Op. cit., 1904, Ixxxv. p. 243. THE CHEMISTRY OF TANNINS, ETC. 115 this substance very fully, and have devised a method of preparing it in considerable quantities by electrolytic oxidation of pyro- gallol. It appears to be a naphthalene derivative. Pyrogallol forms compounds with aldehydes; with formaldehyde, a body which reacts like a tannin and precipitates gelatin is produced. If pyrogallol be heated with hydrochloric acid and aldehyde, chloral or acetone, a red substance is produced. The less volatile portions of crude beech-tar creasote contain ethers of pyrogallol, methyl-pyrogallol and propyl-pyrogallol, and of the corresponding catechol derivatives, from which these bodies may be obtained by the action of hydrochloric acid under pressure. The methyl and propyl derivatives differ from the original phenols in having one or more atoms of hydrogen replaced by the groups CH 3 or C 3 H 7 respectively ; and it is very possible that some tannins are derived from them. Tribrom- and tri- chlor-pyrogallols are formed by the action of bromine or chlorine on pyrogallol. If pyrogallol be heated rapidly to 250 C. it parts with the elements of water, and is converted into metagallic acid, C 6 H 4 O 2 , a black amorphous body, insoluble in water, soluble in alkalies. When pyrogallol is made in the ordinary way, by heating gallic or tannic acid to 210 C., much metagallic acid is formed, even if the process be conducted in a stream of carbonic acid, and the yield of pyrogallol usually amounts to only about 5 per cent, of the gallic acid employed (see p. 1 20). On boiling pyrogallol with hydric potassic carbonate, pyro- gallo-carboxylic acid is produced, which only differs from gallic acid in the position of its CO. OH group (p. 119), and like it is resolved into pyrogallol and CO 2 on heating. Pkloroglucol (Phloroglucin), 1:3:5 C 6 H 6 O 3 , is a trihydric phenol, isomeric with pyrogallol. It crystallises with two mole- cules of water, which it loses at 1 00 C. It melts at about 220 C., sublimes without' odour, and solidifies again on cooling. It is soluble in water, alcohol, and ether, and by agitation with the latter it may be removed from its aqueous solution. It is not precipitated by any metallic salt but basic lead acetate. When pure it gives no reaction with ferric salts. If bromine be added to its concentrated solution in water, it absorbs three atoms, forming tribromophloroglucol, C 6 Br 3 (OH) 3 , which separates in crystalline needles, with evolution of heat and a very irritating I 2 u6 LEATHER INDUSTRIES LABORATORY BOOK. odour. In dilute solution it gives a bulky white precipitate with bromine water. If deal shaving be moistened with solution of phloroglucol, and then with strong hydrochloric acid, it soon takes a deep violet colour, from the formation of phloroglucol- vanillein with the trace of vanillin contained in all coniferous woods. Neither catechol nor pyrogallol gives this reaction when free from phloroglucol ; though Etti has stated that pyrogallol forms a somewhat similar bluish compound when mixed with vanillin and hydrochloric acid. Resorcinol gives a similar reaction but of a much bluer shade, but this may be due to the presence of phloroglucol as an impurity. Nierenstein, however, obtained the reaction from resorcinol, purified with extreme care.* This reaction may be used to detect phloroglucide tannins without the troublesome fusion with potash, and is strongly given by gambier and cutch, and more faintly by many other tanning materials (see pp. 157-167, et seq). Many colouring matters, however, such as quercetin (p. 142), and possibly some tannins, yield phloroglucol on fusion with potash, but contain the phloro- glucol nucleus in such a form that it does not react with vanillin. For the reasons given, the results must be accepted with some reserve. If to a dilute solution of phloroglucol a solution of aniline or toluidine nitrate be added, and then a trace of potassic nitrite, the liquid gradually becomes yellow or orange, then turbid, and finally deposits a cinnabar-red precipitate. This reaction is given by many tanning materials, but is somewhat unsatisfactory, as gall tannin, pyrogallol and other substances give similar but browner precipitates. Phloroglucol is more commonly obtained by fusion with potash than by direct heating, but as it may occur in the pro- ducts of dry distillation, it is described here, together with the other polyhydric phenols. Catechol (Pyrocatechin), I : 2 C 6 H 4 (OH) 2 , is not only produced by the decomposition of certain tannins by heat (see p. 1 1 3), but by the dry distillation of protocatechuic and 1:3:4 dioxybenzoic acids, and of catechin and some colouring matters which contain the catechol groups, and which frequently accompany the tannins. It is also formed .together with pyrogallol and its homologues {see above) by the dry distillation of wood, wood-tar creasote consisting largely of ethers of catechol and '-its homologues * Collegium, 1906, p. 14. THE CHEMISTRY OF TANNINS, ETC. 117 methyl- and propyl-catechol, etc., and is hence found in crude pyroligneous acid, and causes the greenish-black colour of " iron- liquor." It has also been produced by long continued heating of starch, cellulose and other carbohydrates with water under pres- sure, and is found ready formed in Virginia-creeper (Ampelopsis hederaced), and probably in other plants. It has also been formed synthetically by fusion of the I : 2 chlor- and brom- phenols, and by fusion of I : 2 benzenedisulphuric and phenol- sulphuric acids with potash. Catechol melts at 104 C., and sublimes at about the same temperature, condensing in brilliant laminae like benzoic acid. Its actual boiling point is 245 C. (Richter). It is readily soluble in water, alcohol and ether, and is extracted from its aqueous solution when shaken with the latter. Its aqueous solution precipitates lead acetate, by which it is distinguished from resorcinol and quinol. It does not precipitate gelatin or alkaloids. Its slightly alkaline solution exposed to air becomes first green, then blue and black. With excess of lime-water or caustic soda solution it becomes reddish, but remains clear for some time. It does not colour ferrous salts, but gives a dark green with ferric (avoiding excess) and after some time a black precipitate. The green is changed to a fine violet-red by alkalies and hydric sodic carbonate, and restored by acids. It has been stated that it gives a violet coloration to fir wood moistened with hydrochloric acid, but this is due to contamination with phloroglucol, and is not given by pure catechol. Catechol gives a red coloration with citric acid, and the solution, after standing, ceases to react with iron. Like pyrogallol, it reduces silver salts in the cold, and has been used as a developer in photography. Alkaline copper solutions (Fehling's solution) are only reduced by the aid of heat. Resorcinol (Resorcin), I : 3 C 6 H 4 (OH) 2 , though not common as a constituent of tannins, is stated by Dr. Nierenstein to be present in quebracho tannin, together with quinol (p. 104), and it is quite possible, therefore, that both may be present in other tannins. It melts at 118 C., volatilises at a somewhat higher temperature, and boils at 276 C. (Richter). It is easily soluble in water, alcohol, and ether, and may be removed from its aqueous solution by shaking with the latter. It is known to be a product of the dry distillation of brazil-wood extract. It is obtained synthetically Ii8 LEATHER INDUSTRIES LABORATORY BOOK. from the meta- (1:3) benzene and phenol compounds in the same way as catechol from the ortho- (1:2) derivatives. Commercially it is prepared from meta-benzenedisulphuric acid, and is important in the preparation of many artificial colouring matters. Lacmoid, used as an indicator, is produced by heating resorcinol with sodium nitrite ; resazurin by the action of nitrous acid on its ethereal solution. It is possible that one or other of these reactions might be used as a means of identification. Re- sazurin crystallises in lustrous green prisms and dissolves in alkalies to a violet solution with brown fluorescence. Resorcinol is not precipitated by lead acetate. Quinol^Hydroquinone), I 14 C 6 H 4 (OH) 2 , melting point 169 C., soluble in water, alcohol, and ether. It is precipitated by lead acetate in presence of ammonia only. It reduces silver in alka- line solution, and is used as a photographic developer. It has two crystalline forms, subliming in monoclinic scales, and crystal- lising from solutions in hexagonal prisms. DISTINCTION BETWEEN PYROGALLOL, CATECHOL AND PHLOROGLUCOL. i PER CENT. SOLUTIONS. Reagent. Pyrogallol. Ferric alum .. .. ; Blue-black, turning green and brown. Lime-water .. .. Violet coloration, rapidly turning brown. No colour. Catechol. Phloroglucol. Dark green. No reaction. Deal shaving and HC1 Bromine- water No marked reaction. No marked reaction. No colour. Violet-red coloration No precipitate. No precipitate. Bulky white precipi- tate. Decomposition of Tannins by Dilute Acids. It has been stated that tannins, when heated with dilute sulphuric or hydro- chloric acids, are decomposed, yielding either gallic or ellagic acids or red anhydrides, and frequently glucose. To determine whether glucose is produced, the tannin must first be carefully purified from glucose, gums or other bodies likely to interfere, by the methods mentioned on p. 1 10. Either the tannin itself or its washed lead-salt may be used, and must be heated to 100 C. for an hour or more in a sealed tube or tightly closed bottle 'I HE CHEMISTRY OF TANNINS, ETC. 119 or boiled in a flask with reflux condenser, with dilute hydrochloric acid (2 per cent. HC1). Complete decomposition is, however, somewhat difficult. After cooling, the mixture must be allowed to stand for some time to separate any sparingly soluble products, which must be filtered off. The filtrate must be shaken with ether to remove gallic acid, etc. (p. 120), the aqueous solution must be boiled, neutralised with sodic hydrate, precipitated with basic lead acetate to remove any traces of tannin or colouring matters, the liquid again filtered, and excess of lead removed with dilute sulphuric acid, excess of acid and all elevation of temperature being carefully avoided. The mixture is neutralised with sodic hydrate, and the clear filtrate heated to boiling, with Fehling's copper solution, when a red precipitate of cuprous oxide proves the formation of glucose. (Cp. also p. 270.) The precipitate separating from the hydrochloric acid solu- tion on cooling may consist of lead chloride, if the lead salt has been used, of ellagic acid, or of insoluble red anhydrides or phlo- baphenes of the tannin. The lead chloride may be removed by washing with boiling water ; and the phlobaphenes by cold alcohol, which leaves them on evaporation, or on dilution with water. If the remaining precipitate has a pale yellow or fawn colour it probably consists of ellagic acid (see p. 121), which may be crystallised from hot alcohol or pyridine. It is soluble in ammonia and reprecipitated by acids, and dissolves freely in strong nitric acid, forming a crimson or deep yellow liquid. The ethereal solution which contains the gallic acid, if any has been formed, must be evaporated to dryness, and the residue taken up with cold water and filtered. Addition of a few drops of solution of potassium cyanide will produce a fine red colora- tion if gallic acid be present, which rapidly fades, but is restored by shaking in presence of air. The sodium arsenate test is also a delicate one. For this and other reactions of gallic acid see p. 1 20. Products of the Decomposition of Tannins by Acids. Gallic Acid (Trihydroxybenzoic Acid), I 13:4:5 C 6 H 2 (HO) 3 CO. OH, exists ready formed in some plants, and is a product of the hydrolysis of gallotannic acid under the influence of the un- organised ferment pectase, or of its decomposition by boiling with acids or alkalies. It crystallises in white or yellowish-white 120 LEATHER INDUSTRIES LABORATORY BOOK. needles, containing I mol. (9*5 per cent.) of water, which it loses at 100 C. It is soluble in 100 parts of cold, or 3 of boiling water,, in alcohol or glycerin, and slightly so in ether, by agitation with which it may, however, be removed from its aqueous solution. Gallic acid fuses at a temperature of about 232 C.,* but at about 2ioC. begins to lose carbon dioxide, and yields a crystalline sublimate of pyrogallol (see p. 113). If the heat be raised sud- denly to 250 C., a considerable quantity of black shining metagallic acid is formed. When gallic acid is heated with concentrated sulphuric acid to 140 C. dark red rufigallic acid is formed, and is precipitated on dilution with water. Rufigallic acid is a hexaoxyanthraquinone related to alizarine. It forms a blue solution with alkalies, and dyes brown on chrome mordants. A. J. Perkin f has employed this reaction as a means of identifying gallic acid, and it is strongly recommenced for this purpose by Nierenstein. Aqueous solution of gallic acid gives the following reac- tions : Solution of ferric chloride gives a deep blue coloration which is destroyed by boiling. Ferrous sulphate, if free from ferric salts, gives no reaction in dilute solutions, but a white precipitate in strong ones. The mixture rapidly darkens by oxidation. In alkaline solution gallic acid absorbs oxygen from the air and darkens from the formation of galloflavin, and after- wards of tannomelanic acid. Lime-water produces a white precipitate which rapidly becomes blue from oxidation (distinc- tion from pyrogallol). The same reaction is produced by baryta-water, or by the chlorides of barium or calcium on addition of ammonia. Gallic acid is distinguished from gallotannic acid by the following : It does not precipitate gelatin, except in the presence of gum. It does not precipitate tartar emetic in presence of ammonic chloride, though both tannin and gallic acid are precipitated by tartar emetic alone. It precipitates lead acetate, but not lead nitrate, while tannin precipitates both.. A dilute solution of potassium cyanide gives a red coloration which dis- appears on standing, but is restored by shaking with air. If to even a very dilute solution of gallic acid, sodic arsenate, or some other faintly alkaline salts be added, the mixture absorbs oxygen and becomes a deep green. Aqueous solution of picric acid to * Etti, Chem. Soc. Abst., 1879, p. 160. t Proc. Chem. Soc., 1896, p. 157. THE CHEMISTRY OF TANNINS, ETC. 121 which excess of ammonia has previously been added gives a red coloration, changing to green. Tannic and pyrogallic acids produce no reaction with cyanide, and with ammonium picrate a reddish coloration only. Gallic acid reduces silver nitrate and gold chloride rapidly when hot, but not Fehling's solution, and decolorises acidified potassic permanganate. If tannin and other oxidisable bodies be removed from its solution, it may be estimated quantitatively by titration with permanganate in presence of indigo (see p. 224). Tannin may be separated from it by gelatin or hide power (see pp. 208, 224) but much gallic acid is also absorbed. Gallotannic and quercitannic acids may also be removed by precipitation with ammoniacal solution of cupric sulphate, or by cupric acetate, in presence of excess of ammonic carbonate (see also p. 171). Many other tannins, how- ever, give precipitates with cupric salts which are soluble in ammonia and ammonic carbonate. Gallotannic acid may also be separated from gallic acid by- lead acetate strongly acidified with acetic acid, by which tannic acid is precipitated, while lead gallate is dissolved ; and by many organic bases, such as cinchonidine, brucine, morphia, and many of the basic colouring matters ; but none of these separations are very satisfactory. Strychnine, however, precipitates gallo- tannic acid quantitatively (see p. 169). Ellagic Acid C U H 6 O 8 , is a decomposition product of ellagi- tannic acid (p. 109), and may be obtained from the tanning materials which contain this tannin. It is deposited from, aqueous infusions of these as a yellowish white solid, known as "bloom." It may be prepared by pouring a hot concentrated alcoholic extract of divi-divi into water, when it separates, and may be filtered off and re-crystallised. It may also be obtained by boiling divi-divi, myrobalans^ pomegranate rind, etc., with dilute hydrochloric acid, and ex- tracting the washed and dried insoluble residue with hot alcohol,, from which it crystallises on cooling. Another convenient source of ellagic acid is the " Alizarine yellow in paste " of Messrs. Meister, Lucius and Briining, which after drying may be readily purified by crystallisation from pyridine. Lowe * was the first to prepare ellagic acid synthetically by oxidising gallic acid with arsenic acid, and this was subsequently accomplished by Ernst * Zeitschrift fur Chemie, 1868, iv. p. 603. 122 LEATHER INDUSTRIES LABORATORY BOOK. and Zwenger* by heating the acid ethyl gallate with sodium carbonate solution in the presence of air, and by Griessmeyer f by heating gallic acid with water and iodine. The best method, however, for its synthetical preparation is one due to A. G. Perking who oxidises a 10 per cent, solution of gallic acid in acetic acid with potassium persulphate and a little sulphuric acid. When the reaction is complete, the mixture is poured into water, and the resulting sandy precipitate is collected, washed, dried, and purified by crystallising from boiling pyridine, in which it dissolves somewhat readily. The prismatic needles eventually obtained are a pyridine salt of ellagic acid, most of the pyridine being removed by washing with alcohol and the final traces by drying a,t 1 60 C. Thus obtained, it forms a crystalline powder, melting above 360 C., soluble in dilute caustic alkaline solutions with a pale yellow colour. It is itself very pale yellow, and is almost insoluble even in boiling water, and only slightly so in cold alcohol and ether, though by agitation with the latter, small quantities may be completely removed from aqueous solution. In hot alcohol, it dissolves with a yellow colour and crystallises on cooling. Solid ellagic acid gives with ferric chloride at first a greenish, and then a black coloration. With nitric acid con- taining nitrous acid, and subsequent dilution, it gives a deep crimson coloration (Griessmeyer's reaction), though that obtained from some natural sources is rather orange. A constitutional formula was first suggested by Schiff, afterwards by Graebe,|| and it was finally proved by A. G. Perkin If to be a double lactone of a hexa-hydroxydiphenyl di- carboxylic acid. The formula is as follows : . OH /' C \ < > <> \ /OH XX) . O/ It will be noticed that the formula is symmetrical, and the formation by the oxidation of gallic acid is readily understood. * Annalen, 1871, clix. p. 32. t Annafcn, 1871, clx. p. 55. I Journ. Chem. Soc., 1905, Ixxxvii. p. 1412. Ber., 1879, xii. 1534. ., II Ber., 1903, xxxvi. 212. f Journ. Chem. Soc., 1905, Ixxxvii. 1412. THE CHEMISTRY OF TANNINS, ETC. 123 It forms a tetra-acetyl compound (m.p. 343-346 C.) and a tetra- benzoyl compound (m.p. 332-333 C.), and on distillation with zinc dust yields fluorene On boiling with concentrated caustic potash, one lactone group is opened, and the carboxyl group eliminated, yielding on neutra- lisation and dilution pentahydroxydiphenylmethylolid. OH OH OH/ _/ OH co . o This compound melts above 300 C., and yields penta-acetyl and penta-benzoyl derivatives, and fluorene on distillation with zinc dust. On fusing ellagic acid with alkalies, the reaction is carried a stage further, and the other lactone group similarly attacked, the final product being hexa-hydroxydiphenyl. OH OH OH OH When heated at 230 C. with 100 per cent, sulphuric acid, ellagic acid is oxidised to form a new compound* C 14 H 6 Oi , which crystallises from pyridine and is soluble in strong alkalies with a greenish yellow tint changing to blue on dilution. This compound gives a hexa-acetyl derivative. Flavellagic acid also yields this compound. Ellagic acid is a feeble colouring matter, dyeing yellows with tin and alumina mordants, and various duller shades with chromium and iron mordants. Catellagic Acid C U H 6 O 6 was obtained by Schifff on heating protocatechuic acid with arsenic acid, and recently by Perkin,J by the oxidation of protocatechuic acid in sulphuric acid solution with potassium persulphate, and was purified by crystallisation * Proc. Chera. Soc., xxii. 308, p. 114. t Ber., 1882, xv. p. 2590. \ Journ. Chem. Soc., 1905, Ixxxvii. p. 1418. 124 LEATHER INDUSTRIES LABORATORY BOOK. in pyridine as with ellagic acid, which it very strongly resembles in its properties. It is a pale yellow crystalline powder, melting above 360 C, subliming much more readily at higher temperatures than ellagic acid. It dissolves in concentrated nitric acid, forming a magenta coloured liquid, which on standing changes to a dull brown, and from which on dilution with water, a yellow crystalline precipitate, probably a nitro-compound, is deposited. Caustic alkalies dissolve it with a pale yellow colour, which is discharged on warming. Its constitution is as follows : ,0 . CO OH OH A diacetyl derivative is easily prepared (m.p. 322-324 C.), and fluorene obtained on distillation with zinc dust. It does not dye mordanted calico. Catellagic acid is also formed by the persulphate oxidation of para-hydroxybenzoic acid in sulphuric acid solution, another hydroxyl group being first introduced forming protocatechuic acid, and the diphenyl condensation afterwards taking place. Another compound is also formed in this reaction, by the diphenyl condensation of para-hydroxybenzoic acid and proto- catechuic acid. It has the constitution : OH ___ ^ . COOH O This compound does not melt below 360 C. and does not sublime. It is very readily soluble in pyridine, yields a diacetyl compound (m.p. 267-268 C.), and also fluorene on distillation with zinc dust. It dissolves with bicarbonates of sodium and potassium, and in hot potassium acetate, and forms salts. Nitric acid gradually attacks it, and yields an orange red liquid. It does not dye mordanted calico. Metellagic acid Ci 4 H 6 O 5 was prepared by. Perkin* by the * Journ. Chem. Soc., 1905, Ixxxvii. p. 1426. THE CHEMISTRY OF TANNINS, ETC. 125 persulphate oxidation of meta-hydroxybenzoic acid in sulphuric acid solution. It is isolated from the mixture of products so obtained by the repeated re-crystallisation of its acetyl derivative and subsequent hydrolysis. Its constitution is as follows : ,0 .CO OH It resembles catellagic acid in that at high temperatures, it sublimes readily in the form of needles. It is soluble in pyridine and deposits long prismatic needles therefrom. Sulphuric acid dissolves it with a pale yellow colour, and nitric acid on gentle warming with an orange yellow tint, and on diluting this solution with water, a yellow precipitate, probably a nitro compound, is deposited. With caustic alkali solutions, it behaves like catellagic acid, and fluorene is obtained on distillation with zinc dust. It forms a crystalline mono-acetyl derivative, melting at 273-276 C., and does not dye mordanted calico. Flavellagic Acid (hydroxyellagic acid) Ci 4 H 6 O 9 was obtained by Perkin * by the persulphate oxidation of gallic acid in the presence of dilute sulphuric acid. If 96 per cent, sulphuric acid be used, ellagic acid is produced also. It is purified by crystal- lisation from pyridine, in which it is less soluble than ellagic acid Its constitution is shown thus : O . CO. OH / OH OH \ / OH NCQ . o/ Small, pale yellow, prismatic needles are obtained from pyridine, which latter is removed, as with ellagic acid. The pure compound does not melt below 360 C., and is insoluble in the usual solvents. It is more yellow than ellagic acid, and its solu- tion in dilute alkalies has a well-marked green tint. It gives the Griessmeyer reaction with nitric acid, and fluorene by dis- tillation with zinc dust. It dyes mordanted woollen cloth in shades resembling, but stronger than, those produced by ellagic * Journ. Chem. Soc., 1906, Ixxxix. p. 251. 126 LEATHER INDUSTRIES LABORATORY BOOK. acid. It yields a sparingly soluble penta-acetyl compound (m.p. 3I7-3I9 C, and a penta-benzoyl compound (m.p. 287-289 C.) On boiling with 50 per cent, caustic potash, a reaction similar to that with ellagic acid occurs, one lactone group being opened, the carboxyl eliminated, and hexa-hydroxydiphenylmethylolid obtained. OH OH OH/ V /- "V ._/ OH \ /OH \CO . O' This compound melts above 300 C, dissolves readily in pyridine, from which it cannot easily be crystallised. Caustic alkalies dissolve it with a dull orange-yellow tint, which on dilu- tion and exposure to air, becomes bluish violet. Alcoholic lead acetate gives an orange-yellow precipitate which rapidly passes into olive green, and alcoholic ferric chloride a bluish green colour. Its solution in sulphuric acid is orange-yellow. Nitric acid dissolves it with an orange colour which becomes redder on dilution, but is devoid of the blood-red tint of the Griessmeyer reaction. It dyes mordanted calico and woollen cloth. The hexa-acetyl compound melts at 232-234 C., and the hexa-benzoyl compound at 261-263 C. A somewhat full account of the ellagic group has been given (and perhaps out of proportion to its direct interest to the tanner), because, owing to the researches of Perkin, its constitution is now fully cleared up, and will probably serve as a starting point for further work elucidating the structure of the tannins themselves. The phlobaphenes or reds are chemically the anhydrides of certain tannic acids from which they are derived, or in other words they are formed from the tannins by the loss of one or more molecules of water. It is in this way that they are pro- duced by the action of acids ; and similarly they are often formed when alcoholic or highly concentrated aqueous extracts are poured into cold water, under which circumstances a part of the tannin seems unable to take up water again, and separates as a red precipitate. It is also possible that in some cases oxidation as well as dehydration takes place. Freezing also seems to have a dehydrating effect. Most solutions of catechol tannins are decomposed by long-continued boiling or heating THE CHEMISTRY OF TANNINS, ETC. 127 under pressure, with formation of reds ; which also exist ready formed in most tanning materials capable of producing them. They are soluble in alcohol, by which they may be extracted from tanning materials or dried residues containing them. They are also dissolved by dilute alkalies and alkaline carbon- ates and sulphites, and by borax, which is said to be used in the preparation of some extracts, and was tried by Sadlon as a means of rendering phlobaphenes available for tanning, but without much practical success. More successful results have been, obtained by Lepetit Dollfus and Gansser's* process of heating with bisulphites under pressure, which give up base to the phlobaphenes, while the sulphurous acid prevents oxidation, and apparently to some extent combines with the tannins. Many phlobaphenes are scarcely soluble in water even at a boiling temperature, though they become more so in the presence of sugar, tannic acid, and some other substances. Their solubility in water depends on their degree of hydration, many tannins giving a series of anhydrides, of which those containing only one molecule of water less than the original tannin are quite soluble in water, while the higher members of the series become less and less soluble as they lose water. Those which are soluble form the principal colouring substances of tanning materials, and generally are practically tannins, precipitating gelatin and combining with hide to form leather. Hemlock bark yields a series of such bodies, of which the lower members are deep red soluble tannins, while the higher form the red sediment, so well known to extract- tanners. Thus it is chemically impossible to decolorise hemlock extract without at the same time greatly lessening its tanning power, though by careful manufacturing and concentration at low temperature, the proportion of the higher anhydrides formed may be kept at a minimum. In many cases it is known, as in gambier, and in others it is probable that the tannin itself is merely the first anhydride of the series, and derived from a catechin which itself is a white crystalline body destitute of tanning properties (see p. 137). The more soluble " reds " are identical with the so-called " difficultly soluble tannins." Decomposition of the Phlobaphenes by Fusion with Caustic Alkalies. It has been mentioned (p. 1 16) that the reds of different tannins yielded, in addition to protocatechuic acid, either phloro- * E. P. No. 8582, 1896. Journ. Chem. Soc. Abst, 1896, p. 436. 128 LEATHER INDUSTRIES LABORATORY BOOK. glucol, or acetic acid, or some other member of the fatty acid series, and some tannins, as those of alder and hop, give more than one of these, but it is possible that this may arise from the presence of two distinct tannins in these materials. It is stated that all those tannins which yield acetic acid on fusion with potash, also yield considerable quantities of glucose to dilute acids, while the phloroglucide tannins do not do so. It is stated by Earth and Schroder * that a small quantity of phloroglucol is produced by the fusion of gallic acid, resorcinol and common phenol with caustic soda. It is therefore possible that in some cases where phloroglucol is detected, it may have been formed by the action of the alkali, and not have been originally a con- stituent of the tannin. Note also its formation from the flavones j&-i>jf C. ; (c) m.p. 163 C. He further describes several other catechins of different melting points derived from acacia catechus. Perkin and Yoshitake (loc. cit.) working on cube gambier, which was extracted with ethyl acetate and the product purified by crystallisation from water after frac- tional precipitation with lead, obtained a catechin agreeing in melting point with Gautier's (b) y which appeared to be a chemical individual, but could not detect the (a) catechin. Dried at 100 C. the air-dried crystallised subtance lost about 20 per cent, of water of crystallisation, and on subsequently heating to 1 60 C. no further loss or change beyond slight browning occurred. Analysis gave numbers closely approximating for the crystalline substance to those required for C 15 H 14 O 6 . 4H 2 O, and this formula was further confirmed by analyses of azobenzene benzoyl and acetyl derivatives. Molecular weight determinations by the freezing point method in naphthalene gave results confirmatory of the formula just quoted, and acetyiation and benzoylation indicate five replaceable OH groups. By treatment with lead acetate in solution slightly acidified with acetic acid, a crystalline lead salt, corresponding to the formula Ci 5 H 14 O 6 . Pb(C 2 H 3 O2)2 was obtained. The catechin was sparingly soluble in cold water, freely in alcohol, and if air-dried, soluble in ether and ethyl acetate, but not if dried at iooC. It gave the phloroglucol reaction with deal-shaving and hydrochloric acid, and on fusion with potash, phloroglucol and protocatechuic acid, and a volatile acid, probably acetic. With ferric chloride a green, and in pre- sence of sodium acetate a deep violet coloration was produced. From the mother liquors formed in purifying this (b) cate- chin a second catechin was isolated, with a melting point of 235-237C., and crystallising without water of crystallisation. In its other properties, and decomposition-products, as well as in the results of ultimate analysis, it was identical with catechin (b). From acacia catechu (" kath ") the same authors obtained a * Bull. Soc. Chim., 1878 (ii.), xxx. p. 567. THE CHEMISTRY OF TANNINS, ETC. 139 catechin, which sinters at about 140 C, resolidifies, and finally melts at 2O4-2O5 C. without decomposition, and which was possibly identical with Gautier's (a) catechin. Except in melting point, it appears to be identical in properties and composition with catechin (fr). The authors state that all these catechins are probably represented by the formula C 15 H 14 O 6 , but are different substances, since they differ not only in melting point but in water of crystallisation, and similar differences occur in their derivatives. It is possible that the volatile acid obtained on potash-fusion is in some cases formic and not acetic. The work of Perkin and Yoshitake has on many points been confirmed by that of Kostanecki and Tambor,* and Karnowski and Tambor.t Ordi- nary gambier catechin yields a series of reds, probably anhydrides, on boiling with dilute sulphuric acid. Heated in dilute sulphuric acid I : 8 in a sealed tube to 140 C. catechol and phloroglucol together with a red anhydride are obtained. In concentrated sulphuric acid, it dissolves with a deep purple coloration. It is readily soluble in alcohol and in boiling water, but requires about 1 100 parts of cold water for its solution. Hence it separates on cooling from a hot solution of gambier, and may be purified by redissolving in hot water and treatment with animal charcoal, and subsequent crystallisation. The white deposit on pit sides and on the surface and interior of leather, which is often found when gambier is largely used, and which is sometimes called the ''whites," consists of crystallised catechin. It may be dissolved by warm water or liquors, or may be decomposed by warm sul- phuric acid. This deposit is favoured by the use of hot gambier liquors, from which it crystallises on cooling. Its occurrence has also been attributed to some kind of fermentation of gambier liquors, but this is very doubtful. Catechin may be extracted from its aqueous solution by agitation with ether. It possesses no acid properties, though some writers have incorrectly called it catechuic acid. The aqueous solution gives precipitates with albumin, lead acetate, and mercuric chloride, and reduces ammonio-nitrate of silver ; but, unlike tannins, it does not pre- cipitate gelatin, alkaloids, or tartar emetic. It is oxidised by permanganate in presence of free acid, and hence, when in solution, is estimated by the Lowenthal method as " non- tannin." It is also partially absorbed by the hide powder filter * Ber., 1902, xxxv. pp. 1867 and 2408. f Ibid. p. 2410. HO LEATHER INDUSTRIES LABORATORY BOOK. ^o that the analysis of gamblers by this method is rendered difficult and uncertain. It probably possesses no tanning pro- perties itself, but its anhydrides will tan. In experiments made by the writer a slight tanning effect was observed in a cold saturated solution, which had a faint yellow colour and probably contained a trace of anhydrides, and on boiling the solution for some time, the tanning effect was increased and both the liquid and the leather took a deeper and redder colour. Its value is therefore uncertain, but the practical results from gambier seem larger than corresponds to the ready formed tannins present, and it is not improbable that in the tanning process the catechin is gradually converted into available tannins. Hunt has, however, shown * that, as estimated by the Lowenthal process, the tannin in gambier is to some extent lessened by boiling, probably by conversion into an insoluble anhydride. Although the tannins of gambier and cutch are so nearly related, great differences exist in their tanning properties. Natural Colouring Matters allied to Tannins. These are -so constantly present in vegetable tanning materials, and so frequently influence their colour reactions, that a section on the Chemistry of Tannins can hardly omit some notice of them ; and as they are closely allied to tannins in their products of decom- position, and are moreover crystalline bodies, the structure of which has been closely investigated by many chemists, and especially in recent years by A. G. Perkin and by Kostanecki and their pupils, and which is now comparatively clearly under- stood, their study can hardly fail to cast light on that of the tannins themselves.f Their relation to tannins has already been mentioned on p. 104. Like the tannins, they usually occur in nature as glucosides, generally crystalline, which are broken up by digestion with dilute acids, or by ferments, leaving the colouring matters themselves as crystalline powders, usually very sparingly soluble in water, but pretty freely so in alcohol, espe- cially with the aid of heat. They are usually precipitated by lead acetate, which, however can frequently be used in their purifica- * Journ. Soc. Chem. Ind., 1885, p. 266. t The best summary of present knowledge of the natural colouring matter is "Die Chemie der natiirlichen Farbstoffe," by Dr. H. Rupe (Vieweg und Sohn, Braunschweig). Their occurrence in tanning materials has been specially studied by Perkin in a series of papers communicated to the Chemical Society ; Chem. "Soc. Trans., 1896, Ixix. pp. 1289, 1299, 1303; 1897, Ixxi. pp. 1131, 1194; 1898, Ixxiii. pp.374, 1016. THE CHEMISTRY OF TANNINS, ETC. 141 tion, tannins and phlobaphene bodies usually coming down in the first fractions. The colouring matters found with tannins mostly dye more or less pronounced yellow shades on tin and alumina mordants and olives on chrome and iron, and belong to one group, the flavones. Others such as haematoxylin and brasilin from logwood and Brazil-wood are more or less asso- ciated with tannins, but hardly come within the scope of this section. Flavone, the at first hypothetical type of the group actually- synthesised by Kostanecki, * and its derivativey?<2^0;^/ possess the constitutions shown below ; and the colouring matters in which we Flavonol. are specially interested, quercetin, fisetin, myricetin, and morin; are all derived from the latter, and differ only in the number and arrangement of OH groups on the two benzene rings. All but fisetin, the colouring matter of Rhus cotinus ("Venetian sumach "' or " young fustic "), have OH groups in the i, 3 positions. When these are fused with potash, the intermediate (pyrone) ring is broken and OH substituted for the OC in the 5' position, so that phloroglucol is formed, though it does not actually exist in the colouring matter, which therefore does not give the pine- shaving reaction. Fisetin has no OH in the I but only in the 3 position, and therefore yields resorcinol on fusion. Quercetin and fisetin have OH groups in the 3' and 4' positions on the second ring, which on fusion becomes oxidised to protocatechuic acid, while morin, the colouring matter of Morus tinctoria (fustic), has OH groups in the 2' and 4' positions, and yields resorcinol but no protocatechuic acid in addition to phloroglucol. Myricetin, the colouring matter of many sumachs and of Myrica nagi, has OH groups in the I, 3, and 3', 4', 5' positions, and consequently yields phloroglucol and gallic acid on potash-fusion. Several of these flavones yield, in addition to the products named, acetic,, glycolic or oxalic acids from the residues of the pyrone ring, * Ber., xxvi. p. 2901 ; xxviii. p. 2302. 142 LEATHER INDUSTRIES LABORATORY BOOK. It is interesting to compare the behaviour of these flavones with maclurin, or so-called " moringatannic acid," a second crys- talline colouring matter of fustic, which not only yields phloro- glucol on potash-fusion, but gives the pine-shaving reaction. OH OH The structure of maclurin is probably OH<( )> C O \ /OH, OH so that the phloroglucol exists ready formed in the substance. In this respect it resembles the catechins. Quercetin, with its glucosides, is the commonest colouring matter in tanning materials, and is the principal one of quercitron {Quercus tinctoria), and is also found in quebracho, cape sumach (Osyris compressd}, gambier, the acacias, and many other materials, while myricetin is common in the sumachs, and some other tans of the pyrogallol class. Other Natural Tannins. It can hardly be said that any natural tannins except digallicacid have been thoroughly investi- gated, and most of them have not even been prepared in a state of approximate purity. It is known that tannins of the pyrogallol class exist in many plants, but whether these are merely impure forms of digallic acid and its glucosides or allied tannins of different constitution containing perhaps methyl or other alkyl derivatives of pyrogallol, is by no means decided in most indi- vidual cases. There is probably little doubt that the principal if not the only tannin of sumach and most galls is gallotannic acid, and apart from the soluble tannin of ellagic acid, the same is probably true of myrobalans, but is by no means so certain with regard to divi-divi, or chestnut-wood and oak-wood tannins, and many others. Bottinger * believes oak-wood tannin to be a methyl derivative of digallic acid, of the formula C 15 H 12 O 9 . 2OH 2 , which loses one molecule of water at 100 C. and the second at 135 C. He obtained a pentacetyl derivative C 15 H 7 Ac 6 O 9 ,and two brom- ine derivatives from this, Ci 5 H 10 BrAcO 9 and C 15 H 7 Br 4 AcO 9 . Koernerf acetylated oak-wood and chestnut-wood tannins, and found for the former an acetyl derivative containing 5 acetyl groups (31*8 per cent.) and with a molecular weight (by freezing- point in phenol) of 689 approximately. Bottinger's formula gives a molecular weight of 546 and 39*4 per cent, of acetyl (C 2 H 3 O), while if Koerner's results can be accepted as reliable, * Ben, xx. 1887, pp. 761-766. f Koerner, Collegium, 1903, p. 163. THE CHEMISTRY OF TANNINS, ETC. 143 the constitution must be much more complex. He, however, points out himself that with such substances cryoscopic methods are not very reliable. Koerner (loc. cit.) found for the acetyl compound of chestnut-wood tannin a molecular weight of 870 and 36*5 per cent, of acetyl, so that he considers it to contain 7 acetyl groups, and states that the acetyl compound differs in solubility and other properties from the oak-wood derivative, and that therefore the two tannins are distinct. Considering the many sources of error to which such determinations are liable, these results must be accepted with a certain reserve. It is possible that many apparent discrepancies arise from the diffi- culty of distinguishing between the tannins and their glucosides. With regard to the tannin of algarobilla pods (closely allied botanically to divi-divi) Zolfel* states that they yield 8-10 per cent, of gallotannic, and a much larger quantity of ellagitannic acid. The gallotannic acid exists as a glucoside with dextrose. The ellagitannic acid (p. 136) is free from glucose, and takes 5 acetyl groups. He believes it to be isomeric with digallic acid C U H 10 O 9 . As ellagic acid is now known to be C 14 H 6 O 8 , and to separate almost spontaneously from the ellagitannic acid, it is difficult to see how this constitution is possible. Zolfel believes the tannins of algarobilla to be identical with those of myrobalans. Some pine-tannins apparently contain pyrogallol derivatives. The bark of the silver fir (Abies pectinatd) is said to give a blue- black with ferric salts. G. Heylf has investigated the tannin of the cones of Sequoia gigantea, the giant pine of California, which gives a brownish-black with ferric chloride, and insoluble preci- pitates with salts of Ba, Mg, Cu, and Pb, which, when heated to 200 C. yield pyrogallol. Boiled with dilute sulphuric acid, in- soluble reds, and only small quantities of glucose and of gallic acid are produced. Heyl gives the constitution of the tannin as C 2 iH 2 oO 10 , and of its bromine and acetyl derivatives as C 2 iH 15 OioBr 5 and C 21 H 14 Oio(C2H 3 O) 6 respectively, but these formulae, like most others which have been given for tannins must be accepted with much reserve. Turning to catechol tannins, it is pretty certain that wider divergences exist. None of the natural tannins can be simply diprotocatechuic acid Ci 4 H 10 Of (p. 137), since it contains only * Arch. Phar., ccxxix. pp. 123-160; Journ. Chem. Soc. Abst., 1891, p. 918. t Chem. Centralb., 1901, ii. p. 312. 144 LEATHER INDUSTRIES LABORATORY BOOK. 57 '93 P er cent, of carbon, while Trimble points out that the natural catechol tannins contain about 60 per cent. (p. 108). It does not appear that they are usually glucosides, the principal pro- ducts on boiling with dilute acids being insoluble reds, which are probably condensation products, possibly anhydrides, of the tannins. Quercitannic Acid. Oak-bark tannin. This tannin may be prepared from oak bark by Prof. Trimble's method, described on p. in, who obtained the oak tannins as light-coloured powders very similar in appearance to gallotannic acid, and easily soluble in water and other solvents of gallotannic acid. It may also be prepared by evaporating the alcoholic extract, and extracting with water, which leaves the phlobaphenes or higher anhydrides undissolved. The first anhydride, which is partially soluble, may be precipitated by the addition of salt, and the quercitannic acid extracted by shaking the filtered solution with acetic ether. In very dilute alcohol it yields a pure yellow precipitate with lead acetate. In aqueous solution the precipitate is light brown. When pure, quercitannic acid gives a green-black with ferric salts, and yields nothing to pure ether or to benzene. The blue-blacks given by many oak-bark infusions have been shown by Trimble to be due to the presence of colouring matters, which in some cases can be isolated by removing the tannin with neutral lead acetate, and after filtration precipitating the colouring matter with the basic salt. If quercitannic acid be heated to I3O-I4O C., it loses water and yields a red anhydride slightly soluble in water, which con- stitutes the red colouring matter of oak bark and which has also been called " difficultly soluble tannin." It precipitates gelatin, and is one of the class which Eitner well designates " tanning colouring matters." It gives a brownish red with lead acetate and a blue-black with ferric salts ; it is difficultly soluble in water and ether, but readily so in alcohol of all strengths. Together with other anhydrides, it exists naturally formed in the bark. At higher temperatures, or by boiling with acids, a series of higher anhydrides may be obtained which are quite insoluble in water, but are soluble in alcohol and caustic alkalies. No glucose is produced by treatment of pure quercitannic acid with acids, that found after so treating oak-bark extract being due to the laevulose present. If oak-reds are fused THE CHEMISTRY OF TANNINS, ETC. 145 with potash they yield, according to Johansen,* protocatechuic and butyric acids, while both Etti and Bottinger obtained proto- catechuic acid and more or less phloroglucol. It must not be forgotten that many if not all oak-barks contain quercetin, and that this if not completely removed must yield protocatechuic acid and phloroglucol on potash-fusion, but many oak-barks give a faint reaction with pine-shaving and hydrochloric acid, while quercetin does not do so. If heated in sealed tubes with fuming hydrochloric acid, methyl chloride is evolved and red anhydrides and traces of gallic acid formed. The constitutional formula of quercitannic acid is as yet very uncertain. It is known that the tannin of oak wood differs materially from that of the bark, the former being a pyrogallol and the latter a catechol derivative, and it is doubtful whether either has been isolated in a state of purity, or free from traces of colouring matters. Trimble's tannins from different oaks, which were probably among the purest yet prepared, showed slight differences in properties which may have been due to difference of constitution, but more pro- bably to residual impurities. Trimble detected catechol in the product of oak-bark tannins heated in glycerin, and protocate- chuic acid in those of the action of hydrochloric acid and of fused potash. Pyrogallol, gallic acid and phloroglucol were not de- tected. The average of nineteen ultimate analyses of tannins of different species of oak is given by Trimble as C = 59 * 79, H = 5' 08, G = 35 "13 per cent. This agrees fairly well with Etti's formula for the tannin of Quercus pubescens, C 24: H 20 O 9 , but he states that his tannin was insoluble in water (and hence perhaps an anhydride). Elaborate researches have been made on the constitution of oak tannins by Lowe, Etti and Bottinger, with somewhat contradictory results. For details the original papers must be consulted, references to which are given below.t A * Arch. Pharm., [3] ix. p. 210; Chem. Soc. Journ., 1877, i. p. 721. t Papers referred to on the constitution of quercitannic acid : 1880. C. Etti, 'Monatshefte fur Chemie,' i. p. 262; Ch. Soc. Abs., 1881, p. 277 ; C. Bottinger, * Annalen der Chemie,' ccii. p. 269. 1881. C. Etti, Berichte d. Deutsch. Ch. Ges., xiv. pp. 1398, 1826 (Ch. S. Abs., 1881, p. 1021) ; ibid. p. 2390; Dingier, ccxli. p 69 ; J. Lowe, Zeits. Anal. Ch., xx. p. 208. 1883. Biittinger, Berichte, xvi. p. 2710, and Ch. Soc. Abs., 1884, p. 321 ; Etti, Monatshefte, iv. p. 512 (Ch. Soc. Abs, 1883, p. 994). 1884. Etti, Ber., xvii. p. 1820 (Ch. Soc. Abs., 1884, P- I 355) '> Bottinger, Ber., xvii. pp. 1041, 1123; Ch. Soc. Abs., 1884, p. 1025. 1887. Bottinger, Annalen, ccxxxviii. p. 336 ; ibid. ccxl. p. 330 ; Ber., xx. p. 761 1888. Etii, Monatshefte, x. pp. 647 and 805. L 146 LEATHER INDUSTRIES LABORATORY BOOK. good summary is given by Trimble.* For the reactions of various oak tannins and infusion see p. 163. On the whole, it seems probable that the principal tannin of oak bark is a purely catechol tannin, and that the gallic and ellagic acids, which have been detected in it, are due to an admixture of the gallotannic and ellagitannic acids present in oak wood, galls, and probably more or less throughout the tree. Apparently phloroglucol is present, but in very much smaller quantity than in gambier. It is not improbable that the tannins of hemlock and mimosa are identical with the catechol tannin of oak barkf Bottinger, however (loc. dt.\ gives the following percentages of bromine as contained in the precipitates from aqueous solutions of their tannins, which do not support their identity. Oak bark, brownish yellow precipitate . . . 28*4 per cent. Br. reddish yellow precipitate. 4,- -43' 5 Mimosa and chest- nut SECTION IX. THE QUALITATIVE RECOGNITION OF TANNING MATERIALS.^ THIS has become a matter of considerable importance, since the commercial value of an extract is dependent, not only on the percentage of tanning matter as determined by analysis, but on the nature of the tannin present which influences the character of the leather produced. It is also frequently desirable to ascertain what material has been used in tanning a given leather, and though in mixed tannages it is impossible to deter- mine this accurately, chemical tests applied to the extract of the leather will frequently give much useful information. It is obvious that, failing a clear knowledge of the tannins themselves, any really scientific scheme of separation is for the present impossible ; since, although we know that the tannins are * 'The Tannins,' i. p. 94. f See Bottinger, Ber., 1884, xvii. p. 1123. } Cp. Journ. Soc. Chem. Ind., 1894, p. 487 ; 'Gerber,' 1894, p. 195. QUALITATIVE RECOGNITION OF MATERIALS. 147 a class, we are quite ignorant of their number, which may be a comparatively small one, because many members of the group which at present receive individual names are probably mere mixtures. All that can be so far done is to separate the tanning materials into groups or individuals by reactions which have been empirically determined ; of which in many cases we do not know the chemical significance, and which are often due to other constituents of the materials than the tannins themselves. The question is still further complicated by the fact that wood, bark and fruit of the same tree often yield tannins of distinctly different character. Thus the bark of Acacia arabica (Babul) gives the general reactions of the Australian acacias, and evi- dently contains as its main constituent a tannin of the catechol group; while the pods of the same tree have a tannin which differs from any other which I have examined, but which is apparently a pyrogallol derivative. The differences between the tannins of oak bark, galls and wood are equally striking. And while cutch from the wood of Acacia catechu gives the phloro- glucol reaction intensely, the bark of the same tree hardly con- tains a trace. In the previous Section it has been shown that the founda- tions of a clearer knowledge are being laid, and while we must still take advantage of any reactions which enable us to recognise individual materials, there are certain broad distinctions of a chemical nature which we can already utilise as the basis of a classification, which we may hope to subdivide further as our knowledge increases. It must, however, be borne in mind that, for commercial purposes, the problem is not always the recogni- tion of chemical individuals merely, but of their origin. It is very probable, for instance, that the principal tannins of oakwood and divi-divi are identical ; but their commercial value in an extract may differ considerably from the presence or absence of other constituents characteristic of the several materials. Some of these constituents may be removed or altered by the processes of decolorising and rendering soluble now used in extract manu- facture, so that it often happens that reactions quite characteristic of the original material may be absent in its purified extract. We may make the beginning of a scientific classification by dividing "catechol tannins," derivatives of the dihydric phenol catechol, and of its corresponding protocatechuic acid, from L 2 148 LEATHER INDUSTRIES LABORATORY BOOK. " pyrogallol tannins " derived from the tryhydric phenol, pyro- gallol, and its corresponding gallic acid. To do this no elaborate separation is necessary, for not only do the two classes differ markedly in general character, but they are distinguished by the fact, not yet explained, but apparently without exception, that while dilute solutions of catechol tannins give an immediate precipitate with bromine water, and form insoluble bromine- derivatives, the corresponding derivatives of pyrogallol tannins are soluble and do not precipitate except in a few cases after long standing and probable oxidation. Again, when all pyrogallol tannins give blue-blacks with neutral ferric salts, most catechol tannins (and perhaps all if properly purified) give green-blacks. The principal exceptions to this rule are the Australian mimosas which give violet-blacks, and oak-bark of Qitercns robur which gives a blue-black, probably from the presence of pyrogallol deriv- atives, since its purified tannin gives the usual green-black. A crimson reaction with concentrated sulphuric acid is also usually a distinguishing mark of the catechol tannins. Another instance in which definite knowledge of constitution may be obtained by a qualitative test is that of the catechol tannins which also contain phloroglucol, and which consequently give a pink or violet coloration to deal-shaving when moistened with hydrochloric acid. The most pronounced instances of this are the true catechus from Nauclea gambir and Acacia catechu^ though many other materials give a faint reaction. It is pointed out on p. 145 that many bodies such as quercetin, which yield phloro- glucol on fusion with potash, do not contain it ready formed, and consequently do not react with deal-shaving, and the same is probably the case with some of the tannins. It must also be remembered that resorcinol, and probably some of its deriva- tives, give a bright blue coloration under the same test The other reactions used in the following tables are more of the nature of special tests, and in most cases the chemical changes involved are not understood. With many materials characteristic colour-changes take place from gradual oxidation when a tanning solution is treated as in the Payne and Parker method of analysis with large excess of a sugar-lime solution, or of baryta water. In some cases valuable information as to constituents and even as to proportions of mixture may be obtained from quantitative methods. The relation between the quantities of tanning and soluble non-tanning substances present is often characteristic ; QUALITATIVE RECOGNITION OF MATERIALS. 149 and inferences may be drawn from the relation between the results of the gravimetric (hide-powder) method of estimation and those of Lowenthal's or Payne and Parker's, since in both of these a different factor is required for catechol and pyrogallol tannins. For purposes of recognition, infusions of tanning materials may therefore be divided into three classes, as in Table I., viz. (i) Those which give a precipitate with bromine water, and green- black with iron alum, and which are pretty certainly catechol tannins : (2) those which give a precipitate with bromine water, but bluish- or violet-blacks with iron, and which for the present must be regarded as mixed or uncertain ; and (3) those which give no precipitate with bromine, and blue-black with iron, and which, so far as we know, are all derivatives of pyrogallol. These classes may be further subdivided by other reactions, but until the causes of these reactions are better understood, the classification must be regarded as somewhat artificial, since though convenient as an index, it sometimes happens that tannins probably really closely allied fall into different groups. The strength of the infusions employed is understood to approximate to that used in the LA.L.T.C. method for purposes of quantitative analysis, that is, containing about 0*4 per cent, of tanning matter ; and the table on p. 191 will give a general idea of the weight of material which should be used. The method of carrying out the various reactions is as follows : {; . Bromine Water Reaction. It is best to add the bromine water drop by drop to 2-3 c.c. of the infusion in a test-tube, until the solution smells strongly. In some cases the precipitate is slight, or forms slowly, and occasionally it is crystalline, and on this account less conspicuous, but it is usually a distinct yellow or brown flocculent one. In general terms bromine may be said to be a reagent for the catechol tannins ; precipitating all which give green-blacks with iron, and many which give blue- or violet-blacks, which are reasonably suspected of containing catechol. It does not precipitate any recognised pyrogallol tannin, but several which yield ellagic acid (bloom), such as oak barks. No notice need be taken of any precipitate which only forms after long standing, as bromine acts as an oxidising agent as well as directly, and many pyrogallol derivatives yield in- s oluble oxidation-products. For this reason it is desirable that 150 LEATHER INDUSTRIES LABORATORY BOOK. the bromine water should not be too strong ; 4-5 grm. bromine per liter being suitable, and the solution of the tannin should be faintly acid, acetic acid being added if required. Ferric Alum, Ammonium Ferric Sulphate. A I per cent, solu- tion of this salt has been chosen as easily obtainable, neutral and in a state of approximate purity. This salt forms violet crystals yielding a pale yellow-brown solution, which should not be kept too long, as it deposits iron as a basic salt. Ferric chloride, which is generally used, is almost invariably strongly acid, and of a very persistent yellow ; and in excess will give a green-black reaction with gallotannic acid itself, and its results as a test are often misleading. Ferric acetate, which has also been recommended, gives heavy precipitates with most materials, and is very strongly coloured, so that it is found less convenient than the iron alum. As the colorations are generally very powerful, they are best seen by diluting the tanning infusion freely, and adding the iron solution very cautiously. In many cases the coloration first produced, and which is the characteristic one, rapidly fades to a dull olive or brown in presence of excess of iron, probably by oxidation and destruction of the tannin. By far the largest number of tannins fall into Class I., giving precipitates with bromine and green-blacks with iron, but these may be pretty satisfactorily subdivided by a reaction with Cupric Sulphate and Ammonia. When solution of cupric sulphate is added to a tanning material infusion, it sometimes gives a precipitate, and sometimes not. This distinction is of little importance, as the presence of any salt of a weak organic acid would in all cases cause a precipitate, by neutralising the sulphuric acid of the sulphate. Cupric acetate produces a precipitate with all tannins, so far as I have observed, and it is interesting to note that in this case, as in many others of a like character, the tannin combines with, and precipitates the salt as a whole. On the addition of ammonium carbonate, effervescence takes place, and the precipitate darkens, and becomes a true tannate. When ammonia is added to the mixture of a tannin and cupric sulphate, a precipitate is invariably formed in the first instance, consisting of a tannate of copper with variable quantities of blue cupric hydroxide. With excess of ammonia, the whole of this precipitate is in many cases dissolved, while Cn others an in- soluble tannate remains. The latter is the case with all tannins QUALITATIVE RECOGNITION OF MATERIALS. 151 derived from gallotannic acid, and with many containing proto- catechuic acid, and forms a useful means of classification, though as yet the difference of structure to which it corresponds is unknown. The fact that hemlock and other pine barks, as well as cutch and gambier, yield cupric tannates soluble in ammonia, is an instructive commentary on processes which have been proposed for quantitative estimation by precipitation by ammonio-cupric solutions. Where the precipitate redissolves, it generally does so with the production of either a greenish brown, or a purplish brown coloration, which in presence of a small amount of the blue ammonio-cupric sulphate, appears either as a green or a reddish violet. It need hardly be pointed out that to see this coloration well, the copper should be in the least possible excess. Hence it is desirable to use a very weak copper solution, say at most I per cent. We thus obtain two sub-classes ; I a (Table II.), in which copper produces a precipitate soluble in ammonia, and I ft (Table III.), in which an insoluble precipitate is formed. It must be admitted, however, that while in many cases the dis- tinction is well marked, there are cases in which the solubility or the reverse is only relative, and others in which the precipitate appears to be soluble or insoluble according to the amount of phlobaphenes and impurities present. The excess of ammonia used is also not without influence in some cases. No attempt has been made to arrange systematically the results of the further reactions which are suggested to determine individual members, since in the main they correspond to no broad chemical demarcation, and in others they shade off by imperceptible degrees, so that it is only by taking all character- istics into account that any safe conclusion can be drawn. This is specially the case with the vanillin reaction for phloroglucol, which is given strongly by cutch and gambier, markedly by several other materials, and so faintly by many others that no sharp line can be drawn. The following reactions are those which I have employed for further distinction : Nitrous Acid Reaction. This reaction is obtained by adding to a few c.c. of the very dilute infusion in a porcelain basin a distinct excess of freshly prepared solution or crystals of sodium or potassium nitrite, and then 3-5 drops of decinormal sulphuric or hydrochloric acid. In typical cases the solution instantly 152 LEATHER INDUSTRIES LABORATORY BOOK. turns pink or crimson, and slowly changes through purple to a deep indigo-blue, but in others, as sumach, where the reaction is feeble, and masked by other changes, the final colour is green or even brownish. In a large number of cases, nitrous acid produces a yellow or brown coloration or precipitate, but " reaction " in the tables invariably means a series of colour- changes as above described. Nothing can as yet be said on the nature of the chemical changes produced. The reaction is ap- parently given by all tanning materials which yield ellagic acid or " bloom," but not by ellagic acid itself, either natural or syn- thetic, nor by pure gallotannic acid. It is, therefore, probably a reaction of ellagitannic acid, and is valuable for subdividing the mixed and pyrogallol tannins. It may also be obtained faintly from some of the oaks, which give green-blacks with iron, and are therefore classed as catechol tans. Stannous Chloride and Hydrochloric Acid. This reagent consists in a strong solution of stannous chloride in concentrated hydrochloric acid. If about 10 c.c. of this are added to I c.c. of the tannin material infusion, in a porcelain basin, and allowed to stand for 10 minutes, coniferous tans, mimosas and some others give a very marked pink coloration. This is specially distinct in the case of larch bark. If a small piece of larch- tanned leather be steeped in the reagent, the coloration appears very strongly. Deal-Shaving and Hydrochloric Acid, Phloroglucol Reaction. This reaction is due to the condensation of phloroglucol with the traces of the methyl-protocatechuic aldehyde (vanillin) always present in the wood of the Coniferce (cp. p. 116). A shaving or slip of any pine wood is moistened with the infusion, and then, either before or after drying, is again moistened with concen- trated hydrochloric acid. In the case of cutch and gambier and a few other materials, and of solution of phloroglucol itself, the spot becomes at once a bright red or violet, but in many cases the reaction is faint, and only appears after some hours. A decided red reaction probably always indicates the presence of phloroglucol. The statement that catechol gives a similar re- action appears to be an error ; but a violet-blue is given by resorcinol. Sodium Sulphite. A few drops of the tanning solution are placed in contact with a crystal of sulphite on a tile or in a basin . With valonia, a bright purplish pink is rapidly developed, QUALITATIVE RECOGNITION OF MATERIALS. 153 apparently by oxidation. Many other tanning materials pro- duced red or pink colorations, but in no case so marked a reaction as valonia. Concentrated Sulphuric Acid. A test-tube is rinsed out with the infusion and drained, so that only about a drop remains, and about I c.c. of concentrated acid is cautiously poured into the sloping tube, so as to form a layer underneath the tanning solution. The ring of colour at the junction of the two liquids is noted, and then they are mixed by shaking and diluted with water. About half the materials give a deep purplish crimson, called simply "crimson " in the tables, and in many cases this is decidedly pink on dilution, while in others it is obscured by brown products formed by the heat of mixture. A large number of other materials give browns or yellows only, which, when intense, may appear red, but dilute to yellows or yellow- browns. Allen recommends the addition of the acid to the residue previously evaporated to dryness, to avoid the heating produced by the combination of the acid with water, but in the table the reaction with liquid is used. The crimson reaction is also given by the " non-tannin " residues of many catechol tannins, possibly from the presence of catechins. If the residue is rubbed up with a few c.c. of alcohol, a red colour after filtration denotes the presence of non-tanning matters yielding " reds," and a portion evaporated to dryness, or nearly so, will give a crimson coloration with sulphuric acid. This reaction was suggested by the late J. E. Hughes as a test for quebracho in oak-wood extract, and has proved very useful. Of course, it might be caused also by the presence of other catechol tans. Lime Water is a most useful reagent, being very varied in the colour of the precipitates it produces, and the changes they undergo by oxidation. The reaction is best seen in a shallow porcelain basin, and time must be allowed for the changes to take place. It may be safely added in considerable excess. Allen recommends the use of a nitrometer for this and similar experiments, as the reaction can first be studied in absence of oxygen, and then air admitted. Similar colour-changes occur in Payne and Parker's lime- water process of tanning estimation, and are utilised by them as a means of distinction between different materials. General Remarks. The infusions are assumed to be of about the strength used in determinations by hide-powder, viz. about 154 LEATHER INDUSTRIES LABORATORY BOOK. cr6 grm. of dry soluble matter in 100 c.c. ; but moderate diver- gences from this do not influence results. Of course, reactions which are feeble may be imperceptible with very dilute solutions, and, on the other hand, strong solutions will sometimes give precipitates where only colorations are noted with weak ones. As some of these reactions are dependent, more or less, on the concentration of the infusion employed, it is desired to experi- ment in the first instance with known materials in order to acquire a little experience of the exact conditions. In some cases these tests are sufficient to detect the con- stituents of mixtures, but usually this will remain impossible till further and more distinctive reactions are discovered. They can, of course, be applied to the recognition of tannages when excess is present in leather, which can be dissolved out with water ; but the effect of dilute alkalies and other solvents on the reactions demands further experiment. For this purpose Andreasch's alcoholic reactions (pp. 164-167) are often more suitable. There are one or two special reactions of tanning materials which cannot be conveniently included in the tables, and may therefore be named here. The bark of chestnut-oak (Q. prinus) contains a substance which possesses a strong blue fluorescence, especially in alkaline solution. It is well seen after precipitating the tannin and colouring matter with ammoniacal zinc solution. A somewhat similar body, aesculin, is contained in the bark of the horse-chestnut. Dieterich* has shown that if an alcoholic solution of gambier from Nauclea (Uncaria) gambir be rendered strongly alkaline with sodium or potassium hydrate, and shaken with petroleum ether, the latter takes a powerful green fluorescence. The same result is obtained with an unfiltered aqueous solution to which alcohol is added, but not with a clear filtered solution, or with an unfiltered one without alcohol. It is, therefore, due neither to the tannin nor the catechin, but to some body soluble in alcohol, but insoluble in water. Dieterich believes that this is quite characteristic of Naiiclea gambir, and this accords with the writer's experience, though Eitner states that he has obtained the reaction from a sample of Acacia catechu, which he believes to be genuine. According to Dieterich the fluorescence is due to * Pharm. Central- H.N.F., 1896, xvii. p. 855. Ber. Pharm. Ges., 1897, vii. P- 153- QUALITATIVE RECOGNITION Of MATERIALS. 155 a substance which can be shaken out of the petroleum ether with dilute sulphuric acid, and recovered by neutralisation and shaking out with ether, but which gradually decomposes, leaving a red phlobaphene-like body. Andreasch, in his pamphlet on " Sicilianischer SumacJi und seine Verfdlschung" * gives special tests for Pistacia lentiscus and Tamarix africana, materials which are constantly employed as sumach-adulterants. He states that if a concentrated solution or small fragments of potassium cyanide be added to an infusion of tamarix alone or mixed with sumach, and not containing at most more than 0*75 per cent, of tannin, a dirty yellow flocculent precipitate is obtained on stirring, which settles rapidly ; but no precipitate or only traces are obtained where pure sumach alone is used. The potassium cyanide must be pure and without excess of potassium carbonate. In the case of solutions containing Pistacia lentiscus, 40 per cent, formaldehyde solution added drop by drop in small quantity to the neutral infusion of the sumach produces a light ochre yellow precipitate which occurs even in the presence of great quantities of sumach, and renders possible the detection of small quantities of pistacia with certainty. The solution first becomes turbid, and then deposits a gelatinous pre- cipitate, which is insoluble in cold water. The present writer has not succeeded in obtaining very decisive results with these reagents, but any considerable admixture of pistacia may be detected in pure sumach by the bromine-water test with which the catechol-tannin of pistacia yields a precipitate, while the pyrogallol-tannin of pure sumach does not, though on long standing it becomes turbid from oxidation-products. On the microscopic examination of sumach, see p. 407. Ferdinand Jean f suggested the detection and even the titration of chestnut-wood extract in mixture with oak-wood by the fact that when iodic acid is added to an infusion of chestnut- wood extract, a certain quantity of iodine is liberated, which can be removed by shaking the mixture with a little bisulphide of carbon, and recognised by its violet colour, or titrated in the usual way with thiosulphate. Jean stated that this reaction was not given by oak-wood, quebracho, and a variety of other tannins ; but Mr. Blockey and the writer observed distinct liberation of iodine, not only with oak-wood extracts of the best marks, but * "Der Gerber," Vienna, 1898. t Collegium, 1902, i. p. 283 ; and 1903, ii. pp. 25 and 95. 156 LEATHER INDUSTRIES LABORATORY BOOK. with infusions made direct from well verified specimens of oak- wood (both English and Slavonian) in the laboratory. It would be interesting, however, to inquire further into the causes of this reaction. Formaldehyde reacts on phenolic bodies in the presence of a mineral acid such as HC1 giving condensation products of the type HO.C 6 H 4 .CH 2 .C 6 H 4 .OH, and interesting results have been obtained by applying this reaction to the tannins. Jean and Frabot* have stated that this reaction causes a complete precipitation of all catechol tannins when the solution is dis- tinctly acid, and that no pyrogallol tannin is precipitated under these circumstances, though there is a partial precipitation when the solution is neutral. They have proposed, therefore, to estimate the catechol tannins in tanning materials by weighing the precipitate produced, and to detect the adulteration of pyro- gallol tanning matters with materials containing catechol tannins, e.g., sumach with pistacia. Stiasny f and Nierenstein,J however, have not been able to confirm the value of this reaction, as they have obtained precipitates with pyrogallol tannins alone. A method of distinguishing tanning materials which is sometimes useful, though not generally sufficient for definite identification, is that of dyeing the mordanted striped cotton used for a similar purpose by dyers. Various patterns are in use, usually mordanted with iron and alumina, and mixtures of the two, the iron giving blacks of varying shade with the tannins, and the alumina showing the presence of mordant colouring matters such as quercetin (p. 142). In this connection it must be remembered that ellagic acid dyes a feeble yellow. The strips are conveniently boiled for 15 minutes with solutions of the tannins of a similar strength to those used for analysis In some cases a further distinction may be obtained by boiling a portion of the dyed strip in a O'l per cent, solution of potassium bichromate. Mordanted strips have also been pre- pared with a much greater variety of metallic salts, and conse- quently a greater variety of colour reaction, but the writer has not been able to obtain these commercially. * Collegium, 1906, p. 435 ; and 1907, p. 1 88. t Gerber, 1905, p. 187. J Collegium, 1907, p. 52. In ordinary dyeing of unmordanted cotton with cutch, the material is boiled for an hour with about 10 grm. of cutch and 1-2 grm. per litejr of copper sulphate, allowed to cool in the solution, and " saddened " by boiling with bichromate solution of -1 grm. per liter. QUALITATIVE RECOGNITION OF MATERIALS, 157 TABLE I. PRELIMINARY CLASSIFICATION OF TANNING MATERIALS. CLASS i. (CATECHOL TANNINS.) Br. Water produces a Precipitate. Iron-Alum gives Greenish Blacks. Copper Sulphate followed by Ammonia in Excess. Precipitate redissolves. i a TABLE II. Precipitate does not redissolve. TABLE III. CLASS 2. (MIXED AND DOUBTFUL.) Br. Water produces a Precipitate. Iron- Alum gives Blue or Purplish Blacks. TABLE IV. CLASS 3. (PYROGALLOL TANNINS.) Br. Water produces no Precipitate. Iron-Alum gives Blue Blacks. TABLE V. 158 LEATHER INDUSTRIES LABORATORY BOOK. . 1 fl> K C X > S J W.!- .g b/J bJD ^M rt .S E ^-.S S .* . | CJ "? "^ S . c 1 II *i Remarks. ed coloration quercetin. ex a, ^g^ 1 Isg -8 i s S 1 2 "5 S o ** *c? ^ c ^ o ^a c/3 i; || 1 3 x J; o 8^1 II rj 11 'C "^ |o .j; H . B W * pq o PH "^ r O ^o c ^J x Q, c c ji l| |1| 6 Reddish p O *^ p t PQ I*! QM i! 'a ^ o r o ' o" 11 rt ^ J l| . i 1 i 'i J 1 is 1 1 S o 15 o y ei 'r3 O CO w 6 6 T3 -0 i & _i | ! PH o JS 5 + | if! ii! ill l^ti 'O 6 w 'TJ >r 5 T3 g IS e bfl "o ii | ; P tl eg || O Ikaline solution yields solution with petroleuir p. 154. Contains querc Cape sumach " contains glucoside of quercetin, coside tannin. Natal sumach-substitut gives dark brown. u M d o E 'Q of live-green coloration and with iron-alum. othtanne, Fichte, commoi Iron gives brownish H 2 SO 4 brownish red. I c L- u II L) < ~ ^ ffi 1 ^ u c d --. ^ d B d &-,x ^ ex -P ex P. . - c 2 do d p t ^ t/: ^ ;>, c o c S ^ o 'C o t/: *i QH ^ "^ ^ *^ -* ' p^ ^ p *TlJ O d -G & M ca *-< ^ GJO ; M V5 tj S3 tjo D 2 ,2 ,0 d i 1 1 * " : * Q > 11 I 1 a ^'^ * 1 > . i_ 2. * ^ S"S 'c' % -"> Q *""" PH O 'O T3 g O T^ SZ5 .^^ t/3 ^ t ' r- 1 13 ^^ 1 1 ? d d t/j o 8 d 53 d d ^0 v3 T3 - ^2 ** 'O ^ tJo 2 X c^ fc ^o d d .2 ! ..* 43 t3 ^ C T3 qj f * M *-' 6 .S ^ "g^ S r '^ C) * J , 0) SJ > o o S t/) u *"o S ^ w Q, b/3ji * O ex > ^uo cLt w g dj> wi d 9 t/J rl etn PM !l!l i! ^| 3 ,2 2 "52 "" g g 2 ^ ^ n o o fc'o o 13 No react, darkens. ill* t c 4J ^! H Q OJ 'M ^ O> oS~ S-^ & a>- > > s. Ji jsS l s 'o if H 1 It ll rf s . "Tr^ _ 1 ll -s 1^1 i ^ ^ ,0 > "S ^ | -^ '**^ M 5^ Q rt ,^j c li L If- s ! O O^ = - 3 SB i3 S S 160 LEATHER INDUSTRIES LABORATORY BOOK "c ^ o -B c^^ -6~ B ^^ cJ-f 5 U .S"3 J $3 '5 iS J'So^ -^ PQ U Er 3 o S y S U tL, << o o ^ o.ti S ^ a ^ w .*< li si a| i -Sls S 4 ^- -S*c ! ^ -j^-3 ^3 O, o ^ 5 8'J A Is wf 1< i i i !| ^ ^ ^^^H &L^ r^ 3 ^ So E ai^.J d" C/3 glii'g h'e S^c^g -S { g ^c.S^xi od-o. d Is , it B . <{ o ^ -r ,2; O T3 4 4 .^ o o ! 73 & 8 H a 1 o M CTJ CU " w . ^ o 00 2 8 D, 44 ^ o* d o* ;.* -" - ^ & ^ * <^ ^ ^ ^ 1 s d- * << i or Pipta- denia macrocarpa bark Acacia catechu bark "Thorn tree" bark (Acacia horrid a] Cape. Mangrove bark extract (Rhizophora mangle}. Crude Quebracho wood extract (Qu,'bracho or Loxopterygium Lor- entzii). CJ ^CJ^^ U il jfy] i-l ^|i!}i;i ISI|<^ig5 gtntea). Chestnut Oak (Quercus prinut). Tamarisk (Brusca) (Ta- marix africana). QUALITATIVE RECOGNITION OF MATERIALS. 161 c/j . ^ c3 x o3 ii G || rt G 1 .11 p^ * '&, 1 2 If c g.S s Is s ^^ Q/ 03 "cu 1 ~ ^ 9 -o O -0 CO 3 8 ,Q * "^ a; ** *j *^ ,> ^o "o ^ ^ . g j ^J .s S 2 1 -2 etf >^ III i d S 2 g S B o ** *3 S [o 5 6 "G "^ "S H . .a -d i r H CO (^ fe > t__< i a) rt w J PQ <3 S3 5 S U+U a ^ffi CO * 6 o rt 'rt i bjQ o 5 S 2 ^8 w ? S d o o Q ^ h * S S g * Z^ z OJ ^ CO fc J5 * ? '5 J_, S ^ d cu S & S i ^ & o [o -5 & & && d Cu O oo i Q,, c d d G r Q > fl C ^ > p^ ^o ^ 2 S fl c B > * >H CJ ti , 3 pH u !? rt | 'O Q il| 0. PH D Q II .SPQ 1 SG CO ^^ 1 ' 6 ^ El 3 tn 1 || 2 al a S . be 2 " d OJ * 6 d rt ^ rt 3 .S *^ <3 O c T3 O rt o T3 cS S ^ *^ "> ^ K ^ Q SZ| o 2i w *5b w wT c M 41 u ^^ X 3 .a g o 9 rt u "3 rt'-~- rt 1 , , S-4 C 'S d IE 3 d d ^ d d d d 3 g g3 DH 3 2 (D ^--j i a 1> oj CU o> o> o i Cu i r^ ^ s ^ ^s *C 3 1> X jj a ,3 p S PQ m Q 5 Ja^ca 5 o T3 T3 C a 1 S ci u U5 - B'l a -S f-st III Kliphaut bark (Rhus Thunbergii). Canaigre (Root of ^w- ntex hymenosepalus). "Talwaan"or ' Elands- bontjes" (Root ^ < /^ phantorrhiza Bur- chellii). Mimosa or Wattle barks (various Austral. Ac acice). 3abool bark, India (Acacia arabica). Dark red Austr. bark (probably an acacia). " White bar k " ^/^ar^a blanca^ South America II f| O^j S Ct t 3 . p"i ^W ^ A i M 162 LEATHER INDUSTRIES LABORATORY BOOK. * 5 c^ 2 rt 43 *5b .o -5 ~ *& wT cj rj fl ^3 O 3^'S "o a g 5=3 O C S) ^ O g 43 C! >>, PM 3 1 1 '1 1 ti Q^ D O O c* * 2< "So 'Ei 'w -J3 w s a as ^ siil 2 i ^ ' Pn""g 5^, o ^ > s ^ s I "6 J ;s s QUALITATIVE RECOGNITION OF MATERIALS. 163 s .2 ^ I-a tuo sd . S rf 3 J o ,0 5*0 3 ^^5 '3^ o" 1 3 ex, s pp i' a Q C 6 ex, d PL, 2 -~ O 1^ | S r^ ^O ^2 wi C 1 d i -i 1*1 11 d 6 "8 8 i ^ 1 ^ c 'o ^ W 'fn O M ^ ^ J< 5 r2 'fi 3 ^ 'c 13 ^ 'a JH ,2 2 ^ c ^'^ ^s. 5 i i .S -J | i i JK " o d d d > - 1 . ^T ^ 1 J cL : i G ffi -=3 m' 2 4 9 ^.5,2 ill Pinkish colou] ^- 8t1 ^ S a i|sj[ c rS ,2 4 4 4 || III 3 < g W x c ^ p PQ * PH 'o -Qpq >s Q ^-- " PH 5 | S OH S i| d o" d 6 d do - 2 rf CJ ^ 1 ^ 0,^0 . fi !& ^JS g 8^d d o || ^ T3 -^ r-H g" 8 -? n riL ^ri S c 3 o P V' fl .^20 O U CH J p^H ^^ a; rt '3 H o3 -2 *2 oj c3 Q < o 5 i PQ u o I s ^_^ ^-^ , ^ f 5 ^* T* ^ K ^^ ^$5 ^ ^ a B Vo M 1 ^ ! | * 1 1 1 _ sit S ci, ^ oj |> Willow oa 1 51 tS S a ^ g-s ^ O C/2 W i. M 2 i6 4 LEATHER INDUSTRIES LABORATORY BOOK. The following table is given by Andreasch * for alcoholic solutions, and has been found useful at the Vienna Institute for the examination of alcoholic extracts of leathers. The reagents Reagent. Spruce Bark. Oak Bark. Willow Bark. Mimosa Bark. Hemlock Bark Oakwood. Water . . Orange Yellow-white Greenish Yellow- white Dark red- Light yellow turbidity. pp., partly turbidity. pp., brown brown pp. turbidity. soluble. solution. Hydrogen As above. Yellow-white Apple-green pp. As above. Light brown Yellowish -white peroxide. pp., partly pp. and soln. flocculent pp. soluble. Hydrochloric Red-brown Yellow-brown Yellow-white As above. Dark brown pp. Pale buff acid. solution. pp., brown pp., rose-red and solution. flocculent pp. solution. zone. Sulphuric acid. Rust-brown pp. and solution. Yellow-white pp., brown Yellow-brown pp., cherry- Slight rust- brown pp., Dark rust- brown solution. Brown pp.and solution. solution. red zone. daik solution. Nitric acid Yellow-brown As above. Yellow pp. As above. Red -brown pp. Yellow pp., dark and solution. and solution. flocculent pp. brown solution. Acetic acid Yellowish .. .. .. . . , . white pp. Ammonia Brown pp., Dark yellow Turbidity. Violet-red pp., Dark brown Pp. soluble to partly soluble pp., soluble soluble in pp., insoluble red solution in in excess. in excess. excess. in excess. excess. Chloroform . Yellow-red Yellow-white Whitish M Dark brown flocculent pp., pp., yellowish turbidity. deposit. . brown solution. solution. Ethyl-ether . Light brown Light yellow pp .. Grey- violet pp. Brown pp. Slightyellowish- pp. white pp. Acetic ether . Turbidity ft M tt Benzol Reddish-brown Brown , . Reddish-black Brown layer Slight red- sediment. flocculent pp. layer below. below. brown pp. Petroleum Ether not Ether pale .. Ether faint red. .. ether. coloured. yellow. Carbon CS 2 yellow. CS 2 yellow. CS 2 green. CS 2 pale yellow. ,. disulphide. Naphthol . . Brown pp. and Brown pp. and Yellow-brown Brown pp. and , . Yellow- brown solution. solution. pp., dark red- solution. pp., dark brown solution. solution. Glycerin . Yellow tt Greenish-white . . Red flocculent Slight flocculent pp. flocculent pp. pp. turbidity. Tartaric acid . Whitish-yellow Slight whitish- Yellow-green Yellow-biown Red-brown pp. Whitish-yellow turbidity. yellow pp. flocks. pp. flocculent pp. Citric acid As above. As above. As above. As above. As above. As above. Oxalic acid As above. As above. As above. As above. Voluminous As above. red -brown pp. Trinitro-phenol Yellow pp. 0. .. .. Yellow-brown .. and solution. '. pp. ' Gerber,' 1894, pp. 195, 207. QUALITATIVE RECOGNITION OF MATERIALS. 165 were always added in excess, and the mixture allowed to stand overnight. Where spaces are left blank it is understood that no visible change takes place. Quebracho. Valonia. Myrobalans. Divi-divi. Sumach. Knoppern. Birch Bark. Turbidity. Dirty yellow, turbid over Dirty yellow turbidity. Marked yellow- brown tur- Dirty green pp. Yellow white pp. Yellow-brown turbidity. dark zone. bidity. Brown yellow As above. Yellowish pp. Yellowish pp. Green pp. As above. | Rusty-brown flocculent pp. pp. I As above. Light brown Light brown. Whitish-yellow Dark green pp. As above. Yellow-brown turbidity. turbidity. pp. pp. Dark red Slight yellow Slight yellow- Dirty reddish Light green pp., Yellow-grey pp. Dense red- solution. pp., pale brown pp. green solution. brown pp., solution. turbidity. dark solution. Slight pp., red brown Slight pale pp., dark solution. Dull red coloration. Dirty brown Dark green pp. turbidity Dark yellow Red-brown pp. pp. and solution. solution. Yellowish Dark yellow Light brown j Dull dark Yellow-brown turbidity. turbidity. turbidity. green pp. pp. Dark red- Yellowish pp., Yellowish pp., Pale yellow pp.,! Pale green pp., Dense greyish- Dark flesh-red brown pp. partly soluble, turns brown, partially soluble darkens. white pp., pp., soluble in reddens. soln. in excess. in excess, turns reddening. excess. brown. Solution pale Yellow-grey Yellow flocks. Yellow-brown Slight green Dense yellow- Slight brown yellow, above flocks. flocks. deposit. white pp. pp. red-brown. . . .. .. .. Grey-brown pp. Trace flesh- coloured pp. \\ Dirty white Pale yellow Rust-brown pp. Slight yellow Reddish-yellow .. pp., turning flocks. j pp. on long flocks. dark brown. standing. .. .. .. .. Ether yellow- .. green. Dense yellow CS 2 scarcely CS 2 scarcely CS, coloured CS 2 coloured flocks at zone. coloured, yellow coloured, yellowi green. yellow-green. flocks at zone. flocks at zone, j Yellow-brown Slight yellow- Slight yellow- Slight yellow- Green-brown Slight greyish Yellow-brown pp., dark brown pp. brown pp. brown pp. pp. pp. on long pp., dark red solution. standing. solution. tl Long standing Long standing Long standing Long standing Slight turbidity. Turbidity. yellowish pp. yellow flocks. ; slight turbidity, dark green pp. Yellow-brown Yellow-grey pp. ! Yellowish pp. ! Yellowish pp. j Greenish pp. Yellow-green ; Light rust- flocculent pp., pp. brown pp. dark red soln. As above. As above. As above. As above. As above. As above. As above. As above. Sulphur yellow j As above. Yellow-brown As above. As above. As above. pp. pp. .. Brown-yellow j Yellow-brown Turbidity first Apple-green pp. pp., turns pp. turns reddish, then lemon. yellow. yellow. 1 66 LEATHER INDUSTRIES LABORATORY BOOK. TABLE FOR ALCOHOLIC Reagent. Spruce Bark. Oak Bark. Willow Bark. Mimosa bark. Hemlock Bark. Oakwood. Salicylic acid . Light-brown pp. Yellow-white flocculent pp. jGreenish-yellow pp. Slight brown Bulky red- pp. brown pp. Yellow-white pp. j Tartar emetic . Fawn-coloured pp. Greyish-yellow Greenish-white \ Violet-red pp. Dirty brown pp. i pp., deep green pp. layer above. As above. Potassium ferro-cyanide. Yellow-white pp. Yellow-white pp. 1 Green white pp. Flesh-red pp. Red-brown pp* Slight white pp. Potassium sulphocyanide. Yellow-brown flocculent pp. sol. on heating. Yellow-brown flocculent pp. Leaf-green pp. Chocolate pp. Red-brown pp. sol. on heating Yellow-white pp., pale yellow solution. Potassium cyanide. Pale-brown turbidity. Pale-brown turbidity. Leaf-green pp., yellowish sol. As above. Pp. brown below, yellow- white above. Lime . . . Yellow-brown pp., glittering on surface. Pp. yellow- brown below, choco- late above, yellow solution. Dirty sulphur- yellow pp. Violet-blue pp., Violet-brown brown above, pp., dull brown and glittering above. Pp. white below, above blue, later brown. Baryta . . Dirty yellow pp., yellow- white solution. As above. As above. Blue-green pp., As above, brown above. Blue pp. turning hrown, glittering red- brown above. Strontia . As above. As above. As above. i Dirty blue pp. As above. Pp. white below, blue above, turning brown. Magnesia . Light-brown pp. Dirty white Violet-red pp., Grey pp. Red pp pp. ! green solution. Yellow-white pp. i Potassium chromate. Dull brown pp. Yellow-brown pp. Bright yellow pp. Brown pp. Brown pp. ' i Green-brown pp., turning brown. Mercuric chloride. Light red brown pp. Yellow-white turbidity. White pp. Light reddish- Blood-red pp. blue pp. Yellow-white flocculent pp. Mercurous nitrate. Dirty grey- brown pp. Pp. reddish Dirty yellow ! Dirty brown Red-brown pp., yellow turning pp. on long | pp. turning dull brown. standing. brown. Brick-red pp., turning brown- red or yellow - grey. i i QUALITATIVE RECOGNITION OF MATERIALS. SOLUTIONS continued-. 167 Quebracho. Valonia. Myrobalans. Divi-divi. Sumach. Knoppern. ' Birch Bark. Brownish yellow pp., dnrk red -brown liquid. Greyish- yellow pp. Yellowish pp. Yellow-brown pp. Green pp. Greyish-yellow pp. Pale rust -brown pp. ' Fawn-coloured pp. Pale grey -yellow pp. Cream-coloured pp. Ochre- yellow curdy pp. Yellow-green curdy pp. Dirty white curdy pp. Bulky pale rusty pp. Pale red -brown pp. Pale yellow pp. Cream-yellow pp. Orange pp. Pale-green pp. Yellow-green Bulky pale pp. rusty pp. Yellow-grey pp. Yellow pp. Dark yellow pp. Green pp. Orange yellow i Turbidity, pp. Slight pp. amaranth-red solution. As above. As above. As above. As above. Curdy reddish- white pp., darkens on standing. Yellow-white pp., dull brown and glittering above. Violet-brown pp., dark brown above. Pale chocolate pp. Bright yellow pp., colourless solution. Cream- coloured pp., which darkens. Green pp., turning yellow. Olive-brown pp. Flesh-red or scarlet pp. Grey-white pp., glittering chocolate-brown above. As above. As above. As above. Green pp., turning sulphur-yellow. Green pp , turning grey- brown over- night. Grey-white pp., brown above. As above. Chocolate pp., turning black. Dirty green pp., turning brown. Pale red pp., dirty grey above. As above. As above. Greyish -white pp., vermilion above. Violet pp., dark solution. Yellowish pp. Yellowish pp. Grey-brown pp. Dirty green mass. Yellow-white pp. Pale flesh col. pp. Dark, dull brown pp. Yellow-brown pp. Dirty brown pp. Dark-brown pp. Dirty brown pp. Dark red- violet pp., turning chocolate. Chestnut-brown PP- Dark turbidity. Dirty yellow pp., partly soluble. Yellow-brown pp. solution in excess. Brown pp., mostly sol. in excess. Dirty green pp., part soluble in excess, turns yellow. Yellow-green Reddish-yellow pp. pp. Chocolate pp. on long standing. Orange-yellow pp., turning dirty grey. Orange-yellow pp , turning dirty yellow. Orange yellow pp., turning dirty yellow. Grass-green pp. Orange pp., Grey pp. turning grey. 1 168 LEATHER INDUSTRIES LABORATORY BOOK. SECTION X. ANALYSIS OF TANNING MATERIALS. HISTORICAL NOTICE. THERE are few organic determinations which have been the objects of more varied suggestions and proposed methods than the estimation of tanning matters, and unfortunately it cannot be said that the problem is yet solved in a wholly satisfactory manner, although, for commercial purposes, the present methods have proved extremely useful. Of the many proposed, only those depending on the indirect weighing of the tannins absorbed by hide, and on its volumetric estimation with an oxidising solution of permanganate in presence of indigo, have held their own as practicable, accurate and suitable for general use under the test of experience, and in most cases the indirect gravimetric method is to be preferred to every other, though there are excep- tions for which the permanganate volumetric method is still useful. Before proceeding to describe the modern methods in detail, a short sketch of earlier processes is desirable, if only to save younger chemists the labour of re-inventing what has already been tried without success. Taking first the methods dependent on precipitation : the oldest is that of Sir H. Davy,* which has been somewhat im- proved by Stoddart and others, and which consists in precipitat- ing with gelatin, and drying and weighing the precipitate. This, with many materials, is almost impossible to filter off as directed by Davy ; but by the use of a little alum, and by pouring hot water on the precipitate, the latter becomes curdled into a mass which may be washed by decantation. As the precipitate con- tains varying quantities of tannin, according to the strength of solution employed and the amount of washing it has received ; as it is soluble in excess of gelatin solution, and almost if not quite impossible to wash free from uncombined gelatin and alum, the method can hardly lay claim to much accuracy. A some- * Phil. Trans., xciii. p. 233. THE ANALYSIS OF TANNING MATERIALS. 169 what better one (Warington, Miiller) depends on employing a standard solution of gelatin with a little alum, and determining the end of the reaction by filtering off a portion of the tanning solution, and ascertaining if another drop of the reagent pro- duces a further precipitate. This method is very tedious, the end reaction is difficult to hit, especially as the neutral solution gives precipitates with both gelatin and tannin ; the standard solution is very unstable, it is inapplicable to many materials, and especially to gambier and cutch, because the mixture will not filter clear, and its results are irregular, probably from the power of tannin to combine with various proportions of gelatin. Saturating the solutions with ammonium chloride gives in some cases somewhat better results. Attempts to use basic aniline colours (which are also precipitated by tannins) as indicators, by adding them to the gelatin solution, have failed from the fact that, on the one hand, the precipitation is with many tannins imperfect, and on the other, that the colour is absorbed by the precipitated gelatin tannate. Some tannins can be fairly esti- mated by precipitation with a standard solution of a basic colour such as benzaldehyde green in presence of sodium acetate, but the method is not of general applicability nor very satisfactory, and at best is rather adapted to the dyer than the tanner. The tannins are precipitated by most alkaloids, and Wagner * employed a standard solution of cinchonine sulphate coloured with rosaniline acetate as indicator, but the results proved much too low, and quite unreliable. Messrs. Trotman and Hackford f have recently proposed a direct gravimetric method for the analysis of tanning materials, in which extraction is made with alcohol in an electrically heated Soxhlet apparatus, the partially evaporated extract being diluted with water, the resinous matters filtered off, and tannin estimated in the filtrate. 25 c.c. of this is placed in a 250 c.c. flask and diluted somewhat with water. 0^25 grm. of strychnine is then weighed out and dissolved in about 50 c.c. alcohol. To this an equal volume of water is added, the mixture cooled and added to the tannin infusion, the contents of the flask being diluted to the mark and thoroughly mixed. The precipitate is collected in a weighed Gooch crucible, dried first in air, and then in the vacuum oven. Gallotannic acid and sumachs yield * Zeit. Anal. Chem., v. p. i. f Journ. Soc. Ch. Ind., 1905, p. 1098. 170 LEATHER INDUSTRIES LABORATORY BOOK. good results, a definite compound, C 21 H 22 N.,O. J . C 14 H 10 O 9 (strychnine tannate) being obtained for weighing. With most tanning materials, however, lower results are obtained than by the hide powder method, much lower indeed than can possibly be accounted for by the non-estimation of astringent non-tannins, showing clearly that all tannins are not precipitated by strych- nine. On the other hand, the fairly high results obtained with materials containing no gallotannic acid show that some other tannins can be estimated in this way. Messrs. Payne and Parker have also worked volumetric- ally with morphine, brucine, and some of the artificial aromatic bases with some success. Many metallic salts have also been used as precipitants, both gravimetrically and volumetrically, but in most cases without satisfactory result, because either the precipitation was incomplete, or other substances, such as gallic acid and colouring matters, were carried down with the tannins. A further difficulty arises in all these methods from the fact that the tannins form a class with very considerable chemical diversity, and that therefore methods which give accurate results with one tannin, either fail to act, or give quite different results when applied to another. It is, indeed, a universal fault of these methods that even if for one tannin the results are constant, no direct comparison is possible between materials containing tannins of different classes. This difficulty is to a large extent overcome by the gravimetric hide-powder method, which gives the actual weight of any tannin absorbed by hide, and thus allows of at least approximate comparison of the tanning value of totally dissimilar materials. Among the processes based on the precipitation with metallic salts which gave useful results must be mentioned that of Gerland,* in which a solution of antimony potassium tartrate is used in conjunction with ammonium chloride, to assist the settling of the tannin precipitate and prevent the co-precipitation of gallic acid. The process was improved by Richards and Palmer,f who substituted ammonium acetate for the chloride, adding 4 per cent, of a concentrated solution of the salt to the tannin solution, and employing 6*730 grm. per liter of the antimony potassium tartrate (previously dried at 100 C), of * Chem. News, viii. p. 54. t Amer. Journ. Science and Arts (3), xvi. pp. 196, 361. THE ANALYSIS OF TANNING MATERIALS. 171 which solution I c.c. corresponds to O'Oi grm. of digallic acid. The end of the reaction is determined by adding a drop of sodium thiosulphate solution to a little of the clear supernatant liquid placed on a white tile, and warming when the formation of red antimony sulphide indicates the ieast excess of the reagent. The method gives good results with sumachs and other materials containing gallotannic acid, but is somewhat tedious, and does not act well with many catechol-tannins. The tartar emetic solution is apt to be spoiled by fungi, and keeps better if an antiseptic, such as phenol or a trace of mercuric chlo- ride, be added. Copper acetate,* with subsequent addition of ammonium carbonate to remove gallic acid, also gives fair results gravimetrically with gallotannic acid and some other tannins, but it is not applicable to the large class (see p. 158) of which the cupric salts are soluble in ammonia. This criticism also applies to the use of copper salts suggested by Darton to remove tannin for estimation by the permanganate process, which will be alluded to later, since in particular cases it enables the propor- tions of different tannins to be determined quantitatively in mixture. The use of ammoniacal solutions of zinc for the same purpose by Carpene \ and others was carefully examined by Kathreiner,t but without satisfactory results. Another class of methods is "based on the amount of some oxidising substance required to destroy the tannin present. The simplest of these in theory is that of Mittenzwey and Terreil, in which a solution of tannin, rendered alkaline with caustic potash or soda, is allowed to absorb oxygen from a known quantity of air, the loss in volume of which is measured. The absorp- tion is slow, and the method difficult to carry out, and subject to several sources of error. In a more recent modification || the oxygen is furnished by a known quantity of hydrogen peroxide acted upon by lead peroxide in alkaline solution, but this appears to have no important advantage over the original method. Vaubel and ScheurerlT suggest modified apparatus, * H. Fleck, Bay. Kunst und Gew. Blatt, 1860, p. 209. t Ding. Polyt. Journ., 1875 (216), p. 452 ; 1876 (219), p. 471. \ Ding. Polyt. Journ., 1878 (227), p. 481. Mittenzwey, Journ. f. Prakt. Chem., ii. (1864) p. 815; Terreil, Compt. Rend., Ixxviii. (1874) p. 990. || A. Thompson, Compt. Rend., 1902 (cxxxv.), pp. 689-691 ; Colleg., 1902, p. 334. Tf Colleg., 1907. p. 96. i/2 LEATHER INDUSTRIES LABORATORY BOOK. and give interesting test analyses, but admit that the method is not ripe for practical adoption. None of these methods make any distinction between tannins and other astringent products which do not tan, such as gallic acid, so that a second estima- tion must be made after removal of the tannin, and the latter determined by difference. Commaille * oxidised with a known quantity of iodic acid in presence of a trace of hydrocyanic acid, and determined the excess of undecomposed iodate. The pro- cess was troublesome, not very exact, and, like all other oxida- tion processes, involved a separate estimation of gallic acid. The power of iodine to enter into combination with tannin has also been utilised by F. Jean,f who uses a standard solution of iodine, which is titrated with the tannin solution rendered alkaline by sodium carbonate. As the colour of the mixture is too dark to allow of the starch reaction being seen in the usual way, minute drops are brought in contact with a piece of filter paper rubbed with starch. The method gives fair results with some tannins, but does not distinguish gallic acid from tannin (see p. 230). The use of potassium permanganate for the estimation of tannins was first proposed by Monier,J but is not practicable in the simple form of direct addition by permanganate, from the fact that the reaction proceeds very slowly, and in the cold never reaches its theoretical limit. Gantter has since shown that it is possible to obtain useful results by titrating in boiling solution by a method which does not differ in principle from Monier's. The original method of Lovventhal || introduced the valuable improvement of adding indigo solution to the tannin solution to be titrated ; the indigo acting not only as an indicator, but controlling the oxidation of the tannin within certain definite limits. Lowenthal first assumed that only tan- nins were attacked, but as gallic acid is also similarly oxidised by permanganate, a double titration was found to be required where both were present, the tannin being removed in the second * f.'ompt. Rend., lix. p. 398. t Ibid., Ixxxii. p. 982 ; Herichte Deut. Ch. Ges., 1877, p. 430. J Compt. Rend., 1858, p. 577. Zeits. fur Angevv. Chem., 1889, p. 20; cp. also Procter, Journ. Soc. Ch. Ind., 1890, p. 260. |! Journ. Prakt. Chem., iii. (1860) p. 150. THE ANALYSIS OF TANNING MATERIALS. 173 case, and several proposals have been made to effect this object ; but from the time Lowenthal published his method of precipi- tating the tannin by gelatin in the presence of an acidified salt solution,* to the introduction in 1886 of the indirect gravimetric hide-powder method, scarcely any other method than this, with removal of tannin either with gelatin or hide-powder, was in practical use for the estimation of tannin, and in certain cases it is yet the best available. The " indirect gravimetric method " first took practical form at the Vienna Research Station for Leather Industry, and was published in its organ ' Der Gerber ' f by F. Simand and B. Weiss, who first applied it to extracts only, but afterwards, in conjunction with Eitner and Meerkatz,| described its applica- tion to solid materials. Earlier methods, which to some extent involved the same principles, were those of Hammer and Miintz and Ramspacher.|| In the former of these the specific gravity of a dilute tanning infusion was taken with a very deli- cate pycnometer or areometer before and after the tanning matter had been removed by hide-powder. A special table gave the amount of gallotannic acid which corresponded to the difference of specific gravity, according to which I per cent, solution of gallotannic acid has a specific gravity of 1*0040 at I5C. ; and as at least up to 5 per cent, specific gravity is proportional to the quantity of tannin, the percentage may be obtained by dividing the decimals of the specific gravity by 0*004. Extreme accu- racy in taking the specific gravity and in the temperature of the solution is required to obtain even approximate results, and deter- minations are still wanting of the specific gravity of solutions of other tannins than gallotannic acid. In Muntz and Ramspacher's method the liquid was detannised by filtering under pressure through a piece of raw hide, and the tannin determined either as in Hammer's method, or by evaporation. The weak point of the method is that if sufficient of the tannin infusion is passed through the hide to insure the expulsion of the water contained in it, the liquid is rarely completely freed from tannin. In * Zeit. Anal. Chem., 1877, pp. 33 and 201 ; see also p. 223. t 'Gerber,' xii. (1886), pp. I, 26, 35 ; Ding. Polyt. Journ., cclx. p. 564. \ 'Gerber/ xiii. (1887), p. 2. Ding. Polyt. Journ., xlix. p. 300. || Compt. Rend., Ixxix. p. 300. 174 LEATHER INDUSTRIES LABORATORY BOOK. experiments by the author the results, even with the greatest care, varied for the same liquor by 10 per cent, of the total tannin present. Simand and Weiss's method was originally carried out as follows : The " total soluble " in a clear filtered solution of the extract was determined by evaporation of 100 c.c. of the solution to dryness in a platinum basin, and subsequent dry ing at iooC, the solution being of such strength that this residue amounted to I *o-i '2 grm. 250 c.c. of the same solution was treated in a dry conical flask with I grm. of the dry purified hide-powder, and allowed to digest for several hours in a cool place, during which it was frequently and vigorously shaken. It was then filtered through linen, 2 grm. more of the hide-powder added, and the digestion continued for 12-16 hours more, when a further 2 grm. of the powder was added and similarly digested. At the end of this time the liquid was filtered through paper, and 100 c.c. of the perfectly clear filtrate evaporated. The residue consisted of " soluble non-tanning matter," and when deducted from the " total soluble," left " tanning matter absorbed by hide." If the whole powder were added at once to a com- paratively strong tanning solution, the surface would become tanned and no longer capable of rapidly absorbing small traces of tannin. By adding the powder successively, the greater part of the tannin is absorbed by the first portions of powder, leaving only traces to be absorbed by the last addition. With carefully purified hide-powder, free from soluble matter, the method is capable of giving good results, but is surpassed by more recent modifications. Soon after Simand and Weiss's publication, it occurred to the writer* (suggested, perhaps, by Miintz and Ramspacher's method) that the determination might be still more rapidly effected by the use of a cylindrical filter packed with hide- powder, through which the tannin liquor was slowly percolated, so that most of the tannin should be absorbed by the first por- tions of the filter traversed. This method was at once adppted by the Vienna Research Institute, with some improvement in the form of the filter, f and continued to be the standard method of the International Association of Leather Trades Chemists up to the present year (cp. p. 21 1). * Journ. vSoc. Chem. Ind., 1887, p. 94. t 'Gerber,' 1887, p. 65. THE ANALYSIS OF TANNING MATERIALS. 175 More recently a method has been published,* by Dr. Yocum, which reverted to Simand and Weiss's original system, but shortened the digestion required to about ten minutes by vigorous mechanical shaking. The results agree very closely with those of the filter, but are generally somewhat lower in tannin. With some tanning materials, and especially with gambiers, the differences are considerable. This method with modifications was adoptedfby the American Association of Official Agricultural Chemists, but has been superseded by that described on p. 216, as it was found so difficult to prepare a pure hide- powder of constant absorptive power and really free from soluble organic matters, that many chemists have turned their attention to the use of powders made insoluble by the use of chrome salts, and in one form or another these have been generally adopted. Powders made insoluble with formaldehyde have also been tried, but apparently become unabsorptive on keeping. No historical notice of methods of tannin estimation would be complete without allusion to that recently introduced by Parker and Payne f and dependent upon an estimation of the total acidity of the tanning infusion with lime or baryta solution by a method not dissimilar in principle to the determination of the saponification value of a fat (p. 304). The process has not proved wholly satisfactory, but the special information which it aims at giving is so important that it is described in some detail in Section XIV. The work of Prof. Wislicenusf with " fibroid alumina " must also be referred to, since, although in the writer's opinion it is not likely to replace the use of hide powder, it presents certain advantages which would have entitled it to very favourable con- sideration had the hide-powder methods not been previously established in technical use. When metallic aluminium is moist- ened with a mercuric salt, such as mercuric chloride, metallic mercury is reduced, and a galvanic couple is produced, which, under the influence of the air and moisture, oxidises the metal to alumina, which appears as a sort of down on the surface, of a fibrous nature, and remarkably like a growth of some form of * ' Leather Manufacturer,' 1894, No. 9; Journ. Soc. Chem. Ind., 1894, p. 494. t Journ. Soc. Ch. Ind., 1904, p. 648. \ Collegium, 1904, pp. 204, 215, 223, 225; 1905, pp. 85, 94, 213, 230; 1906, PP- 77 &7 3 l6 > 3 21 ; I 97i PP- J 57> l6 9i Zehs. fur Angew. Ch., 1904, No. 25 ; Zeits. fur Anal. Ch., No. 2. 176 LEATHER INDUSTRIES LABORATORY BOOK. mould. This constitutes the " sprouted," or " fibroid " alumina, which is employed by Wislicenus as substitute for hide-powder, and which, from its considerable adsorptive power and large surface, is capable of taking up from solution all tannins, and many of the substances which are usually classed as non-tanning. In the first instance, considerable difficulty was experienced in producing this material of constant and reliable quality ; but detailed instructions have been given by Wislicenus, and the powder is now to be obtained of satisfactory quality from Merck. It has the advantage over hide-powder that it can be used repeatedly, the carbon being destroyed by heating to something below the point of redness ; but strong ignition lessens or destroys the adsorptive powers. Being also entirely incom- bustible and inorganic, it is possible to determine the organic substances absorbed by it by elementary analysis of the powder after use. The substance is employed in a filter-tube, which, owing to the smaller quantity, only 2 5 grm. employed, has been considerably modified from the form used for hide-powder, and in its latest development is described in Collegium, 1907, p. 172. The results obtained show good concordance in them- selves, and do not differ very largely from those with chromed hide-powder, but the absorption is commonly somewhat greater and the non-tannins consequently lower. It is also by no means certain that the selective absorption of the alumina is for the same " non-tannin " substances as that of hide-powder, since its adsorption is mainly a surface action, while in the case of hide- powder there is little doubt that chemical affinities also come into play. It may be mentioned that instead of the candle-filter, Prof. Wislicenus prefers to use a perforated porcelain tube, covered with several thicknesses of filter paper, for preparing his solutions. A method mentioned by Dr. Baum* at the Frankfort Con- ference of 1906, but not since further described, depends on somewhat similar principles to that of Prof. Wislicenus. The tannin solution to be detannised is placed in an aluminium basin and submitted to electrolysis with a current of low voltage, the tannin being precipitated in combination with alumina. * Collegium, 1906, p. 373. i 77 SECTION XL SAMPLING, GRINDING AND EXTRACTION OF TANNING MA TERIALS. BEFORE proceeding to describe the practical analysis of tanning materials, careful consideration must be given to the drawing and grinding of samples, and the extraction of the tanning matter in a proper form for analysis, as in many cases greater errors arise in these preliminary operations than in the analysis itself. In the case of extracts, samples should be drawn by a sampling tube by which a portion of extract can be withdrawn from 6-10 casks of the parcel to be analysed. These portions are carefully mixed, and 4-6 oz. at least transferred to a clean bottle, A suitable sampling tube is designed and sold by Mr. Arthur Priest- man, of 275 Burlington Street, Liverpool, consisting of a small reservoir closed by a valve operated by a rod passing through a long tubular handle. Before sampling, the extract should be thoroughly stirred and mixed by the sampling tool, special care being taken to disturb any sediment which has settled to the bottom of the cask. If the extract has been standing in cask for some time, the only way really to mix properly is to take out the heads of a sufficient number of casks, and stir the contents thoroughly with a suitable stirrer or paddle, rolling full casks being quite ineffective. After thorough mixing, the tool is immersed and moved about in the cask, while the reservoir is filled by depressing the rod which passes down the handle, and which opens an aperture for the purpose. Samples should be drawn not only from the centre but also from the upper and lower corners of the cask. The various samples should be thoroughly mixed in a clean pail, and at once sealed in bottles for analysis. It is desirable, if the casks are numbered, to select for sampling, numbers equally distributed through the parcel, as the casks are usually filled in rotation. It is also of equal importance that samples in bottles or jars should be thoroughly mixed before weighing out portions for analysis, as Dr. Parker has shown * that errors of 2 per cent, in * ' Leather Trades Circular,' 1897, p. 489. N i 7 8 LEATHER INDUSTRIES LABORATORY BOOK. either direction maybe caused through the settling of the " diffi- cultly soluble tannins." The sampling of block gambiers is one of the most difficult problems owing to the irregularity of the moisture in different parts of the block. A somewhat large sample must be drawn either by perforating the blocks with a tubular sampling tool like a large cork borer,* or by cutting sections with a " fleshing knife," and the portions must be mixed by chopping or kneading, with as little exposure to the air as possible to avoid loss of moisture. Similar difficulties with regard to moisture often Centimeters. Inches. 35 FIG. 9. arise with cutch and other solid extracts, which are usually too hard and brittle to cut with the tool, and cannot be pulverised till dry. In this case it is best to weigh the whole sample, dry, and again weigh to determine loss. The sample is then pul- verised and analysed in the usual way, and calculated to original moisture. An example of the reckoning may be given. The original sample weighs 900 grm. and after drying 765 grm. only, * A suitable tool, designed by Kathreiner, is shown in Fig. 9, which can also be used for sampling sumach, if damage to the bags is not objected to. C C' is a brass or gunmetal tube with a collar B and strong handle A. The plunger D is of hard wood, and used to force the ' cores ' cut out of the cutting tube. SAMPLING, GRINDING AND EXTRACTION. 179 the loss being 15 per cent. The percentage is therefore cal- culated on 85 as representing 100 original, and 15 per cent, is added to the water. In sampling solid materials, portions should always be drawn from at least ten sacks or separate parts of the bulk, and, in the case of valonia, special care should be taken to have a fair average quantity of " beard." No attention is usually paid to this point by merchants, and the p'roportion varies greatly in different parts of the same cargo. If several sacks are spread in layers on a level floor, and then " quartered," dividing the pile first in half, and the half again in halves, cill a sufficiently small sample is obtained, this will be accomplished. Where samples must be dealt with which have not been specially drawn, it is safest to weigh out from the same proportion of beard and whole cups, bearing in mind that the beard is always the richest part of the valonia ; or to make the analysis on the cups only. Valonia beard often contains large quantities of crushed lime-stone and other rubbish added as weighting material, but frequently very injurious to the colour of leather. Special care is required in sampling and mixing sumach, as it is very frequently adulterated, and probably a portion of pure is often added on the top of the sack. In sampling myrobalans, it should be remembered that the poor and light nuts will rise to the top, and hence the hand should be plunged well into the sack. The sample, when drawn, may be ground in a small bark-mill, and, if necessary, a smaller sample drawn from this and ground a second time in a sample mill, or even an ordinary coffee- mill, though the latter is scarcely strong enough for many materials. Burroughs, Wellcome and Co.'s drug mills, No. 3 or 4, answer the purpose very fairly. A special mill of small size is also sold under the name of " Tannina " by Robert Paessler, of Freiberg, Saxony, and small sizes of the ordinary toothed grind- ing mills for tanning materials, such as Schmeija's, Glaser's, or the Hardy Pick Co.'s may also be used, especially if power is available. A laboratory disintegrator is made by Grumbach and Son, of Freiberg in Saxony, to a design of Dr. Koerner's.* If a suitable mill is not at hand to reduce the material to a coarse powder, it is necessary to prepare it for these small mills by crushing or breaking into pieces of moderate size. With many * Collegium, 1903, p. 365. N 2 i8o LEATHER INDUSTRIES LABORATORY BOOK. materials this is conveniently accomplished with a flat-faced hammer on a thick cast-iron plate, which should have raised ledges round three of its sides to prevent the fragments from flying about. Many barks are too tough to be pulverised in this way, and are most easily sampled by cutting short lengths off a large number of pieces with strong shears, or with a saw. If the bark can be secured in a firm bundle with a clamp or by binding with wire, a number of thicknesses may be cut at once with the saw. Materials are usually much more easily ground if strongly dried, but this involves determination of the moisture both on a portion of the original sample, and again on the dried and ground material as prepared for analysis, and should therefore be avoided if possible. Extraction. Extracts are most satisfactorily dissolved for analysis by placing the beaker or basin in which they are weighed in a large funnel in the neck of a liter flask, and washing the extract into the flask with boiling water. After filling up with boiling water the neck of the flask is then covered with a small beaker, and the liquid cooled rapidly under a stream of water from the tap, and made up to the mark with cold water. If the tap water is much below 15 C. it is safer to immerse in a vessel of water at about that temperature, to avoid over- cooling or local chilling of the solution. Solid extracts are generally best dissolved in a beaker by the use of successive quantities of boiling water, and treated in the same way. Care must be taken not to pour solid particles into the flask until all matter soluble in hot water has been dissolved, and if it is suspected that solution is not complete, the flask may safely be heated for a short time ori the water-bath. If the amount of insoluble particles is large, as in the case of some gambiers, it may be advisable to strain through muslin, or even to extract like a solid tanning material. A short boiling of the residue, after most of the tannin has been removed, is not generally injurious ; but long-continued boiling of most tannin solutions, and especially in contact with air, leads to the des' ruction of some tanning matter, and almost invariably darkens the colour. The writer is of opinion that with most extracts slightly more uniform results are obtained by cooling at the temperature of the laboratory for some hours instead of rapidly under the tap, and this method has been adopted by the American Association SAMPLING, GRINDING AND EXTRACTION. . 181 of Leather Chemists, but the results are somewhat lower in tannin and higher in insoluble matters ; though experiment shows that the differences are negligible as compared to other sources of error ; and in many cases want of time precludes the slower process, so that for the sake of uniformity it is considered best invariably to cool under the tap. For the same reason it is best, or at least official, to dissolve always in absolutely boiling water, for though many extracts dissolve perfectly well at lower temperatures, and sometimes filter more easily, in some cases the use of lower temperatures leads to a larger precipitation of insoluble 4< reds." It is pointed out by Mr. A. Barrett that loss is incurred with sumach extracts, and others containing ellagi- tannic acid, if boiling water is used, and better results are obtained with water not above 4O-5O C. The insoluble matter found in an extract is considerably influenced by the quantity of water which is employed in dis- solving it. This was long ago pointed out by von Schroeder, and is well shown by the following figures given by Cerych for an oak-wood extract : * i i ' Dissolved per Insoluble liter. Matter. M on-T ' -iiinins. substances. \ \ grm. grm. grm. grm. 5 o - i2 '8-44 28*56 15 I'22 18 50 27-40 30 1*65 1 8 40 27 07 It will be noted that as the " insoluble " increases the tanning substances diminish, as the former consists largely of red anhy- drides or "difficultly soluble tannins." In materials such as hemlock bark and quebracho, which contain these bodies in much larger proportion than oak-wood, the variation becomes still more marked. It is hence very important to keep to a uniform proportion of dissolved matter in the liquid for analysis in order to obtain comparable results, and this is one of the reasons why chemists interested in the analysis of tanning materials have agreed upon quantities to be adhered to (see p. 191), and Section XIII. The solubility of the " reds " is greatly influenced by tem- * 'Gerber,' 1895, p. 242. 182 LEATHER INDUSTRIES LABORATORY BOOK. perature. and consequently the quantity found will vary mate- rially with the temperature at which the liquor is filtered for analysis. Cerych (loc. cit.} found that an oak-wood extract which gave i'68 per cent, insoluble at 16 C. was perfectly soluble at 34 C. These remarks also apply to solid tanning materials. Tanning solutions for analysis should therefore be filtered at a temperature as little above I5C. as possible, and never exceeding 20 C., even if artificial cooling is necessary. The extraction of solid tanning materials presents much more difficulty than the solution of extracts. It has been pointed out by Mr. A. S. Jones that each tanning material has an optimum temperature of extraction, and that in many cases materially less tanning matter is extracted at a boil than at some less degree of heat, and considerable information on this point is given by the researches of )r. Parker and the writer.* In many cases, also, it is difficult to complete extraction without the use of more water, and the consequent production of a more dilute solution than is permissible for the gravimetric method. As a rule, the digestion of the material for some hours at a low temperature or in the cold, and its subsequent percolation with water, of which the temperature is gradually increased to 100 C., will extract a maximum amount of tanning matter, and if all the concentration is done on the later dilute percolate, which contains at best a very small proportion of the total tannin, any loss by decomposition may be neglected. It is found, however, that in nearly every case by careful percolation practically the whole of the tannin can be extracted without exceeding the permissible volume of water, and this is now the rule of the I.A.L.T.C (p. 213). Considerable difficulty arises from the tendency of many tanning materials to choke any filter on which they are placed for percolation, and this is on the whole most successfully met by the use of sand filters. One of the best appliances for the purpose is the bottle extractor devised by Dr. R. Koch,f which may be considered as an improvement and simplification of the so-cailed " Real's press." The apparatus (Fig. 10) consists of a glass wide-necked bottle which should be thin and well annealed in order to stand the heat of the water- bath. A convenient size is about 200 c.c. capacity. It is wise * Journ. Soc. Ch. Ind., 1895, p. 635, and Prin. Leather Mfr., p. 344. t Ding. Polyt. Journ., 1887, cclxvii. p. 513. SAMPLING, GRINDING AND EXTRACTION. 183 before .use to heat these bottles in water to boiling, and allow them to cool in it, to lessen the risk of their cracking during an extraction, and it is often difficult to get bottles which will stand the test. The bottle is closed with a rubber cork, through which pass two glass tubes ; one through which water is introduced extending about half an inch below the cork to prevent the cold water coming in direct contact with the heated sides of the bottle ; while the other, which serves as an outflow, almost reaches FIG. 10. the bottom of the bottle, where it is widened to a funnel-shaped end, which in use is covered with a piece of the silk gauze used by millers. Above the cork both are bent at right angles for connection with other tubes. In use a layer of about 2 cm. of fine sand (purified by digestion with hydrochloric acid and washing), is first introduced into the bottle, and then the weighed quantity of ground tanning material. The funnel-tube and cork are now inserted, any slight disarrangement of the layers being of no consequence, as the sand speedily settles to the bottom on 1 84 LEATHER INDUSTRIES LABORATORY BOOK. filling the apparatus with water, which is done by sucking the short tube, while the funnel-tube is connected by a short length of rubber tube, with screw pinchcock, with a second tube bent at right angles and dipped in a beaker of water. The same tube afterwards serves to convey the percolate into a gauged flask. In special cases, where it is desired to obtain the extracted material free from foreign matter, the sand may sometimes be dispensed with. Where, from the fibrous character of the material, or other reasons, it is found difficult to insert the funnel-tube after filling the flask, it may be inserted first, and the flask filled by sliding the cork up the tube, and using a funnel for the sand and material. FIG. ii. After filling, the cork is secured in the bottle, either by wiring like a champagne bottle, or by the special clamp shown in Fig. 1 1 ; the apparatus is placed in the water-bath, and the shorter tube is connected with a water supply about \\ meter above it by tubing, also provided with a screw pinchcock. The second tube is arranged to drop into a liter flask, as shown in the figure. The water-bath may now be gradually heated, and the percola- tion carried on at any desired temperature. If the gauze chokes, as is not unfrequently the case, it may sometimes be made to run, by sucking a little water backwards, as in filling the appa- ratus. A slight modification of the Author's, which, together with SAMPLING, GRINDING AND EXTRACTION. 185 the Koch apparatus was accepted as official by the Turin Con- ference of the I.A.L.T.C., and which has the merit that it can be arranged from the apparatus found in every laboratory, is shown in Figs. 12 and 13 (from the catalogue of Messrs. Reynolds and Branson, of Leeds, who supply the apparatus). In place of a closed bottle, an ordinary open beaker is used, which is placed in a water-bath. A common thistle-head funnel is bent twice at right angles to form a syphon. The head, covered with silk gauze C, rests on the bottom of the beaker, the other end being lengthened by a piece of slender glass tube, connected by india-rubber tube fitted with a screw pinch-cock A, the C! FIG. 12. whole being held in position by a clamp. The beaker is filled with sand B and tanning material, as in Koch's apparatus, water is poured in, and the whole raised to the requisite temperature in the water-bath, when percolation is started by sucking, and warmed water added as required. As compared to Koch's, the apparatus has the advantage that, being open at the top, it can be stirred with a rod in event of the material clogging the filtering layer of sands, but on the other hand the beaker has to be watched and filled up as required, instead of having a supply cistern of a definite quantity of water which can be left to percolate automatically. In ordinary complete extrao i86 LEATHER INDUSTRIES LABORATORY BOOK. tion in the Leeds University laboratory, the requisite quantity of material to give an extract containing 3 5-4*5 grm. of " tanning matter," see p. 191, per liter is weighed out into the apparatus, and allowed to macerate in water for some hours in the cold. The temperature is then gradually raised, and the percolation begun at about 50 C. About 500 c.c. is run through in i-i hour, after which the temperature of the bath is rapidly raised to 100 C. In the case of sumachs, a good part of the first 500 c.c. should be percolated at or below 40 C., and with canaigre and other starchy materials the whole extraction is best done below 50 C. The Turin Conference decided that only I liter should be used for extraction, and the small quantities of tannin which could be extracted with further portions of water should be neglected, but if solutions must be concentrated by evaporation, the boiling, if not done in vacuo, should be protected as much as possible from excess of air, by placing a funnel in the neck of the flask, and should be confined to the later and more dilute percolates. After making up to I liter, the solution, whether of an extract or solid material, requires filtration, and for the sake of uniformity this is desirable even when the solution is quite clear and trans- parent. In some cases it is very difficult to obtain such a filtrate, and special thick filter-paper is required, Schleicher and Schull's " No. 605 extra-hart " being very suitable. Even with this the patience of the operator is often severely taxed, as it is often necessary to return to filtrate again and again to the filter till absolute clearness is obtained. As a last resort, a teaspoonful of well-purified kaolin may be shaken up with the solution before nitration. It was early pointed out by von Schroeder that small traces of tannin are removed by the kaolin as well as by the filter paper itself, and many attempts have been made to over- come this source of error, which usually amounts to per cent, or more on liquid extracts. A method of determining a correction was devised by the author,* and adopted by the I.A.L.T.C. in their conference in 1902 ; given up in 1904 for the Berkefeld " filter-candle," f but again accepted by the Tannin Analysis t Collegium, 1902, pp. 119, 145, 151. * Berkefeld Filter Company, Celle. SAMPLING, GRINDING AND EXTRACTION. 187 Commis.ion of 1906 (see p. 213). The "candle" was first sug- gested by Kathreiner, but successfully introduced by Parker and Payne.* This filtering medium is very indifferent to tannins and other matters really in solution, and the filtrate is generally perfectly clear, but in some cases, especially with quebrachos, is slightly opalescent. For the sake of uniformity it was agreed by the conference of Turin, 1904, and by the commission it then appointed,! that all matter which has passed through the filter should be regarded as "soluble." It has, however, since been found that in some cases as much as 10 per cent, of matter will pass through the candle, which is removed by filtration to optical clearness. The " candle " adopted consists principally of porous infusorial siliceous earth and is about 7 cm. long, 3 cm. external, and i cm. internal diameter. Various methods of mounting have been suggested. Dr. Parker cements with shellac on a glass tube with a projecting collar, and fits with a rubber cork into a wide and deep ordinary funnel, which in turn is fitted with a rubber cork or collar into the neck of the usual vacuum filtration flask, from which the air is exhausted either by a water-pump or a syringe. The pumping must not be continuous, but as soon as the necessary vacuum is formed the connexion with the pump is closed with a pinchcock, to pre- vent loss by evaporation, and only opened if the vacuum becomes insufficient. Another, and, on the whole, preferable arrangement is to invert the candle in a narrow cylinder, connecting it by a syphon tube with the vacuum flask ; in which case the tube should be prolonged to the bottom of the " candle," so as to insure complete and rapid removal of liquid filtered. In the author's laboratory the open end of the filter is somewhat enlarged with a conical reamer (or even the tapered end of a screwdriver) to fit a suitable india-rubber cork, and is hardened by painting round the edge with shellac varnish. Before use, the filters should be thoroughly cleansed from soluble matter, by boiling for two hours with 10 per cent, hydrochloric acid, and thorough washing, and are then dried before washing. About 300 c.c. should be rejected (and if a dry filter is employed, may be used for non-tannin determination), before the portion is taken for evaporation. After use the filters are cleansed by brushing * Collegium, 1904, p. 249. t Ibid., 1904, p. 343, 1905, p. 45. t Il>id., 1905, p. 66. 188 LEATHER INDUSTRIES LABORATORY BOOK. with a soft tooth-brush, and filtering distilled water till it comes through colourless. If, as rarely happens with very muddy solutions, the filter chokes and filters too slowly, the surface may be cleansed by brushing during filtration. It is best to dry the filters after washing, as if kept wet they become very soft. If the filter appears choked with organic matter it may be washed with very dilute ammonia, then with dilute acid and finally with water till neutral before drying. This should always be done if a candle which has been used for extracts with insolubles is employed for solubHised extracts, as otherwise the latter are liable to dissolve out'the difficultly soluble matter in the candles. As it sometimes occurs that filter-candles and the necessary apparatus cannot be obtained, it is well to give some particu- lars of the method of filtration through paper with correction which was proposed by the author, and which previous to the adoption of the " candle " was the official method of the I.A.L.T.C., and is again made official in cases of difficulty, and the results agree quite accurately with those obtained with the " candle," when the filtration is repeated to optical clearness. The present official directions will be found on p. 213. Those of 1902 were as follows : " The filtration of solutions for analysis can be carried out with any paper which is considered most suitable, and if desired kaolin may be used. The amount of tannin absorbed by the filter-paper must be taken into consideration, and a correspond- ing correction made, which is ascertained by the similar filtration of a perfectly clear tannin solution. " In order to estimate the error caused by the filter-paper, 500 c.c. of the tannin solution to be analysed is filtered until perfectly clear, preferably by the method of filtration for which correction is to be made. 50 c.c. of this clear solution is now evaporated to dryness, dried and weighed, to determine " total soluble No. I." A second portion of the same filtrate is now filtered exactly in the manner for which correction is to be made, and 50 c.c. of this filtrate evaporated, dried and weighed. This residue is No. 2. Subtracting No. 2 from No. I gives the correction which is necessary to add to the " total soluble " found in analysis. " It is advisable not only in the analysis but also in the second estimation for correcting the filter-error, that at- least the first 1 50 c.c. which passes through the filter-paper should be discarded, SAMPLING, GRINDING AND EXTRACTION. 189 or used -for detannisation, and the next 50 c.c. used for evapora- tion. During the whole filtration it is advisable that the filter should be kept as full as possible. Where kaolin is used to get a clear filtrate the same quantity should be weighed in each case. One or two grammes are recommended. This is washed by de- cantation with about 75 c - c ' f the tannin solution, and then washed on to the filter-paper with a further quantity of the tannin solution, of which 200 c.c. are filtered as above." As filtration by the candle may be assumed to need no correc- tion, an alternative method is to filter solutions to optical clear- ness by the candle, which is usually possible by repeatedly return- ing the filtrate ; and also by the method for which correction is to be made, when the difference in weight on evaporation of 50 c.c. is the required correction. It must be clearly understood that an average correction only is required, since of course the determina- tion for every sample of material and every sheet of paper is impracticable. In practice it has been found that on the average of many determinations a correction of about +0*005 g rm - must be added to the residue from evaporation of 50 c.c. of any tanning material through Schleicher and Schull's "605 extra-hart" of 17 cm. dia- meter ; for S. and S. 590 the correction is negligible ; DreverhofTs 311, 15 cm., requires -f- 0*002, and the use of 2 grm. kaolin about -f 0*0025 grm. in addition to that required by the filter-paper. Of course different filters, even out of the same packet, will often yield different corrections, but as the whole correction amounts to under 0*7 per cent, on an oakwood extract, slight fluctuations are not of great practical importance. Perhaps, for the sake of completeness, some allusion should be made to the earlier methods of extraction, which for ordinary purposes have fallen out of use, since it has been recognised that with many materials heat must be sparingly employed. In early times, and for the weak infusions required by the Lowenthal method, it was customary simply to boil a weighed quantity of the finely powdered material with a liter of water for half an hour or more, and to make up the whole, including the solid matter, to i liter. This method gives in many cases very constant results, though somewhat below those obtained by percolation, but from its simplicity may sometimes prove useful wliere merely comparative results are required. A much more accurate method 190 LEATHER INDUSTRIES LABORATORY BOOK. was that formerly employed by Prof, von Schroeder, who, however, relinquished it in favour of Koch's method, described above. Von Schroeder's extraction apparatus consisted of a perfectly cylindrical vessel of cast tin, about 12 '5 cm. deep and 7 cm. diameter. A strainer covered with fine muslin fitted it like a piston. The powdered material was placed in the cylinder, and stirred up with 200 c.c. of cold water. At the end of an hour the piston was inserted and pressed down gently, the clear liquor poured off, and the process four times repeated with hot water at intervals of half an hour, placing the cylinder in a water-bath. The liquor was made up to I liter, and, if necessary, filtered. To determine separately the " easily soluble tannin " (viz. that extracted by cold water), Real's press was employed, which consisted of a cylinder, fitted with a tap at the bottom, and a water-tight cover at the top, with a tube through which water was forced by the pressure of a column of liquid. A small sieve, covered with a disc of linen, was placed in the bottom of the cylinder, and on it the tanning material previously tho- roughly moistened with water, and the tap closed. The press was then filled with water and left fifteen hours under a pressure of ij meter. The tap was then opened and I liter allowed to run through in the course of about two hours, and the infusion so obtained was thoroughly mixed by shaking. The material was finally exhausted like a new material in von Schroeder's apparatus to extract the difficultly soluble tannin. Both of these pieces of apparatus are practically superseded by Koch's bottle extractor and its modifications. An apparatus on the principle of Soxhlet's extractor (p. 285), but made of copper, was introduced by the Vienna Research Institute,* and modifications have been largely used in America, but from the facts already mentioned it will be obvious that continued boiling of tanning extracts is inadmissible. If used at all it should only be for the last traces of tannin after the bulk has been removed at a lower temperature,f but it has been shown by Eachus % that continued steam extraction removes matters which are absorbed by hide-powder, but which do not precipitate gelatin. * ' Gerber,' 1887, p. 67. f Journ. A.L.C.A., i. p. 88. \ Collegium, 1907, pp. 384, 386. SECTION XII. TANNIN DETERMINATION BY THE HIDE-POWDER GRAVIMETRIC METHOD. THE official methods of the International Association of Leather Trades Chemists, and of the American Association of Leather Trades Chemists are given in Section XIIL, but it is necessary here to explain a few details more fully to give reasons for the exact methods prescribed, and to discuss the various suggested modifications. An infusion of the tannin material, or a solution of extract, of such a strength as to contain O'35 to 0*45 grm. of tannin matter per IOO c.c. (or as nearly as possible 0*4 grm.) is prepared by the methods described in the preceding section. The following table gives the approximate quantities required. Variations from the prescribed dilution are specially apt to lead to errors with materials like quebracho, which contain much " difficultly soluble " tannin. THE APPROXIMATE AMOUNT OF DIFFERENT MATERIALS TO BE WEIGHED OUT FOR ANALYSIS TO MAKE UP ONE LITER OF SOLUTION. Barks, etc. Extracts. Grammes. Grammes. Algarobilla . . 8-9 Oak wood, sp. gr. I '2 or Canaigre 15-18 over 15 Divi-divi 9 Chestnut (liquid) . 14 Hemlock bark . 32-36 (solid) . 7 Mimosa bark . ii Quebracho (solid). 6 Myrobalans . 15 (liquid) 9-13 Oak bark . 3-3 6 Mimosa D . IO-I2 Oak wood . . 50-100 Gambier (block) . IO Quebracho wood . 20-22 ,, (cube) . 7 Sumach 15-16 Mangrove (liquid) 9 Pistacia lentiscus . 20-22 (solid) . 7 Pine bark . 32 Cutch 7 Willow bark 36 Myrobalans liquid) 16 Chestnut wood '. 45 Hemlock . 10-14 Mangrove bark 10 Pine Bark . 16 Valonia 14-15 Valonia beard IO-II Spent tans . 50-100 192 LEATHER INDUSTRIES LABORATORY BOOK. Determination of " Total Dry Matter" It is necessary for the calculation of analyses to determine the amount of moisture present in the sample, since this varies in solid materials to a considerable extent with atmospheric conditions, and in extracts according to the degree to which they have been evaporated. The total weight of substance taken, less the moisture, gives, of course, the " total dry matter." The drying, however, presents difficulties in some cases. On the one hand many tannins are easily alterable by heat, parting with water of combination and yielding anhydrides, and on the other, hygroscopic moisture is obstinately retained. Thus different temperatures and different methods yield varying amounts of moisture, and it is often impossible to define the point where hygroscopic water is driven off, and the loss of chemically combined water begins. A standard method would be to dry at ordinary temperature, and preferably in vacuo, over concentrated sulphuric acid, but this is too slow for practical requirements. In dealing with dry ma- terials therefore, it is better in every case, to work with as small a quantity as is permissible, taking into account the accuracy of the balance employed. As it is not possible in the analysis of the tanning materials to obtain an accuracy greater than O'l per cent, or one part per thousand, I grm. is quite sufficient with a balance weighing to milligrams, and with balances of greater accuracy, proportionately smaller quantities may be employed. In the case of liquid and pasty extracts, I grm. should never be exceeded, and if possible it is better to work with smaller quantities. The only exceptions to this rule are such materials as cannot readily be accurately sampled in small quantities, and in this case it may be necessary to weigh out larger quantities even at the expense of additional time and trouble, or to dissolve and dry an aliquot part of the solution as described below. The drying of solid materials is accom- plished in basins or other vessels which are preferably fitted with a cover to lessen absorption of moisture during weighing, though with rapid weighing this is not essential. The drying of liquid and pasty extracts is more difficult, and is best done in the flat bottomed shallow basins which are used for the subsequent determinations, and which should be 7 or 8 centimeters in diameter and capable of holding at least 50 c.c. ,In such basins the extract spreads itself to a thin layer ; while, if the basins ar e THE HIDE-POWDER GRAVIMETRIC METHOD. 193 rounded, it collects at the bottom and dries on its surface, forming a coherent coating from which the moisture of the interior of the mass escapes with great difficulty. In cases where extracts are solid or too thick and pasty to spread themselves without assistance, the addition of 20 or 25 c.c. of water, in which the extract is dissolved on the water-bath in the basin, will con- siderably facilitate drying. The author proposed the use of alcohol for the same purpose, which certainly leads to rapid drying, but on the whole he is of opinion that water is to be preferred. Where extracts will dissolve in water to a turbid solution which can be thoroughly mixed by shaking, the best method of determination of both water and dry matter is to mix the liquid thoroughly, after making up to the liter (or 2 liters where it is desirable to weigh a larger sample), and before filtration, and to pipette 50 c.c. into a basin as usual. The residue obtained in this way X 20, or the sum of duplicate determinations x IO, is obviously the dry matter contained in the substance weighed out for analysis, and the difference between this and the original weight is the moisture ; while the difference between the " dry matter " and the " total soluble " obtained by evaporation after filtration is the matter "insoluble at 15 C." As regards drying apparatus, the German Section of the I. A. L. T. C., have selected the Moeslinger drying oven, which is simply a water jacketed oven with small compartments, the interior of which cannot reach a temperature exceeding 95-98 C. In this apparatus drying is, in the author's opinion, much too slow, at least 8 or 9 hours being required with many materials to obtain approximately constant weight. This length of time is objectionable not only on account of the delay, but because in most tanning materials oxidation, leading to a gain of weight, goes on at the same time as drying ; so that in some cases the sample begins to gain weight instead of remaining constant, and even the lowest weight found must be too high, as oxidation no doubt occurs more or less during the entire drying. The American Leather Chemists' Association have officially adopted Reed's Combined Evaporator and Dryer,* which seems open to the same objections. In the writer's opinion a preferable method is the use of an air-oven heated to a temperature of IOO-IO5 C., in which the drying is much more rapid, while * Journ. A. L. C. A., 1906, p. 32 ; see also Collegium, 1907, p. 140. O 194 LEATHER INDUSTRIES LABORATORY BOOK. the danger of oxidation is probably not increased. Great care, however, is required even in ovens with automatic regulation of temperature that 105 C. is not exceeded in any part of the oven \ and that the vessels, especially if of metal, do not come in contact with the heated walls. There can be little doubt that the vacuum drying oven is the most perfect means for drying tannin residues, accomplishing the work in two or three hours with as little exposure to oxidation as possible. There seems no reason why the vacuum-oven should be used at a temperature below 100 C., as from its expense it cannot become the ordinary means of drying in small laboratories, which must dry at about 100 C., and greater differences are likely to arise between this and working at a temperature at 60 or 70 C. than would occur by drying at 100 C. for the shorter period required by in vacuo. After drying for a sufficient length of time, the sample is cooled in a desiccator, and rapidly weighed ; and is then returned to the oven for half an hour or more, and the process repeated, placing the weight previously found upon the scale, so that the weighing can be almost instantaneous. If the loss of weight exceeds 2 milligrams, the sample should again be returned to the oven and the process repeated. If larger quantities than I grm. are being dried, a somewhat larger margin may be allowed. Determination of " Total Soluble" Matter. 50 c.c. of the solution filtered by one of the methods described in the previous section, is measured with a pipette into a weighed glass or porcelain basin as described in the previous paragraph, and evaporated to dry ness on a water-bath.* It is then dried eight hours, or overnight, in a water-oven at 100 C., or, preferably, three hours in an air-oven heated to IOO-IO5 C., or two hours in a vacuum-oven at 100 C., cooled in the desiccator! and weighed till constant as before. About three hours in the vacuum-oven * Platinum, nickel, or aluminium basins may be substituted for porcelain with some gain in rapidity of evaporation. Platinum is important where the ash of the residue must be determined, as in the analysis of old liquors, and but for its cost is most ad- vantageous ; nickel is less fragile than porcelain, but not suitable for ignitions, and alu- minium is not to be recommended unless lightness is of great moment, as it is often attacked by the liquors to a slight extent ; and this is specially apt to occur with the impure commercial aluminium used for cooking utensils. If it is used, care must be taken to prevent contact with the copper water-bath, which sets up galvanic action and increases oxidation. t It is important that dry calcium chloride, or some other good water-absorbent, should be employed in the desiccator, as the dry residues are very hygroscopic. THE HIDE-POWDER GRAVIMETRIC METHOD. 195 and five hours in the air-oven are generally sufficient. Ultimately the residue may begin to gain weight from oxidation, and absolute constancy cannot always be obtained. In case of gain being noticed, the minimum observed weight must be taken, and this itself will obviously tend to be in excess of the truth, either from oxidation or imperfect drying, or both. The weight obtained after deducting that of the basin is that of the soluble matter of ^o part of the substance originally weighed out, and hence the percentage is readily obtained by dividing double the weight of the residue in milligrams, or the sum of two duplicate determina- tions, by the weight in grams of substance employed. Determination of "Soluble Non-Tanning Matter'' Up to June 1907, the official I. A. L. T. C. method of detannisation has been by the hide-powder filter (see p. 211). The syphon-filter generally employed (Fig. 14) is of a bell -shape, of a total length of 7 cm. with a diameter of 3 cm. in the cylindrical part, and I 8 cm. in the neck, and resembles a bottle from which the bottom has been removed. Into the neck of this is fitted an indiarubber cork, with a piece of fine glass tubing about 30 cm. long projecting slightly through the cork, and bent into the form of a syphon. The short limb of this within the bottle is plugged with a tuft of cotton or glass wool, and the bottle is uniformly filled with between 6 * 5 and 7 5 grm. of hide-powder. The satisfactory filling demands some practice, and depends much on the quality of the powders. If the powder is too tight, or swells too much on being wet, the filter refuses to run, or runs too slowly ; if too loose, the liquid forms channels and is not properly detannised. With pure hide-powders there is a tendency for the liquid to pass up the sides of the glass, and it is best to pack up pretty firmly round the sides, leaving it as loose as possible in the centre ; but this difficulty has been almost entirely overcome by the admixture of cellulose (not exceeding 20 per cent.) with the hide-powder, so that it can usually be packed firmly and evenly. The use of cellulose for this purpose was first proposed by F. Cerych.* When full, the powder is kept in place by a piece * ' Gerber,' 1895, p. 241 ; cp. also ibid., 1896, p. 62. O 2 y 30cc_ FIG. 14. 196 LEATHER INDUSTRIES LABORATORY BOOK. of muslin secured over the bottom of the bottle with an india- rubber band. If the powder does not entirely fill the bell, the vacant space may be occupied with purified cotton or glass wool. The packed filter is placed in a beaker or tumbler, with its open bottom resting on the bottom of the glass, and the tanning infusion is added very gradually little by little, so as to wet the powder by capillary absorption. When the filter and tumbler are filled with liquid, the syphon is started by sucking with a piece of india-rubber tube, and the filtered .liquid is allowed to drop into a gauged cylinder. The first 30 c.c. are placed on one side, and the next 60 collected, from which 50 c.c. are pipetted, and evaporated for non-tannins. This should be quite or nearly colourless, and if a little of the liquid is now allowed to drop from the syphon into the first 30 c.c., no cloudiness should be produced. The soluble matter of the hide-powder contained in the first 30 c.c. is in this way a much more delicate test for tannin than any gelatin solution. Should the filtrate not be perfectly free from tannin, the entire operation must be repeated, using a more diluted solution. The time required to obtain the necessary quantity of filtrate with a well-filled filter does not exceed I hour after the hide- powder has been thoroughly wetted ; but a somewhat longer time is not injurious. As it is extremely difficult to regulate the rate of filtration by the tightness of packing, it is usually best to use a piece of rubber pipe and a pinchcock on the syphon to regulate the flow to a rate of one drop in two seconds, which will give about 90 c.c. per hour. Veiy slow filtration, extending over more than two hours, usually increases the hide-substance dissolved, and consequently the amount of non-tannins found. Cerych shows that high temperature of the laboratory during filtration also increases the amount of dissolved hide-substance, and advises that the temperature should be kept below 20 C, ven if in hot weather artificial cooling has to be resorted to. The temperature also influences the amount of " insoluble reds " >C ji II ^ vO OO OO TJ- 10 \O R^ ll VON^ONVOt^ t^t-0 211 w d rt & O fOO r-fOi-irOOO xoxo LO vOvOOO Os-^--^-r^ o ^J 2 vOvOt- t^ON NVOVON '-" ' <& & ' z z & % . . vOrh*- SOC^NOON^H, O u < 4 1 ^1 2 ^?^ - 8^2?S^^^ OTANNIC i a 1 Washed 1 1 klU -TSIIj Si nJ <; ."2 y I C ON^J-I-I COOO w Nt^QOOO lOio O p 3 C 3 S 1 l| ""=" 2 "--""" C/5 *K 1 03 B - 0*^0 *^0 ^<^N 25 cS >, J3 CJ | O^ fO GO s * ON O ^^ OO CO ^^ ON ^ 2 8 u5 ONOTJ-HH NTJ-^OONON M 1 1 | S l 1-1 ON to O s < > 1 S s I E annins. zz 1 z .. z zg.1 $zi to H 3- O O O vo 00 >o ON O^*~" - ri- l>. O H | . 8 % & N < 8^-^ -> c? & 8 Q < h t > J5 c OOOO N t^ O O vooO'-' < ^-t^ O O O O ** N N O N ro to t^ to T^ vO ^ OO OO ^O TJ- to to ^j-vO to VO tr> vo _i c5Q ^ fc OO O O O O OOOOO O O 8 E h W i x*! to vO vO ^j"vO tovo to^O O *O tovo to O O to O *o O N O to j -a ThTh^n-rJ-^Nn-fO^-N^N^OVONxoNiO-tON s -d-0 .-d -TJ -d-C-O .-d --0-0 -T3 --d "0 - Material Empl rtoj .oJ rt .rt^rt rt .rtrt .rt .^ .rt . OO O 0^>0 00 0_0 O 111 I a! silTL'l o 11|1'111'|| OOOOQOOOOO 0*0 u J^S c3 o o o d d u -N rorfvovo r-ooONO N THE HIDE-POWDER GRAVIMETRIC METHOD. 209 w n c * Ss-l S'S'S III II ro ow oo p OO t* oo oo ^ CO O\ CX5 O\ i/5 ' &.S I *J o ON u->ba\cooo o M Cl J5 O M P >H MVD Os W O> fO OO to VQ b i-ii-i o wONCo ro vnrhr>.Tf NONi-iTr oo O\ 8a\o t^vo loN CO ON T?- O N CO oo v ^- >v3 *? Nt^fOi-i OvO ThOvOOO OMD b b b b b b b b b b b b o 0*00*00 0*00*00 ^ o VO N *0 1-1 *0 esc c e d c c gs S.2 2 'a'fi '" '"a ''SlS'S *'S ''S'S *'5 *'a p g 3(3 C C CUC C CC C C3 (3 S. rt ..<5 .^3*^ .^5 '-2f5 -^ --5 -.2 OOT3O O O'.O O OO OTJOT5O -'- " - - '''* > *' c^o 210 LEATHER INDUSTRIES LABORATORY BOOK. The following table by Prof. Weiss*, referring to the dry chromed powder in use at the Vienna K.K. Institute for Leather Industry may also be quoted. Dry Residue per 100 c.c. Original Solution. Chrome Powder Filtrate. Plain Powder Filtrate. Tannin 0*418 0-418 \ 2-806 / 0-418 \ 1-035 / 0-418 \ 0-405 / 0-418 \ 0-778 / 0*623 0*015 2-770 1-043 0-385 0-817 0-618 O'OII 2-768 I'037 0-367 0-772 Q'599 Tannin + Cane-Sugar .... Tannin + Glucose . . Tannin . + Gallic Acid .... Tannin + Dextrin Extractum Graminis . Further tables will be found in Collegium, 1906, pp. 421, 432,. 442, 443; and 1907, pp. 50-56, and 165. It may further be stated that all acids and strong bases are absorbed by hide (though the latter can hardly be present in a tanning liquor) ; and that many metallic salts and most dyes are also absorbed. SECTION XIII. THE OFFICIAL METHODS OF TANNIN ESTIMATION. IN the preceding Sections the development of the hide-powder method of tannin estimation, and the principles on which it is based, have been discussed at some length, and all that seems further necessary is to give the full text of the present official methods of the International Association and the American Associations of Leather Trades Chemists, together with a few notes on their differences and on special points in their manipulations. These notes are referred to 'by means of * 'Gerber,' 1905 (No. 746), p. 277. OFFICIAL METHODS OF TANNIN ESTIMATION. 211 numbers in the text. The divergences between the two methods are small and unimportant, and the International Association accepts the analyses of members of the American Association as official when done by the American method. The following is the text of the method recommended by the "International Association, and described in their organ " Collegium," 1907, p. 249. THE NEW "INTERNATIONAL" METHOD OF TANNING ANALYSIS. By resolution of September 19, 1906, the LA. L. T. C. in conference at Frankfort a. M., appointed an International Com- mission to investigate the various methods of tanning analysis and test by means of comparative work, and finally to report their conclusions to the Executive Committee on or before April i, 1907. The Executive Committee was empowered, if necessary after further reference to the Commission, to adopt their report and to make the same the official and compulsory method of the I. A. L. T. C. The Conference specially stated that the retention of the filter bell was not a sine qua non. The Analysis Commission consisted of the following : For Scandinavia, V. Boegh ; Italy, Baldracco ; France, Meunier ; Belgium, Wauters ; Germany, Paessler ; America, Reed ; Austro- Hungary, Eitner ; England, J. G. Parker, as representing the extract trade, Lepetit ; with Procter as chairman ; Berger as collaborator in place of Klenk. The following are now the final recommendations of the Commission : General Regulations. The Executive Committee have decided that any method which conforms to the conditions of Sections I to 4 of the following statement may be regarded as conforming to the recommendations of the International Commission on Tannin Analysis, but that members of the International Association must work according to the detailed directions contained in Sections 5 to 8. I. The Solution for Analysis. This must contain between 3' 5 and 4-5 grm. of tanning matter per liter, and solid materials P 2 212 LEATHER INDUSTRIES LABORATORY BOOK, must be extracted, so that the greater part of the tannin is removed at a temperature not exceeding 50 C. 2. Total Solubles. These must be determined by the evapora- tion of a measured quantity of the solution previously filtered till optically clear both by reflected and transmitted light ; that is, a bright object such as an electric light filament must be distinctly visible through at least 5 cm. thickness, and a layer of I cm. deep in a beaker placed in a good light on black glass or black glazed paper must appear dark and free from opalescence when viewed from above. Any necessary mode of filtration may be employed, but if such filtration causes any appreciable loss when applied to a clear solution, a correction must be deter- mined and applied as described in Section 6. Filtration must take place at a temperature between 1 5 C. and 20 C., or the actual temperature shall be stated on the Report. 3. Total Solids. These must be determined by drying a weighed portion of the material, or a measured portion of its uniform turbid solution at a temperature not exceeding 100 C. in vacua or 105 C. in air till constant. " Moisture " is the difference between 100 and the percentage of total solids, and " Insoluble " the difference between " Total Solids " and " Total Solubles." 4. Non-Tannins. The solution must be detannised by shaking with chromed hide-powder till no turbidity or opal- escence can be produced in the clear solution by salted gelatin. The chromed powder must be added in one quantity equal to 6 to 6*5 grm. of dry hide per 100 c.c. of the tanning solution, and must contain not less than o * 5 and not more than 2 per cent, of chromium reckoned on the dry weight, and must be so washed that in a blank experiment with distilled water, not more than 5 mgr. of solid residue shall be left on evaporation of 100 c.c. All water contained in the powder should be determined and allowed for as water of dilution. LA.L.T.C. Official Method. The following sections give the detailed method of carrying out the analysis adopted by the LA.L.T.C. for the use of its own members. 5. Preparation of Infusion. Such a quantity of material shall be employed as will give a solution containing as nearly as OFFICIAL METHODS OF TANNIN ESTIMATION. 213 possible 4 grm. of tanning matter per liter, and not less than 3 5 nor more than 4*5 grm. Liquid extracts shall be weighed in a basin or beaker and washed with boiling distilled water into a liter flask, filled up to the mark with boiling water, and well mixed, and rapidly cooled to a temperature between 15 and 2O C, after which it shall be accurately made up to the mark> again well mixed, and filtration at once proceeded with ; sumach and myrobalans extracts should be dissolved at a lower tem- perature. Solid extracts shall be dissolved by stirring in a beaker with successive quantities of boiling water, the dissolved portions being poured into a liter flask, and the undissolved being allowed to settle and treated with further portions of boiling water. After the whole of the soluble matter is dissolved the solution is treated similarly to that of a liquid extract. Solid tanning materials previously ground till they will pass through a sieve of 5 wires per centimeter, are extracted in Koch's or Procter's extractor with 500 c.c. of water at a tem- perature not exceeding 50 C. (2), and the extraction continued with boiling water till the filtrate amounts to I liter. It is desirable to allow the material to soak for some hours before commencing the percolation which should occupy not less than 3 hours, so as to extract the maximum of tannin. Any remain- ing solubles in the material must be neglected, or reported separately as " difficultly soluble " substances. The volume of liquid in the flask must after cooling be accurately made up to i liter. 6. Filtration. The infusion shall be filtered till optically clear (3) (see Sect. 2). No correction for absorption is needed for the Berkefeld candle, or for S. and S. 590 paper if a sufficient quantity (250-300 c.c.) is rejected before measuring the quantity for evaporation ; and the solution may be passed through re- peatedly to obtain a clear filtrate. If other methods of filtration are employed, the average correction necessary must be deter- mined in the following manner. About 500 c.c. of the same or a similar tanning solution is filtered perfectly clear, and after thorough mixing 50 c.c. is evaporated to determine " total soluble No. I." A further portion is now filtered in the exact method for which the correction is required (time of contact and volume rejected being kept as constant as possible) and 50 c -c. is evapo- 214 LEATHER INDUSTRIES LABORATORY BOOK. rated to determine " total soluble No. 2." The difference between No. I and No. 2 is the correction sought, which must be added to the weight of the total solubles found in analysis. An alter- native method of determining correction, which is equally accu- rate and often more convenient, is to filter a portion of the tanning solution through the Berkefeld candle till optically clear, which can generally be accomplished by rejecting 300 or 400 c.c. and returning the remaining filtrate repeatedly ; and at the same time to evaporate 50 c.c. of clear filtrate obtained by the method for which correction is required, when the difference between the residues will be the correction sought. Note. It is obvious that an average correction must be ob- tained from at least 5 determinations. It will be found that this is approximately constant for all materials, and amounts in the case of S. and S. 605, 150 c.c. being rejected, to about 5 mgr. per 50 c.c. and where 2 grm. of kaolin are employed in addition, to 7^ mgr. The kaolin must be previously washed with 75 c.c. of the same liquor, which is allowed to stand 1 5 minutes and then poured off. Paper 605 has a special absorption for a yellow colouring matter often contained in sulphited extracts. It is proposed that the Commission should be asked to deter- mine average corrections for the more important methods of fil- tration and report at an early date. 7. Detannisation. The hide-powder employed shall be of a woolly and not granular texture, thoroughly delimed, preferably with hydrochloric acid, and not requiring more than 5 c.c. of N/io NaOH or KOH to produce a permanent pink with phenolphthalein on 6J grm. of the dry powder suspended in water (4) ; and the detannisation shall be conducted in the following manner : The moisture in the air-dried powder is determined and the quantity equal to 6*5 grm. actual dry hide-powder is calculated, which will be practically constant if the powder be kept in an air-tight vessel. Any multiple of this quantity is taken according to the number of analyses to be made, and wet back with approximately ten times its weight of distilled water* ; 2 grm. per hundred of dry powder of, crystallised chromic chloride (Cr 2 Cl 6 i2H 2 O)t is now dissolved in water and made basic with * Very woolly powders require slightly more than 10 times their weight of water, t Kahlbaum. OFFICIAL METHODS OF TANNIN ESTIMATION. 215 0*6 grm. Na 2 CO 3 by the gradual addition of 11*25 c.c. of N/i solution, thus making the salt correspond to the formula Cr 2 Cl 3 (OH) 3 (5). This solution is added to the powder, and the whole churned slowly for I hour. In laboratories when analyses are continually being made it is more convenient to use a 10 per cent, stock solution, made by dissolving 100 grm. of Cr 2 Cl 6 i2H 2 O in a little distilled water in a liter flask, and very slowly adding a solution containing 30 grm. of anhydrous sodium carbonate, with constant stirring, finally making up to mark with distilled water and well mixing. Of this solution 20 c.c. per 100 grm. or 1*3 c.c. per 6*5 grm. of dry powder should be used. At the end of one hour the powder is squeezed in linen to free it as far as possible from the residual liquor, and washed and squeezed finally with distilled water, until on adding to 50 c.c. of the filtrate, I drop of 10 per cent. K 2 CrO 4 and 4 drops N/io AgNO 3 , a brick-red colour appears. Four or five squeez- ings are usually sufficient. Such a filtrate cannot contain more than o-ooi grm. of NaCl in 50 c.c. The powder is then squeezed to contain 70-75 per cent, water, and the whole weighed. The quantity Q containing 6*5 grm. dry hide is thus found, weighed out and added immediately to 100 c.c. of the unfiltered tannin infusion along with (26*5 Q) of distilled water (6). The whole is corked up and agitated for 15 minutes in a rotating bottle at not less than 60 revolutions per minute (7). It is then squeezed immediately through linen, I grm. of kaolin added to the filtrate, stirred and filtered through a folded filter of sufficient size to hold the entire filtrate, returning till clear, and 60 c.c. of the filtrate is evaporated and reckoned as 50 c.c. or the residue of 50 c.c. is multiplied by . The non-tannin filtrate must give no turbidity with a drop of a I per cent, gelatin 10 per cent, salt solution (8). 8. The Analysis of Used Liquors and Spent Tans. This shall be made by the same methods as are employed for fresh tanning materials, the liquors or infusions being diluted, or concentrated by boiling in vacuo, or in a vessel so closed as to restrict access of air, until the tannin matter is, if possible, between 3*5 and 4-5 grm. per liter, but if from any cause it is impossible to reach this concentration the weight of hide-powder used shall not be varied from 6J grm. in consequence. 216 LEATHER INDUSTRIES LABORATORY BOOK. The results shall be reported as shown by the direct estima- tion, but it is desirable that in addition efforts shall be made by determination of acids in the original solution, and in the non- tannin residues, to ascertain the amount of lactic and other non-volatile acids absorbed by the hide-powder, and hence returned as " tanning matters." In the case of tans it must be clearly stated in the report whether the calculation is on the sample with moisture as received, or upon some arbitrarily assumed percentage of water ; and in that of liquors whether the percentage given refers to weight, or to grm. per 100 c.c. ; and in both cases the specific gravity shall be reported. ***** In carrying out the above resolutions, this method may be brought into immediate use from that date, but becomes compul- sory for all members of the LA.L.T.C. eight weeks after this publication, viz. September 75, /po/. Analyses carried out according to the findings of the Com- mission must be marked on the report "Analysed according to the resolutions of the LA.L.T.C. of September ip, 1906." The resolution of M tiller, " That reports of analyses shall only be sent out on a mean of at least two separate analyses which closely correspond," should be again carefully noted. I take this opportunity of expressing to all the members and workers in the Commission the thanks and appreciation of the whole LA.L.T.C. for their unselfish and painstaking work. PROF. DR. H. BECKER, President of the LA.L.T.C., on behalf of the Executive Committee. FRANKFORT ON MAIN, July n, 1907. OFFICIAL METHOD OF THE AMERICAN LEATHER CHEMISTS ASSOCIATION FOR TANNIN ANALYSIS. I. Crude Materials. 1. Moisture Determination. Upon receipt of the. sample, grind promptly and dry 10 grm. in the manner and for the period specified for evaporation and drying in extract analysis. 2. Preparation of Sample for Extractions. Sample must be OFFICIAL METHODS OF TANNIN ESTIMATION. 217 dried at- a temperature not exceeding 60 C, and then ground to such a degree of fineness that the entire sample will pass through a sieve of 20 meshes to the inch (linear). 3. Amount of Sample and Proportion of Water for Extrac- tion. For fresh materials the amount of sample and proportion of water for extraction should be such as to give between 0*35- 0-45 grm. tannin per 100 c.c. of solution. For spent materials this proportion should be approximated as closely as practicable. 4. Extraction of Sample. Extraction should be conducted in a form of apparatus that permits the removal of the extractive solution from the influence of sustained high temperature, and shall be continued till a portion tested with gelatin salt solution fails to give a precipitate. At least 400 c.c. of the first portions of extractive solution should be removed and not subjected to further heating. A thin layer of cotton must be used in order to prevent fine material passing over. 5. Analysis. After extraction and dilution, solutions must be heated to 80 C, and analysis conducted as per Official Methods for Extracts. In case of weaker dilutions than the Official Method specifies, the amount of hide-powder must be reduced in proportion to the reduction of tannin. Ten grm. of the air-dried sample should be dried as in (i)< to determine moisture content of the portion extracted, and the analysis calculated and reported upon a " dry " basis. The tannin in fresh materials should also be reported on the basis of the moisture content of the sample " as received." II. Analysis of Extracts. 6. Amount and Dilution for Analysis. Fluid extracts must be allowed to come to room-temperature and weighed in stop- pered weighing bottle. Such quantity shall be taken as will give from 0*35-0*45 grm. tannin per 100 c.c. of solution, dis- solve in exactly 900 c.c. of distilled water at 80 C., and make up to mark after standing not more than 20 hours, nor less than 12 hours. Temperature must not go below 20 C. 7. Total Solids. Thoroughly mix solution, pipette 100 c.c. into tared dish, evaporate and dry as directed under " Evapora- tion and Drying." 8. Soluble Solids. To I grm. of kaolin in a beaker add 75 c.c. of solution ; stir and pour on 590 S. and S. 15 cm. pleated ^218 LEATHER INDUSTRIES LABORATORY BOOK. filter paper ; return filtrate to paper for one hour, keeping filter full. At end of hour pour solution from filter or remove with pipette. Bring 800 c.c. of solution to 20 C. ; refill the filter with this solution and begin to collect filtrate for evaporating -and drying so soon as filtrate comes clear. Keep filter full. Evaporate and dry the first 100 c.c. of filtrate, as per " Evapora- tion and Drying." Funnels and receiving vessels must be kept covered during collection of filtrate for evaporation. 9. Non- Tannins. A quantity of hide-powder sufficient for the number of analyses to be made shall be prepared in the following manner : Digest with twenty-five times its weight of water till thoroughly soaked. Add 3 per cent, of chrome alum in solution. Agitate by either shaking or stirring occasionally for several hours and let stand over night. Wash by squeezing through linen, continuing the washing until the wash water gives no precipitate with barium chloride. Squeeze the hide, using a press if necessary, so that the wet hide will contain between 70 and 75 per cent, of water. Use approximately 20 grm. of wet hide for moisture determination. Add to 200 c.c. of the original solution such quantity of the wet hide as represents from 12 to 13 grm. dry hide. Shake for ten minutes in some form of mechanical shaker and squeeze immediately through linen. Add 2 grm. kaolin to the filtrate, stir and filter through folded filter (No. I F Swedish, recommended) of size sufficient to hold entire filtrate, returning till clear. Evaporate 100 c.c. of the filtrate. The weight of the residue must be corrected for the dilution caused by the water contained in the wet hide-powder. The non-tannin filtrate must not give a precipitate with a I per cent, gelatin 10 per cent, salt solution. 10. Tannin. The tannin content is shown by the difference between the soluble solids and the corrected non-tannin. III. Analysis of Liquors. 11. Dilution. Liquors must be diluted for analysis so as to give as nearly as possible 0*7 grm. solids per 100 c.c. of solution. 12. Total Solids. To be determined as in Extract Analysis. 13. Soluble Solids. To be determined 'as in Extract Analysis. OFFICIAL METHODS OF TANNIN ESTIMATION. 219 14.' Non-Tannins. To be determined by shaking 200 c.c. of solution with an amount of wet chromed hide-powder, containing 70 per cent to 75 per cent, moisture, corresponding to an amount of dry hide-powder shown in the following table : Tannin Range per 100 c.c. Dry Hide-Powder per 200 c.c. 0-35-0*45 grm. , . . . 8-10 grm. 0-25-0-35 grm. .. ,' 5- 8 grm. 0-15-0-25 grm. . . '. 2- 5 grm, o'oo-o' 15 grm. . , o- 2 grm. Solutions to be shaken for non-tannins as in Extract Analysis ; 100 c.c. must be evaporated as in Extract Analysis. IV. Evaporation and Drying. 15. Evaporation and Temperature. All evaporations and dryings shall be conducted in the form of apparatus known as the " Combined Evaporator and Dryer," at a temperature not less than 98 C. The time for evaporation and drying shall be 1 6 hours. 1 6. Dishes. The dishes used for evaporation and drying of all residues shall be flat-bottom glass dishes of not less than 2| inches diameter nor greater than 3 inches in diameter. Notes on Official Methods. i. Variations from the prescribed strength cause the largest errors where difficultly soluble tannins such as those of quebracho are present, as the more dilute the solution the larger proportion is dissolved. In this case also differences of temperature have the largest effect on the result, since these difficultly soluble tannins are readily brought into solution at higher temperatures, thus causing a lessening of the so-called u insolubles " and an increase of tannins. It must be remembered that a large propor- tion of these insolubles are really of the nature of tannins, and under suitable conditions contribute to make leather. The writer has shown that skin can be thoroughly tanned by agitation with the insolubles of quebracho extract merely suspended in water. It has been found that with most extracts a slightly larger quantity is brought into solution if the whole of the water used is at boiling temperature, though the difference between this and the American prescription of 80 C. is quite negligible. The same may be said with regard to the question of cooling rapidly 220 LEATHER INDUSTRIES LABORATORY BOOK. or slowly to the required temperature ; the International method has the advantage of much greater rapidity of execution, and in cold weather there is less danger of cooling during the night below the prescribed temperature, which would lead to slightly lower results ; but with matters like quebracho, which contain much difficultly soluble tannin, the separation with slow cooling is probably slightly better, and the filtration easier. With materials containing ellagitannic acid, the separation of bloom is greater the higher the temperature, and the longer the solution is allowed to stand. 2. It is important that the greater part of the tannin of solid materials should be removed at a low temperature, and not afterwards subjected to heat, which especially with bloom-giving materials precipitates a portion of the ellagitannic acid as bloom. Boiling has also been shown to destroy tannin in many catechol tannins, and it is probable also that at higher temperatures a portion is fixed by the fibre or proteids of the tannin material. The reason for restricting extraction to I liter is principally to- increase uniformity and save time, and with careful extraction and fine grinding all tannin available for the tanner can be extracted with this quantity, though by more prolonged extrac- tion a slightly larger amount of solubles and a trace more of the difficultly soluble tannins is obtained, but the troublesome and tedious concentration of the later and weaker runnings is involved. 3. By the American " contact filter " method it is in many cases impossible to obtain optically clear (i.e. brilliant) filtrates, and it is assumed that by uniformity of operation, a uniform degree of opalescence can be obtained, and consequently a concordant if empirical amount of " insolubles." It has, however, been shown in the author's laboratory that in some cases a solution may appear transparent to transmitted light, and yet contain as much as 10 per cent, of matter which must be filtered out before a brilliant solution is obtained ; while 5 per cent, is not uncommon ; and as filter paper cannot be obtained of uniform closeness, and trifling differences of temperature and manipula- tion affect the fineness of division of the opalescent material, no real concordance of result can be obtained. The International Association has therefore determined that actually brilliant solutions must be obtained, and it has been sh6wn that such solutions cannot contain more than at most a few tenths per cent, of matters which can be removed by further filtration- OFFICIAL METHODS OF TANNIN ESTIMATION. 221 With most extracts brilliant solutions can be obtained by filtra- tion, repeated if necessary, through the Berkefeld candle, which, if the first 200 c.c. be rejected, does not absorb any appreciable quantity of soluble tannin, but in some cases, such especially as hemlocks and quebrachos, it is necessary to use Schleicher and Schull's " extra-hard " paper, with or without the addition of kaolin, and this paper has an appreciable absorption even for soluble tannins, which must be allowed for by the method given in the text. Most other filter papers with or without kaolin have an appreciable though smaller absorption, which should be similarly corrected. 4. A granular powder * gives equally concordant results, and absorbs a lesser quantity of non-tannins, but as woolly powders are mostly in use, it has been thought best to prescribe their universal employment, although their surface absorption is some- what greater. The degree of original acidity of the powder is of considerable importance, as excess of acid lessens the amount of chrome taken up, and makes the chromed powder more acid, with the result of greater absorption of non-tannins. It is easy, for instance, to make a granular powder as absorbent as a woolly one by additional acidification, and it will be found that most of the slight want of concordance in results of different powders are due either to marked differences of texture, or to irregular acidification. 5. The use of basic chrome chloridef in place of normal chrome alum has the advantage of chroming more rapidly and fully, and it is believed that the results are more uniform and less affected by slight differences in acidity of the original powder, but the particular chrome salt used has only a minor influence on the result. Chlorides have the advantage over sulphates of being easily estimated by N/io silver nitrate, which facilitates the determination of the acidity of the powder and the sufficiency of the washing, which is of great importance. Strictly taken, it is impossible to wash a powder chromed with normal chrome alum till it will give no precipitate with barium chloride, as slow hydrolysis of the basic chrome sulphate fixed in the powder con- tinues for an almost unlimited time, and there is no simple test which shows when the sulphates have been reduced to a negli- gible amount. * Bennett, Journ. Soc. Chem. Ind., 1907, p. 455. t Procter and Bennett, Journ. Soc. Chem. Ind., 1906, p. 1203 ; 1907, p. 79. 222 LEATHER INDUSTRIES LABORATORY BOOK. 6. The method of determining moisture* prescribed by the International Association is believed to be more convenient in practice, and equally accurate with the American method of direct drying, and if both methods are carried out properly, the effect of any small difference on the analysis is within the limits of negligible error. In proof of this, it may be pointed out that a I per cent error in moisture determination would only cause an error of 0*05 per cent, of tannin in a 30 per cent, extract. Small errors in the International method may be caused by excess of solubles in the original hide-powder, or by loss of powder by careless manipulation, but are almost compensated by the larger amount of water which must then be added to make up (26* 5 Q). On the other hand the American method is subject to at least equal errors from the impossibility of insur- ing an absolutely uniform sample of the moist powder, and the danger of driving off water of combination in the drying, especi- ally if the powder is not thoroughly air-dried before submitting it to a high temperature, which from the time involved, is im- possible in practice. 7. The time or violence of the agitation makes no practical difference in the result so long as it is sufficient to insure the complete removal of the tannin, as shown by the salted gelatin test, and it is obvious that any other equivalent form of shaking machine, or even of hand shaking can be used without affecting the results. In fact the words " or other equivalent method of agitation " were inserted in the report, but inadvertently omitted in printing. Earp's statements with regard to the inefficiency of the gelatin test have no foundation in fact, if the test is properly carried out. 8. Nov definite instructions for drying are issued by the I.A.L.T.C , beyond the statement that residues must be dried till constant, and that the temperature must not exceed 105 C. in air, or 100 C. in vacua, since it is felt that the subject requires closer investigation before any more detailed direction is given. It is believed by the Commission that drying in vacuo is the more perfect method, but the cost of the apparatus prevents its being made compulsory in all laboratories. The air-oven requires great care in its use and a good thermostat to secure uniform tempera- ture. The Reed combined evaporator and drye'rf is very con- * Procter and Bennett, Journ. Soc. Chem. Ind., 1906, p. 1203. t Journ. of Am. Lea. Chemists' Assoc., 1906, p. 32. OFFICIAL METHODS OF TANNIN ESTIMATION. 223, venient, and gives uniform results, but involves 16 hours drying, and is open at least to the theoretical objection that change and oxidation may take place in some extracts from slow evaporation and long continued heating. 9. The detannisation of liquors requires further investigation. The amounts of hide-powder prescribed by the American Asso- ciation (par. n) have not proved in all cases sufficient for a complete detannisation, and it was thought best by the Inter- national Commission to prescribe a uniform amount of powder with a view to concordance. It is clear, however, that, especially in presence of large quantities of non-tannins of an acid character,, the determination of small amounts of tannin cannot be regarded as of great accuracy. (Compare p. 197.) SECTION XIV. VOLUMETRIC METHODS OF ESTIMATING TANNINS. IN recent times volumetric methods have been almost entirely- superseded by gravimetric methods with hide-powder, but there are still cases in which they may be employed with advantage, and it is quite possible that in the future they may again assert their importance for general use. Several of the older methods are briefly referred to in Section X. Where only one or two analyses are to be made at one time, the preparation and adjustment of solutions are much more tedious- than gravimetric analysis ; but where a number of successive titrations are required, volumetric methods are considerably more rapid. The most important of the methods hitherto in use is that of Lowenthal. It has the advantage that it can be applied direct to solutions, however dilute, without the need for previous concentration, and if gelatin precipitation is used, it is much less affected by the presence of gallic or other fixed acids than the hide-powder method, and is therefore well adapted for the analysis of weak and waste liquors for technical purposes, for the systematic testing of spent tans, and for the analysis of sumach and myrobalans, which contain much gallic acid, which,. 224 LEATHER INDUSTRIES LABORATORY BOOK. in the gravimetric method, is wholly or partially estimated as " tanning matter." The method of titration with permanganate in presence of indigo was originally published by Lowenthal in the ' Journ. f. Prakt. Chem.,' in 1860,* who at first thought that only tannins were oxidised. It was, however, soon shown that gallic acid and other non-tanning substances present were also oxidised, and Neubauer \ proposed, after determining the total astringents, to remove the tannin by animal charcoal, and by a second titra- tion to determine the tannin by difference. Lowenthal soon afterwards! described the method of precipitation with gelatin in presence of salt and acid, and at the same time pointed out that the tannin might also be removed by the use of hide- powder, as in Hammer's method (p. 173). Some years later, it was shown by Simand that the precipitate of tannin and gelatin was to some extent soluble in the salt and acid solution em- ployed, so that the non-tannins came out always too high, and that the error became greater as weaker solutions were employed. He showed that this solubility might be calculated in any given case by determining the non-tannins with two different strengths of tannin solution, and finding what deduction for solubility would bring them into constant ratio to the tannin present ; but as this complicated the analysis, he preferred to absorb the tannin with hide-powder, or with decalcified bone or horn-piths, ground and purified in the same way, and preferably dehydrated with alcohol. It was not recognised at that time that these materials absorbed material quantities of gallic acid when present, and therefore introduced an error in the opposite direction, a fact pointed out by Procter in i876; and Simand's suggestion was generally adopted. Procter || endeavoured to overcome the * iii. p. 150. t Zeitsch. f. Anal. Chem., 1871, p. I. % Zeitsch. f. Anal. Chem., 1877, pp. 33, 201 ; 1881, p. 91. Proc. Newcastle Chem. Soc., 1876, p. 214 ; Journ. Soc. Chem. Ind., 1886, p. 81 ; 1887, p. 94. || The method is as follows: To 50 c.c. of the infusion 28*6 c.c. of a gelatin solution of 2 grm. per 100 c.c. is added. After shaking, the mixture is saturated with salt, which brings the volume up to 90 c.c., and 10 c.c. of dilute sulphuric acid (con- taining I vol. of concentrated acid in 10) and a teaspoonful of pure kaolin are added. It is best to do this in a flask in which it can be well shaken, after which filtration may be at once proceeded with, although it is safer to let it stand an hour or two (the flask may be cleansed with caustic soda solution). 10 c.c. of this filtrate ( = 5 c.c. of the original infusion) are employed for titration. VOLUMETRIC METHODS OF ESTIMATING TANNINS. 225 solubility of the tannin precipitate by saturating the solution with salt,* which precipitated the tannin much more perfectly, as well as nearly the whole of the gelatin, and at the same time he introduced the use of kaolin, which assists the precipitation, and the production of a clear filtrate in a very marked degree. It was, however, pointed out by Hunt t that under these con- ditions a marked amount of gallic acid was precipitated from liquors containing it, and he went back to a proportion of salt even less than that originally used by Lowenthal, but retained the use of kaolin or barium sulphate. Although Hunt's results can only be rigidly correct by a compensation of errors, his method yields, perhaps, the nearest approximation which can be obtained in the case of materials containing gallic acid, and if it were worth while to strive for greater accuracy in a method which is at best only approximate, it would not be difficult to apply Simand's correction in a simple form by rigidly adhering to constant quantities. Turning from the separation of tannin to the details of the titration process, it must be remarked that Lowenthal himself pointed out that the function of the indigo was not merely that of an indicator, but that it also acted as a regulator, preventing the oxidation of matters of a more stable character than itself, and that therefore, it was necessary to have it present in con- siderable excess. It was also found that the quantity of per- manganate consumed by a given amount of tannin varied materially, according to the rate at which the permanganate was added, rendering all attempts to fix a definite ratio illusory. Kathreiner, who very carefully investigated the method,t endea- voured to render the results more constant by prescribing a fixed time for the titration to occupy ; and von Schroeder pro- posed that the permanganate should be added I c.c. at a time at definite intervals of 5-10 seconds. It was afterwards shown by Procter || that what really made the difference was the speed of mixture, and that therefore the result was as much influenced by the vigour of the stirring as by the rate of addition, and that * Journ. Soc. Chem. Ind., 1884, p. 82. t Ibid., 1885, p. 263. \ Ding. Polyt. Journ., 1878, ccxxvii. p. 481. ' Bericht der Commission zur Feststellung einer einheitlichen Methode der Gerb- stoffbestimmung, Cassel, 1885. j| Journ. Soc. Chem. Ind., 1886, p. 79. Q 226 LEATHER INDUSTRIES LABORATORY BOOK. in this respect the " I c.c." method was much more easily affected than the ordinary " drop method." The explanation of the variation is a simple one. The oxidation in the Lowenthal process should be limited to indigo, and bodies more oxidisable than indigo, but there exist both ready formed in liquor, and among these oxidation products, many substances which, in the absence of indigo, will readily reduce permanganate. When the latter is added rapidly, and with insufficient stirring, it destroys the indigo and tannin in contact with it, and then proceeds also to oxidise the other matters present, although in other parts of the beaker indigo and tannin still exist. Thus more permanganate is reduced than corre- sponds to the indigo and tannin, and this is especially so towards the end of the process, when very little of either remains. The more slowly the permanganate is added, and the more vigor- ously it is stirred, the more closely it will approximate to the theoretical quantity required merely to oxidise the indigo and tannin. A much more valuable suggestion than that of the " i c.c." method was, however, made by von Schroeder,* which has since been universally adopted. It is obvious that if in every case the permanganate solution be standardised against one of pure tannin, under precisely the same conditions under which the analysis is made, any variations dependent on the rate of mixture will mutually compensate each other ; and this method would have been universally adopted, but for the impossibility of obtaining really pure tannin. Von Schroeder showed that the principal impurity in the purest commercial tannins is gallic acid, which is oxidised similarly to tannin in presence of indigo, but consumes slightly more permanganate, and therefore in- fluences but little the amount of permanganate consumed by such a tannin. According to his experiments, I grm. of the purest commercial tannins was equal in reducing power, on the average, to I 05 grm. of theoretically pure, and he therefore proposed to multiply the amount actually weighed out by this figure. Procter suggested the employment of pure crystallised gallic acid as a standard, which, perhaps, hardly received the attention it deserved, since it is oxidised in the same way as tannin in presence of indigo, and is easily dbtained in a pure * Loc. cit. See Note , p. 225. VOLUMETRIC METHODS OF ESTIMATING TANNINS. 227 state. -I grm. of crystallised gallic acid is equivalent to about 1*25 grm. of ''pure tannin," as calculated by von Schroeder's data, but further experiments are required to fix the quantity with complete accuracy. The values obtained by this method are only correct for gallotannic acid, and little is known of the proportionate con- sumption of permanganate by other tannins. For a long time, -a statement of Oser's that oak-bark tannin consumed I * 5 times as much permanganate as gallotannic acid was accepted as a basis of calculation, but after the introduction of the hide-powder gravimetric method it was found that the relation between the two varied irregularly, even for the same kind of material,* and if it is desirable to calculate even approximate percentages in a series of determinations of any other tannin than gallotannic acid, the only way is to determine a factor at the time, by a sufficient number of simultaneous gravimetric analyses (p. 21 2) of the same material. The working details now adopted in the Leeds University Laboratory are as follows. The solutions required are : 1. Pure potassium permanganate, 0*5 grm. per liter. As very weak solutions keep badly, it is best to make up one of 5 grm. per liter, and dilute when wanted. The exact strength of the permanganate is not important so long as it is constant through a series of experiments. 2. Pure indigo carmine (potass or sodium sulphindigotate), 5 grm., and concentrated H 2 SO 4 , 50 grm. per liter. This must be filtered, and should give a pure yellow free from any trace of brown where oxidised with permanganate; 25 c.c. of this solu- tion should equal about 30 c.c. of the permanganate, and, if necessary, must be diluted to that strength. As it is frequently difficult to obtain indigo carmine of satisfactory purity, the use of synthetic indigo, as suggested by Rawson, may be recommended. I grm. " indigo pure B.A.S.F. " is dissolved by heating to a temperature of 75-80 C. with 25 c.c. of sulphuric acid S.G. I '84 for I hour, stirring occasionally with .a glass rod. After cooling, 25 c.c. more acid are added, and the whole made up to I liter. 3. Solution of pure tannin, 3 grm. to I liter. Since absolutely pure tannin cannot be obtained, the follow- * 'Gerber,' 1887, p. 3. ^ Q 2 228 LEATHER INDUSTRIES LABORATORY BOOK. ing method is adopted : A sample of the purest obtainable tannin (not less than 90-95 per cent, pure by hide-powder) is preserved air-dry in a well-stoppered bottle, and the moisture carefully determined. The principal impurity is gallic acid, which acts on permanganate like tannin, but reduces somewhat more, and I part of such tannin, calculated to dry weight, is equal on the average to I "05 parts of pure tannin. Hence it is easy to calculate a quantity of the air-dry tannin equal in per- manganate value to 0*3 grm. of pure tannin, and this is weighed out when required and made up to 100 c.c. The moisture varies very little, but it is well occasionally to re-determine it and calculate afresh. Titration. 25 c.c. of the indigo-carmine solution is mixed in a beaker with about f liter of tap water, and the permanganate added drop by drop from a glass-tapped burette till a pure yellow is obtained, the liquid being stirred steadily the whole time. A disc stirrer or a glass rod bent several times back and forward, is to be preferred to a plain rod ; or some method of mechanical stirring may be adopted. The dropping should be always as nearly as possible at a similar rate for each experi- ment, and should be slower towards the end of the titration. It is convenient to keep a second beaker titrated to a pure primrose yellow as a standard test. Titrations may be accurately performed by artificial light, but usually differ slightly from those by daylight, and hence the light should not be varied in the course of an analysis. For daylight work Kathreiner recommends the use of a white basin instead of a beaker. The permanganate solution is allowed to drop in, with constant stirring, till the pure yellow liquid shows a faint pinkish rim, most clearly seen on the shaded side. This end reaction is of extraordinary delicacy, and is quite different to the pink caused by excess of permanganate, being an effect common to all pure yellow liquids. The titra- tion is done at least twice, and the average taken ; f liter of water and 25 c.c. of indigo are then taken as before, and 5 c.c. of the tannin solution are added and similarly titrated repeatedly. Deducting amount required for the indigo, the remainder is that consumed by the tannin, which should not at most exceed two-thirds of that required by the indigo. A similar titration is made with the tannin infusion to be examined,''of which such a number of cubic centimeters is employed as will consume about VOLUMETRIC METHODS OF ESTIMATING TANNINS. 229 the same quantity of the permanganate as the standard tannin solution. The value of the total astringent is then calculated in terms of tannin. Since tanning matters contain astringents which are not taken up by the hide, but which are oxidised by permanganate like tannins, it is in most cases necessary to remove the tannin from a portion of the infusion, and to repeat the titration to determine the non-tannin. Detannisation by Hide-Powder. This may be done by the^ same method as is used when the tannin substance is deter- mined gravimetrically This is desirable where liquors containing tannins other than gallotannic acid are to be estimated, since it is found that the weight calculated as " pure tannin " (gallotannic acid) is in many tanning materials very different from that determined gravimetrically. This fact may be taken advantage of in recognising the classes of tannins which are contained in an extract. If a gravimetric determination is made at the same time, a factor may be calculated which will give results on other samples of the same material closely approximate to those obtained directly by gravimetric analysis. Detannisation by Gelatine. Hunt's modification. This is generally applicable, and especially so to liquors containing gallic acid, of which it removes less than hide-powder. It is also much more rapid of execution than detannising by hide- powder. Solutions required. I. Pure gelatine, 2 grm. per 100 c.c. 2. Saturated solution of NaCl containing 50 c.c. of concen- trated H 2 SO 4 per liter. Precipitation. To 50 c.c. of the liquor (of about the strength required for the hide-powder process) is added 25 c.c. of the gelatin solution and 25 c.c. of the salt solution, and about a tea- spoonful of kaolin or baric sulphate, and the whole is well shaken for five minutes and filtered. This filtrate, which should be perfectly bright, is titrated for non-tannin by the Lowenthal method, double the volume being taken which was employed for determination of total astringents, and the result is deducted before calculating the tanning value. Calculation of Lb'iventhaPs Estimations. If only the tannin is required, the result is best obtained by a simple proportion sum If 3 grm. per liter of pure tannin consumes a c.c. of perman- 230 LEATHER INDUSTRIES LABORATORY BOOK. ganate, and the solution tested b c.c. (after deducting that consumed by the indigo and non-tannins), the solution will contain ^b\a grm. of tannin per liter. If any other strength than 3 grm. per liter has been used for the standard, or it has been determined by gravimetric analysis, the quantity used or found must be substituted for 3. In a similar way the non-tannin permanganate may be reckoned to its tannin value, and reduced to gallic acid, by division by 1*35, but the value will include .beside real gallic acid, all other easily oxidisable matters present in the solution. The non-tanning oxidisable matter of sumacs differs in many respects from pure gallic acid.* Estimation of Tannin by Jean's Method. F. Jean estimates tannins with a solution of iodine in presence of sodium bicar- bonate. The method is described in his " Industrie des Cuirs et Peaux," p. 116, and with some modifications and improve- ments in Wiss. Beilage der Ledermarkt I. (1899-1900) p. 113. The method is capable of giving good results, at least with gallo- tannic acid, but does not seem to have marked advantages in either accuracy or facility of execution over the Lowenthal method. The iodine solution is made by rubbing up 3*77 grm. of iodine in a mortar with a little water and 5 or 6 grm. of potassium iodide, and making up to I liter. Tannin and gallic acid solutions of I grm. per liter are used for standardizing. 10 c.c. of the tannin solution, and 5 c.c. of a saturated solution of sodium bicarbonate are introduced into a small beaker of 5 or 6 cm. in diameter, and with a mark to contain 50 cm. The iodine solution is now added drop by drop from a burette, con- stantly stirring the liquid with a glass rod, until about 9 cm. have been added. A drop of the liquid is then taken on a glass rod and spotted on to filter paper which has been rubbed over with starch, the spotting being so managed that only the drop and not the rod itself comes in contact with the paper. If iodine is present in excess a grey spot with blue edges will be produced upon the paper, and if this is not the case more iodine solution must be added until the blue edges appear. The glass is then filled up to the 50 c.c. mark with distilled water, and the spotting repeated. If no colour appears, the right quantity of iodine has * Commercial gelatine often contains sulphites and other matters which reduce permanganate, and which must be determined, if present, by a blank experiment. VOLUMETRIC METHODS OF ESTIMATING TANNINS. 231 not yet been exceeded, and 5 or 6 drops of iodine solution must be given before this occurs, the end of the titration being marked by the re-appearance of the blue ring. As the coloration is some- what fleeting, the first appearance of this ring must be taken as the end point. 10 to 10*5 c.c. of iodine solution should be required for the 10 c.c. of tannin solution, and if this is not approximately the case with pure tannin, the iodine solution must be diluted or strengthened as required. In working with gallic acid about 3 c.c. more of iodine solution will be required than for the same strength of tannin, and in titrating unknown solutions the quantity or strength of the solution must be varied so as to make the amount of iodine solution employed as nearly as possible the same as that with the pure tannin. A correction is required for the amount of iodine solution which must be added to 5 c.c. of sodium bicarbonate solution diluted to 50 c.c. with distilled water before the blue reaction is produced. This is determined by a blank experiment, and amounts approximately to 0*4 cm., which must in each case be deducted from the iodine solution used for titration. In order to separate tannin and gallic acid Jean employs egg-albumen, prepared by rubbing 2 grm. of dry albumen with a sufficient quantity of white glycerine of 28-30 Baume to form a thick paste. The mixture is allowed to stand for half an hour and then filled up with tepid water to the volume of a liter. A few pieces of camphor are added as an antiseptic, and the neck of the flask is closed with a plug of cotton wool wetted with a few drops of formalin. In order to precipitate the tannin, 50 c.c. of the tannin solution are introduced into a IOOC.G. flask. 15 c.c. of the albumen solution and 20 grm. of common salt are added, shaken up, filled up to the mark with water, and again shaken to complete the solution of the salt. The solution is filtered through a small filter, the first few centimeters being rejected and 20 c.c. (equal to 10 c.c. of the original solution) are run into a porcelain basin, slightly acidified with acetic acid and heated to boiling to precipitate the excess of albumen. It is then filtered through a smalRpleated filter, the precipitate well washed, and the filtrate and washings which amount to 30 or 35 cm. are titrated with iodine exactly in the method described. A further correction, which usually amounts to about 0-7 cm. is required for the iodine absorbed by the albumen, which is also determined by a blank experiment. Tannin materials which contain water- 232 LEATHER INDUSTRIES LABORATORY BOOK. soluble starchy compounds must be extracted with alcohol in order to avoid the influence of the starch. If sulphites are contained in extracts, which is now very commonly the case, they must be removed by acidification with hydrochloric acid and evapora- tion before making the estimation. The extraction of tanning materials is similar to that used for the gravimetric method, but the solution is only about one quarter the strength, and if tolerably clear does not require filtration, so that in many cases a simple procedure of extraction may be adopted. The Parker- Payne Method of Tannin Estimation. In 1904,* a new method for the analysis of tannin and tanning materials was described by Dr. J. G. Parker and Mr. E. E. Munro Payne, which excited very considerable interest, and which was based on the attractive idea of determining the total acidity of a liquor including the tannins, and afterwards detannising and estimating the tannins by difference. It is in a sense an extension of the Procter method for estimating acidity with lime water, p. 243, and depends on the precipitation of the tannins and allied bodies with a solution of sugar-lime or of barium hydroxide, filtering and titrating back the excess of lime or baryta with standard acid and phenolphthalein. The end-reaction is very sharp, and the process at first sight an extremely promising one, but un- fortunately on investigation sources of error appear which it is difficult or impossible to overcome. The details of the process as originally published are as follows : 300 c.c. of N/5 calcium hydroxide solution is added in a stoppered cylinder to 2OO c.c. of the clear liquor to be examined, which must contain approximately the same quantity of tannin as is required by the I. A. L. T. C. method for the hide filter (about 4 grm. per liter). The mixture is shaken repeatedly and allowed to stand for about four hours with occasional shaking, looc.c. are filtered off and the amount of alkali is titrated back in presence of phenolphthalein. This a'mount multiplied by 5 and deducted from the 300 c.c. of hydroxide solution employed gives the volume of N/5 calcium hydroxide solution which has been absorbed by 200 c.c. of the original liquor. This figure is termed by the authors of the paper the " total absorption value." Since this obviously includes the value of all acids present in the liquor, the authors proceed to remove tannin and 'to repeat the * Journ. Soc. Chem. Ind., 1904, p. 648; Collegium, 1904, p. 261. VOLUMETRIC METHODS OF ESTIMATING TANNINS. 233 determination. As a precipitant, a mixture of gelatin and gelatin-peptones is used, which the authors call " collin," and which is prepared as follows : 60 grm. of a good commercial gelatin, which it is stated must be produced from skin, is soaked in about 500 c.c. of distilled water, and warmed until dissolved. I2O c.c. of N/i caustic soda is added, and the mixture is heated on a water-bath for twenty minutes at 90 C., and is then filtered through linen. The cooled solution is made up to accurately 500 c.c., and 100 c.c. are titrated with N/i hydrochloric acid and phenolphthalein, and the remainder is accurately neutralised with a similar quantity of N/i acetic acid, I c.c. of pure chloroform is added as an antiseptic, and the neutralised portion is made up to a liter. The solution, when acidified by a dilute organic acid, is found to be a very delicate precipitant of tannic acid. To detannise the solution for analysis, 200 c.c. of the original tanning liquor is mixed with 100 c.c. of N/5 acetic acid, and IOO c.c. of the collin solution ; or in case of old acid tanyard liquors which contain but little tannin, with 50 c.c. of each. The solution is filtered, and 200 c.c. of the clear bright filtrate is treated with 200 c.c. of N/5 calcium hydroxide and allowed to stand for an hour, when it is titrated as before. Deducting the amount of hydroxide neutralised by the added acetic acid, the result gives the absorp- tion of the liquor minus the tannin, and of course the difference between this figure and the previous one corresponds to the tannin precipitated. It is not necessary to follow the authors in some theoretical considerations which they subsequently deduce with regard to so-called " colour weight," as these are founded on the radically false assumption that all tannins are simply mixtures owing their tanning value to the presence of digallic acid together with certain weight-giving substances. Much criticism * has been devoted to the suggested method of separating the tannic acid, and it has been pointed out that the gelatin-peptones present are uncertain in their composition owing to varying degrees of hydrolysis, and that they do not necessarily precipitate tannin in the same proportions as hide- fibre. These considerations are valid as applied to a proposed method of estimating the tanning value of the solution by weighing the precipitate produced by collin, but the mixture * Wood, Journ. Soc. Ch. Ind., 1904, p. 1071 ; Collegium, 1905, p. 40. 234 LEATHER INDUSTRIES LABORATORY BOOK. may be taken as at least a fairly good material for detannising, and the authors have suggested several other substances, and especially some of the alkaloids which may be substituted. The actual foundation of the method, however, the precipitation with calcium or barium hydroxide, rather curiously escaped criticism until it was examined in the Leeds University Laboratory,* and the assumption that it was based on the formation of a definite metallic compound had not been questioned. Very careful experiments, however, by Procter and Bennett, extending over several months, have shown that not only is the precipitate with gallic acid, and probably with most other astringent bodies, partially soluble in the precipitant, and having an alkaline reac- tion to phenolphthalein is estimated as a portion of the excess of calcium hydroxide, but that variable quantities of lime or baryta are carried down together with the precipitate, so that with known substances such as gallic acid, exact theoretical results are only obtained by accident. In the case of proto- catechuic acid (the acid corresponding to gallic among the catechol tannins), which is quite a possible constituent of tanning liquors, the precipitate is wholly soluble. The solution containing these partially dissolved compounds when exposed to air absorbs oxygen with great avidity, forming dark coloured precipitates, which carry down with them lime- or baryta-salts, and so in- crease the apparent absorption to an unknown extent. Sugar also apparently introduces unexplained chemical complications, and if the amount is increased leads to the production of a yellow and oxidisable solution ; though with the quantities prescribed, the sugar-lime solution, which is preferred by the authors to barium hydroxide, is certainly the easier of the two to handle. It was hoped by Procter and Bennett that, employing barium hydroxide, and operating under strict exclusion of oxygen, more reliable results would be obtained ; and it was found that the addition of some 20 per cent, of alcohol completely stopped the solubility of the precipitate in the case of gallic acid, but this unfortunately only brought into prominence the fact that variable quantities of the precipitant were carried down with the insoluble gallate, so that the results varied with every change of small details in quantities and manipulation, though they approximated to the theoretical result that all three hydroxyfe of the gallic acid were saturated as well as the carboxyl group. * Journ. Soc. Ch. Ind., 1906, p. 251. VOLUMETRIC METHODS OF ESTIMATING TANNINS. 235 Turning to gallotannic (digallic) acid it was found that, beside the difficulties arising from the gallic acid always present, the results varied through the hydrolysis of the compound. If the precipitated solution was filtered and titrated at once, the results approximated to the saturation of six hydroxyls with the base. Hydrolysis, however, at once set in, and the absorption gradually increased, so that no definite results could be obtained until the solution had been heated on the water-bath for an hour or more, when apparently constant results were obtained, which approxi- mated to those of 2 molecules of gallic acid, combining with 8 equivalents of base. The investigation of some other tannins was begun, but was not carried far, as the difficulties of obtain- ing reliable results with these simplest and best known forms appeared to be insuperable, and the reader must be referred for details to the original paper.* It was stated by the authors that the " total absorption value "" obtained by their method exactly agreed with the results of the ordinary hide-powder analysis. It is most obvious that any such agreement could only be the result of accident, as the basis of calculation was wholly arbitrary, and, as we unfortunately know, the results of various hide-powder methods differ among them- selves. The matter, was, however, investigated by Boegh f, who- showed that actually in the case of the tannins of the pyrogallol group, very fair concordances were obtained, but that known catechol tannins gave results corresponding to only about two- thirds of the weight actually found. This observation is interest- ing, and is probably explained by the fact that, while pyrogallol has three hydroxyls capable of saturation, catechol has only two ; and the reader may be reminded of the similar relation which exists between the two classes of tannins when estimated by the Lowenthal method. It is needless to point out that such a fact involves the determination of a special factor for each individual tannin, and is a complete explanation of the so-called " colour weight." It is unfortunate that so promising a method should prove unreliable, but until radical improvements are made, the reader cannot be too strongly warned against placing any implicit con- fidence in its indications. * Loc, cit. T Nordisk Garvertidende, 1904 ; Collegium, 1904, p. 301. 236 {LEATHER INDUSTRIES LABORATORY BOOK. SECTION XV. ESTIMATION OF COLOUR IN TANNING MATERIALS, ETC* THIS is often of considerable importance in the case of ex- tracts, in which contracts are often made for deliveries equal to the buying sample in colour. It is also useful in estimating the discoloration caused by iron and other impurities. Many forms of colorimeter have been proposed which depend on matching varying thicknesses or dilutions of a standard colour solution against the colour to be measured, but though these render excellent service where the colours are constant and identical, as in nesslerising, and the colorimetric estimation of iron, dyestufifs, etc., they are unsuitable where not only the intensity, but the actual tint varies in the colours to be registered. Spec- trometric methods are no doubt theoretically the most perfect, but at present are too complex and not sufficiently deli- cate for practical use, and the colour estima- tion of the solution is FIG. 16. made therefore best by Lovibond's tintometer. This instrument is shown in Fig. 16, and consists of an oblong box, at the upper end G of which are eye-holes for one or both eyes, while at the lower is a -screen with two square openings, behind one of which can be placed a rectangular glass cell of known thick- ness, while behind the other is a grooved rack J in which slips of coloured glass can be inserted to match the colour of the liquid. These glasses are red, yellow, and blue, of tints carefully graduated in units and decimals of colour, and so adjusted that equal numerical quantities of the three colours will always produce a neutral tint. By suitable combinations of these,, any colour can be imitated which is not brighter than that of the combined glasses, * Parker and Procter, Journ. Soc. Ch. Ind., 1895, p. 124. ESTIMATION OF COLOUR IN TANNING MATERIALS. 237 in which case a certain amount of neutral tint must be added to the colour to be matched, and noted in the result as a minus quantity. This case, however, never practically occurs in the matching of tanning materials. It is found that a set of glasses ranging from o* I to 20 in red and yellow, and from cr I to 10 in blue, is sufficient to match all tanning liquors. The ease of matching greatly depends on practice and a trained eye, and some observers never succeed, owing to defective colour vision. Match- ing is often greatly facilitated by the use of a coloured glass over the eyehole of the instrument (blue in the case of tanning liquors). The match should remain unaltered when this glass is removed. Light is most conveniently reflected into the instrument from a sheet of opal glass, illuminated with good daylight, and satisfac- tory measurements cannot be made by artificial light, or in very dark or foggy weather. The colours of opaque objects may also be measured by placing them in front of one of the openings of the instrument on the opal glass, and adding glasses to the white to effect a match ; or the difference between a dyed and undyed fabric can be measured by placing one before each aperture in equal light, and compensating with coloured glasses. The colour is usually calculated on a solution containing J per cent, of tanning matter, as determined by hide-powder analysis, measured in a J-inch cell. It is found that within the limits of the experiment the colour varies proportionately to the strength of the solution and the thickness of the layer of liquid ; so that, if a be the percentage of tanning matter in the solution em- ployed, and b the thickness of the cell used, the value of each colour for per cent, solution in a J-inch cell may be found by multiplying the observed value by the factor 0*5 x 0-5 a x b The most convenient strength for working with is found to be usually J per cent, solution, measured in a J-inch cell ; but for solutions of sumach this is too weak, and for hemlock extracts and materials rich in colour a weaker solution or a thinner cell is preferable. In most cases the solution used in an analysis for tanning, usually containing about 0*4 per cent, of tanning matter, may be employed, and the results calculated to ^ per cent, solution at will. Results may be calculated from \ inch to I cm. 233 LEATHER INDUSTRIES LABORATORY BOOK. cell with sufficient accuracy by multiplication by 0*8, or more accurately 0*79.* The following method is adopted in the case of a tanning extract. The proper quantity of the extract, calculated accord- ing to analysis, is weighed into a 500 c.c. flask, and made up to 500 c.c. with distilled water as nearly boiling as possible (about a grm. of phenol having been previously added to prevent fer- mentation and consequent change of colour). The solution is allowed to cool, and is then carefully filtered. The filtrate must be perfectly clear or accurate measurements cannot be made. If this cannot be accomplished by ordinary means, the addition of a small quantity of kaolin is often beneficial, and has no appreciable effect on the colour. If there is the least turbidity it is most important that the cell should be shaded from diffused light, by covering all but the front with black card or velvet. It is found that where dilution is necessary the chance of error is greatly reduced by making two or more measurements, using in ach case different strengths or cells, e.g. measuring \ per cent, and J per cent solution first in a J-inch cell, and afterwards in a i -inch cell, and reducing the mean of the result to one strength. By this means it is possible to match most materials to within one-tenth of a Lovibond degree. The measurement is made by placing the standard red, yellow, and blue glasses in the instrument, till the tint of the extract solution is exactly matched ; and the values of the respective colours are read off from the numbers marked on the glasses. As, however, a combination of any equal value of all three colours makes grey or black, the smallest colour-value always combines with equal quantities of the two others to form black. It is therefore entered as black, and an amount equal to it is deducted from each of the two larger before entering. Thus if the reading were 7 red, 14 yellow and I blue, the result would be entered as 6 red, 13 yellow and I black. As light is -lost by reflection from the surfaces of the standard glasses, an equal number of plain glasses must be placed with the cell, the -cell itself being counted as two, though when filled with liquid this is not strictly correct. It is obvious that tintometer * The rule given, while sufficiently near for technical purposes and with small variations, is not found to hold good, if the difference of dilution or of thickness is very large. ESTIMATION OF COLOUR IN TANNING MATERIALS. 239 measurement may be applied to a variety of other purposes in the tannery beside the mere matching of extracts, and especially the examination of oils and the registration of the colours of dyes may be named. Since the animal skin has different affinity for the different colouring matters contained in tanning materials, the colour given to leather is only proportionate to the colour of the solu- tion where the materials are identical ; and it is often desirable to confirm the results of the tintometer by actual tanning experi- ments. The most suitable material for this purpose is calf or sheep grain, which is obtained from the tanner in the limed state after splitting and " scudding " on the beam. The skin is well washed with water, and then with a solution of 2 per cent, boric acid, and I per cent, pure phenol. After some hours in this solution, it is " scudded " on the grain side, on a glass plate, with a blunt brass sleeker or a smooth-edged slate, to remove lime soaps, and is then preserved till required in a fresh boro-phenol solution, in which it will keep for some months. While this method of preparation answers well if properly carried out and the boro-phenol solution frequently changed, trouble is often experienced from the gradual deterioration of the skin. Apparently satisfactory results have been obtained with skins preserved with formaldehyde, and also by pickling with salt and a little formic acid, but it is obvious that if comparable colours are required, some one method must be adhered to. For tanning, a " rnilk-shaker " (p. 202), or the glass churn shown in Fig. 17, is generally used. Similar results may be obtained by suspension, but in a much longer time, about 36 hours being required for tanning ; and in the earlier stages the skin must be frequently moved or shaken. Strips of the delimed sheep grain, about 8 inches by 4 inches are cut, and, to remove the boric and phenol solution, are washed in the churn twice by shaking with distilled water, and lightly pressed out with the slicker. They are then put into the churn with a solution from the material to be used containing \ per cent, of tanning matter, and agitated for 20 minutes, after which the solution is strengthened to \ per cent., and at the end of another 20 minutes to I per cent, strength, and the skin is shaken with this for i^hour the tanning process thus lasting two hours. 240 LEATHER INDUSTRIES LABORATORY BOOK. The leather is then washed for a few minutes in distilled water, slicked out on a glass plate with a glass or brass slicker, lightly oiled with a drop of cod-liver oil, and pinned up in a dark place until dry. This method is satisfactory, the leather drying an even colour, and repeated tests of the same material giving concordant results. The exact strength of the solution, or the length of time employed in tanning, has no appreciable effect on the colour. Dr. Lepetit informs the author in a private letter that he tans pelt (calf or sheep) with extract equal to about 2O per cent, of tannin on its wet weight, with 400 per cent, of water. The whole extract is added to the water at once ; and the tannage lasts 12 hours. The leather is pressed through a rubber wringing machine before drying. The colour is sometimes lighter than if done in the ordinary way. If split calf pelt cannot easily be obtained FIG. 17. thin skins may be shaved with the bate-shaver's knife on the beam to the required thickness. The time required for tannage is somewhat longer than for sheep grain. It is obvious that the above method of experiment may be employed for many purposes beside the mere testing of extracts and other tanning materials ; as, for instance, the investigation of the effect on colour of " assistants " and additions to the tanning bath, of various methods of leaching, and of different modes of bating and preparing the skin. 241 SECTION XVI. ANALYSIS OF USED LIQUORS, AND SPENT TANS. Total Solids. This is only important if it be desired indirectly to determine suspended matter. The liquor is well shaken and 50 c.c., or less, is measured into a basin, evaporated to dryness and dried till constant at 105 to 110 C. In most cases the liquor may at once be filtered, either through a Berkefeld candle (p. 187), or through paper, using kaolin if necessary to get it perfectly clear. The specific gravity should always be taken with a pycnometer or hydrometer. Barkometer degrees correspond to the second and third deci- mals of specific gravity ; thus sp. gr. I -065 = 65 Bkr. Total Dissolved Matter. 50 c.c., or a smaller quantity of strong liquor, is evaporated to dryness in a weighed basin, dried till constant at 100 to 105 C., and weighed. In stronger layer liquors, 5 or 10 c.c. is quite sufficient. Difficulty is likely to arise from the presence of partially volatile substances, and especially of lactic acid. Suspended Matter is determined by subtracting the " dissolved matter" from the "total solids." This calculation is only reliable in dilute solutions, where the volume of the suspended matter may be neglected. In other cases the suspended matter may be collected on a tared filter, weighed wet with the filter and again when dry, and the weight of dissolved matter calculated corresponding to the loss of water, and deducted from the net weight. Ash and Organic Matter. The residue of the " total soluble " (or " total solids ") is cautiously incinerated, moistened with ammonium carbonate solution, evaporated to dryness, and cau- tiously heated to incipient redness to drive off ammonium salts without decomposing carbonates. The loss is total " organic matter," the residue " mineral constituents," which consist mainly of lime, and may be further analysed by the methods given for the mineral matters of water, p. 45. The ash may be dissolved in excess of N/i HC1, and titrated back, when bases found will have existed in the liquors as salts of organic acids. These neutral salts have very considerable influence on the tanning R 242 LEATHER INDUSTRIES LABORATORY BOOK. process, and appear to be one of the principal causes of the "mellow" action of old liquors. The determination of iron in the ash is frequently of great importance, and may be made colorimetrically, p. 40. Tanning Matter. A portion of the filtered liquor is diluted, or concentrated, till it contains about 4 grm. of tanning matter per liter, and is analysed as described in Section XIII. by the chrome-chloride method. Where very large quantities of non- tans are present, it is often necessary to reduce the tannins below the official strength of 4 grm. per liter. The residues of " total soluble " and " non-tannins " may be redissolved in water, and the acid in each determined (see below), and the difference, reckoned as lactic or gallic acid, as the case may be, deducted, from the tanning matter found. Where the proportion of tannin is small, great accuracy is impossible from the disturbing action of non-tannins. (Cp. p. 208.) The results of analysis of liquors may be stated in grm., either per 100 grm. (per cent.), or per 100 c.c. (volume percent.) ; but it must be made clear which is used, and the density of the liquor must always be given to render it possible to calculate from one to the other. The Lowenthal method may also be used for tannin estima- tion, especially for weak and acid liquors, or where a series of determinations are required. The tannin may be reckoned in terms of gallotannic acid, when it will be always lower than the " tanning matter " determined by hide-powder ; or a factor may be obtained by simultaneous hide-powder determinations of one or more of the liquors. The Parker-Payne method has been strongly recommended, but the reader is referred to the criticism in Section XIV. Determination of Free Acids in liquors is of great importance, both on account of their power of swelling and of removing lime, but no thoroughly satisfactory method has yet been pro- posed, largely owing to the impossibility of defining what is really meant by the term. Liquors may contain acids of all degrees of strength, from actual mineral acids to phenolic bodies exerting a feeble acid function, and the solution which may appear perfectly neutral to one indicator, such as methyl orange or congo red, may be strongly acid to another suqh as phenolph- thalein. The question is further complicated by the fact that ANALYSIS OF USED LIQUORS, AND SPENT TANS. 243 the tannins themselves seem to possess feeble acidity, and that means such as gelatin, hide-powder, or animal charcoal used to remove them generally absorb or destroy also a portion of the other acids present. As a rule, any of the methods described below will give results of approximate accuracy if applied to determine a known quantity of one of the stronger acids added to a distinctly acid liquor of which the acidity has been previously determined, but naturally the presence of lime in combination with phenolic bodies may render determination abortive, and in any case no two methods can be compared if they employ different indicators and consequently estimate different acids.* Determination of Total Free Acids (Procter]. Measure 10 c.c. of perfectly clear liquor into a beaker, and add clear saturated lime water from a burette till a trace of permanent cloudiness is produced. Repeat several times, and take the average. The beginning of turbidity is best seen by looking through the liquid at printed paper. For most tannery purposes it is sufficient to merely note the c.c. of saturated lime water consumed for 10 c.c. of liquor, but it may be standardised against 10 c.c. of decinormal HC1, with phenolphthalein or methyl orange, and the acidity of the liquors calculated as equal to grin, acetic or lactic acid per liter ; N/io acid being equal to 6 grm. acetic and 9 grm. lactic acid. If oxalic or free sulphuric acid is present, turbidity may be produced on the first addition of lime water. In this case, add excess of neutral calcium chloride to the liquor, make up to double its original volume, filter, and use 20 c.c. instead of 10 c.c. Pine-bark liquors are said to produce a similar turbidity, which may perhaps be removed in the same way. Half the free carbonic acid is estimated by this method, but may be removed previously by adding salt, and shaking ; the liquor must in this case be made up to a definite volume before titrating. (Simand.) Gallic acid behaves approxi- mately as a monobasic acid. Determination of Free Acids, Koch's Method^ as improved by Paessler and Spanjer.\ This method depends on the removal of tannin by precipitation with gelatin, of which a solution, 2 grm. per liter, is used. 25 c.c. of the filtered liquor is mixed * Cp. Bennett and Wilkinson, Journ. Soc. Chem. Ind., 1907, p. 1186. t Dingl. Polyt. Journ., 1887, cclxiv. p. 395. \ Deutsche Gerber-Zeitung, 1899, Ixxvi., Ixxvii. ; 1900, xlv., 1. to liii., lv., cliii. to clvi. ; Collegium, 1903, pp. 10, 17. R 2 244 LEATHER INDUSTRIES LABORATORY BOOK. with 25 c.c. of the gelatin solution ; or in liquors containing much tannin with a larger quantity, or a somewhat stronger gelatin solution may be used. After thoroughly mixing, the unfiltered liquid is titrated with N/io sodium hydrate solution, or with baryta or lime water which has been standardised against acetic acid, withdrawing after each addition a small drop on a glass rod and spotting it upon azolitmin paper. The end of the reaction is when a blue edge appears round the spot. Should this point have been passed the titration is repeated. The acidity is calculated in terms of acetic acid per 100 c.c. or per liter of the original liquor. As some gelatins have an acid reaction, it is desirable to do a blank with the gelatin solution employed, and to deduct from the amount of lime or baryta water used the quantity which is required for the gelatin alone. All acids which are capable of reddening azolitmin paper are esti- mated by this method, except the tannins which combine with the gelatin. Carbonic acid is almost without effect if the end reaction is taken when the paper ceases to be brightly reddened, but will be included if the titration is taken to a full blue. The authors quote experiments made with known added quantities of acetic and lactic acids in which the results do not vary from the truth by more than a few milligrams of acid per 100 c.c. Most if not all acids which redden azolitmin will swell hide, and consequently, the method may be taken as a measure of the swelling acids present, though their actual swelling power is sometimes destroyed by the presence of salts or putrid ferments. Procter's method, on the other hand, gives all those acids which will dissolve lime, and consequently includes weak acids which have soluble lime-salts, but practically no swelling power. The choice between the two methods, therefore, depends on the object of the determination. Determination of Total Free Acids (Kohnstein and Simand)* This method is laborious and of doubtful accuracy. 100 c.c. of the liquor is heated to boiling with 3 to 4 grm. of pure freshly ignited magnesia free from lime. The liquid is filtered, and in the clear and almost colourless filtrate the magnesia is estimated gravimetrically (after removing lime, if present, with ammonium oxalate). The total acid in terms of acetic acid is calculated from the magnesia so found, by multiplying 'its weight by 3 ; Dingl. Polyt. Journ., 1885, cclvi. pp. 38, 84. ANALYSIS OF USED LIQUORS, AND SPENT TANS. 245 (MgO = 40, 2C 2 H 4 O 2 = 1 20). Sulphuric acid may also be es- timated by evaporating a portion of the filtered magnesia solution to dryness, igniting, moistening with carbonic acid water, drying, and washing on a filter, when any magnesium sulphate will pass through, while magnesium carbonate or oxide corre- sponding to the organic acids will remain on the filter. If the liquors naturally contain magnesium salts, these must be previ- ously determined and allowed for in calculating the acids. (Cp. p. 45.) The method obviously estimates all acids which form soluble magnesia salts, and the salts of such acids which are decomposed by magnesia. Probable sources of error are the instability of the magnesium tannates, and the tendency to form basic salts. Simand * also gives the following method. 50 c.c. of the liquor, which must not exceed 1-004-1-005 S.G. (4-5 Bkr.), is boiled for five minutes with return-flow condenser with 5 grm. of pure freshly ignited animal charcoal, which has been freed from mineral constituents by digestion with acid and washing. After cooling, the condenser and flask are rinsed out on to a filter, and the animal charcoal washed with boiling water till the filtrate amounts to nearly 500 c.c., and after heating on the water bath to drive off carbonic acid, and cooling, is made up to exactly 500 c.c. Of this liquid, 200 c.c. is titrated with N/io sodium hydrate, or lime water and phenolphthalein. Minute traces of acid are retained by the charcoal, and for very exact analyses this must be determined by treatment of a known acid solution in a similar manner and allowed for, but the amount is so small that it may generally be neglected. This method has been adopted by the American Leather Chemists' Association, who dilute liquors I : 5 and use 2 grm. animal charcoal to 100 c.c., titrating an aliquot part ; but it has not proved very satisfac- tory. Hoppenstedtf has suggested a method in which the tannin is precipitated by alcoholic solution of quinine, for which he claims good results, which have not been wholly confirmed. The determination of free acid is of great practical import- ance, especially in sole-leather tanning, where a deficiency invariably leads to discoloration, which is sometimes not notice- able till the leather is dried. * Bockmann's Chemisch-Technische Untersuchungsmethode, 552. t Journ. Am. L.C.A., 1906, pp. 192, 221. 246 LEATHER INDUSTRIES LABORATORY BOOK. Carbonic acid only exists in liquors in which fermentation of carbohydrates is going on, and is not found in putrid liquors. It is also present in drenches. It may be determined by boiling IOO c.c. of the fresh unfiltered liquor in a flask connected with a U-tube or other absorption apparatus containing an ammoniacal solution of barium chloride, which is afterwards rapidly filtered with as little exposure to the air as possible, washed with boiling distilled water, dried at IO5C., and weighed. Simand found an average of about 0-15 grm. per 100 c.c. in sole-leather colouring liquors. Volatile Acids (Acetic, etc.). These may be determined by distilling IOO c.c. of the liquor in a tubulated retort with a good condenser to small volume, say 25 c.c., titrating the distillate with lime water or N/io caustic soda and phenolphthalein, adding ioo c.c. distilled water to the retort, and repeating the operation so long as any material quantity of acid comes over. The quan- tity so found may be deducted from the total, and the remainder calculated into lactic acid (2 of acetic equal 3 of lactic acid). In sumach, myrobalan, or divi-divi liquors, the fixed acid is mainly gallic, of which N/io soda is equal to about 17 grm. per liter when phenolphthalein is used, but the end-point is very un- certain. If it be desired to determine the volatile acids present in combination with bases, slight excess of phosphoric acid (free from volatile acids)* may be added to the residue in the retort, and the distillation repeated. The separation from lactic acid is not quite exact, some lactic acid being carried over. Microscopic Examination. The liquor is allowed to settle best in a conical glass and the sediment, as well as any scum is examined under a thin cover-glass, first with a low, and then with the highest power at command, and the forms and sizes of bacteria and ferments noted. Occasionally crystals (needles) of catechin may be noted if gambier is largely used. For more complete bacteriological examination, cp. Section XXVI. Viscous Fermentation, probably due to various bacillus fer- ments, produces a " ropy " or treacly condition of liquor, from the formation of gummy matter, which may be separated and estimated by the method given for dextrin in the examina- tion of leather. Valonia liquors are most liable, to this disease. * Allen, 'Commercial Organic Analysis,' i. 383. 2nd Ed. ANALYSIS OF USED LIQUORS, AND SPENT TANS. 247 (Cp. p. 435). It must not be forgotten that lime salts of tannins are also precipitated by strong alcohol. Valuation of Liquors. The chemist is often called upon to make an estimate of the value of liquors in a tannery, but so many circumstances are involved that at most a somewhat rough approximation can be given. It is generally out of the question to analyse each liquor individually, and it is sufficient to deter- mine the amount of " tanning matter " per barkometer degree in a few different stages of the process, and calculate the value of the other liquors from their barkometer strength. It will be found that the tanning strength diminishes much more rapidly than the barkometer degree as the liquors become weaker, because only the tanning matter is absorbed by the skin, while the non-tannins accumulate. The proportion of the two may be approximately estimated from the hide-powder analysis. The different tanning value of liquors in different parts of the yard will be best adjusted by deducting a fixed number of barkometer degrees in each case before estimating value. Thus if a layer liquor of 50 barkometer contains 4 per cent, of tanning matter, and a handler liquor of 20 only I per cent., we may deduct 10 throughout, and say that each degree above 10 is equal to C'l per cent, of tannin. If the analysis and value of the materials are known, it is easy to determine the cost value of I percent, of tanning matter ; remembering that carriage, and with solid materials cost of grinding, must be added to the original price, and that the tanning value of materials shown by analysis must be diminished by an allowance for loss in leaching. If the actual cost or value of the materials is not known, average strengths and market values must be taken. The weight of the liquors may be conveniently taken at 10 Ib. per gallon, or 62*25 Ib. per cube foot, and this will be strictly accurate if the liquors are measured instead of weighed for analysis. Pits containing hides and dusting materials do not usually contain more liquor than two-thirds of their total capacity. Value of dust may generally be included with that of the liquor. For stock-taking purposes the value of all material in the leaches may be taken at half-cost, but this is generally an over-estimate. The Valuation of Tannage of leather in process of manu- facture is very difficult, but for stock-taking it is convenient to estimate it in terms of the liquor which has been used in the 248 LEATHER INDUSTRIES LABORATORY BOOK. process. Thus, if the liquors in a handler shift fall from 3 per cent, to I per cent, of tanning matter, and one fresh liquor is given for each pack of goods, the pack will gain the value of one liquor of 2 per cent, in passing through the shift. Where dust is used its value must be added. For most stock-taking pur- poses it is sufficient, in a systematically worked yard, to estimate such a value once for all for each stage of the process, and adhere to it so long as the method of work is unaltered. The difference of cost of tanning materials from year to year is, of course, taken into account in costing the liquor as already described. Labour should be separately allowed for. Unless the value of tannage during manufacture is properly taken into account, any attempt at costing by setting annual expenses against production is illusory, except over a long average of years. The degree of tannage actually attained can of course be accurately determined by Kjeldahling. For the complete analysis of leather see Section XXII. Spent Tanning Materials must generally be dried and ground and extracted like other solid materials, somewhat large quan- tities being used, and the weak liquors concentrated by boiling. For a series of determinations it is often convenient to use the Lowenthal method, which gives good comparative results. In this case concentration is unnecessary. It is generally most convenient to calculate the percentage on the dry substance, but in comparing the results with the original material, the loss of weight in leaching and drying must not be overlooked. If the composition of the unleached material is known, the original weight of the leached sample may be calcu- lated from the insoluble matter present. Thus, if the original material contained 45 per cent, of insoluble matter, and the spent material 83 per cent, the tannin found in the latter would be multiplied by ff rds, to reduce it to that of an equal original weight. This correction is of the greater importance the stronger the material. Where the actual analysis of the original material is unknown, the correction must be calculated from that of an average sample. 249 SECTION t XVII. ANALYSIS OF THE MATERIALS USED IN ALUM AND CHROME TANNAGES. Alum Tannages. In these the principal materials employed are salt, with ammonia or potash alum, in conjunction with which flour and egg-yolk are frequently used as fillings. Recently, sulphate of alumina has largely superseded alum, of which it is the only useful constituent, since improved materials and process of manufacture have enabled its production in a practically pure state, and free from iron. In the manufacture of those leathers where colour is of secondary importance, less pure forms (such as " aluminoferric ") containing traces of iron are equally suitable, if there is any material advantage in price. The value of these materials is principally dependent on the amount of alumina present, and on the absence of iron as an impurity. It is also sometimes necessary to determine the acidity, but this is usually of minor importance, except in the examination of used liquors for the study of the tawing process. Alumina is A1 2 O 3 , and all the salts follow the same type, the chloride being A1C1 3 (or A1 2 C1 6 ), the sulphate A1 2 (SO 4 ) 3 , and so on. In addition to normal salts, alumina is very prone to form " basic " salts, which may be viewed as compounds of the normal salts with additional quantities of aluminium hydrate* The " basic chlorides " are thus really oxychlorides, of which several exist. Such basic salts are formed by dissolving freshly precipitated alumina in solutions of the normal salts, or by neutral- ising a portion of their acid with an alkali. The greater the basicity, the less stable is the compound formed : thus the more basic salts are readily precipitated by boiling, or in some cases, even by diluting their solutions. This instability renders them more easily taken from solutions by the hide fibre than the normal salts, and hence their use in tanning, and in mordanting fabrics. It is in many cases uncertain whether the basic salt is absorbed by the fibre as a whole, or whether the base and the acid are fixed by the carboxylic and amido groups indepen- dently, leaving a more or less basic salt in solution, or whether 250 LEATHER INDUSTRIES LABORATORY BOOK. the process is not one of adsorption rather than chemical com- bination. What has been said in this respect of alumina salts is also applicable to those of iron and chromium. Determination of Alumina. In most cases where other bases are present, gravimetric methods are most satisfactory. To the nearly boiling, and somewhat dilute solution, ammonium chloride is added to prevent possible precipitation of traces of lime, and then ammonia in very slight excess, and the boiling is continued till the free ammonia is almost completely expelled. The oper- ation is best performed in a platinum or porcelain basin, as glass is apt to be somewhat attacked by the hot ammoniacal solution. The precipitate should be washed several times by decantation, and then thoroughly on the filter with boiling water (p. 12). For accurate work, after removing most of the residual solution, the precipitate should be redissolved off the filter with dilute hydro- chloric acid, and reprecipitated as before and returned to the same filter, thus removing traces of co-precipitated bases. If the analysis is to extend to these, the washings and filtrate from the second precipitation should be united with those from the first. If a suction or vacuum filter is available, its use is con- venient, as the precipitate is bulky and gelatinous, and is so far freed from water by suction that the still damp precipitate may be ignited with care without further drying. If the vacuum filter is not used, the precipitate must be thoroughly dried, and transferred with its paper to a platinum or porcelain crucible, where it must be ignited at first very cautiously with the cover on to avoid loss by splintering, and then very strongly over the blowpipe. If the alumina salt has been a sulphate, this strong ignition must be continued for five or ten minutes, to drive off the last traces of sulphuric acid. As the precipitate is difficult to wash, small quantities of substance (cr 5-1 'O grin.) only should be used. If iron is present, it is best to precipitate it as Fe 2 O 3 along with the alumina, and afterwards to estimate its amount either in a separate portion of the solution, or after dissolving the very insoluble ignited precipitate by digestion with sulphuric acid as described on p. 364. For very small quantities the colorimetric method (p. 40) is most satisfactory, while larger amounts are determined volumetrically. A very accurate and convenient method of estimating small ANALYSIS OF ALUM AND CHROME. 251 quantities of iron in the ferric state, unaffected by the presence of salts of aluminium or chromium, is that of employing it to liberate iodine from a solution of potassium iodide, and titrating the iodine with sodium thiosulphate and starch. The iron must be wholly in the ferric state, but may be sulphate or chloride, and the solution must not contain free chlorine, bromine, or nitric or chromic acids. If the iron oxide has been dissolved in concen- trated hydrochloric acid after ignition, the liquid must be diluted and well boiled to expel any chlorine evolved. If the iron is not wholly ferric it may be oxidised by the addition of bromine -water, and subsequently boiled to expel all free bromine. A portion of the solution, preferably about 10 c.c. and containing as nearly as possible o* I grm. of Fe is placed in a bottle with a well-fitting stopper, and 30 c.c. of a 10 per cent, solution of potassium iodide is added. It is generally directed that the bottle should be heated to 5O-6o C. for 15-20 minutes, in which case the stopper must be securely tied or wired in, but in the author's laboratory no increased result was obtained by heating over that got by allowing the mixture to stand for half an hour at laboratory temperature. The contents of the bottle are now rinsed into a beaker with distilled water and titrated with N/io thiosulphate solution in the ordinary way (cp. pp. 258 and 306). Each c.c. of N/io thiosulphate corresponds to 0*0056 grm. of Fe or 0*008 grm. Fe 2 O 3 , or the thiosulphate may be standardised directly against ferric solution of known strength. As hydro- chloric acid frequently contains traces of chlorine, and potassium iodide of iodate, both of which liberate iodine, a blank determina- tion should be made : and the thiosulphate used (if any) deducted before calculating the iron. According to Fresenius the iron solution should be nearly neutral, but in the writer's experience the result is more accurate if it is freely acid. If neutral, there- fore, 20 c.c. of N/i HC1 may be added to the mixture, and even large excess produces no ill effect. A method more suitable for larger quantities and especially when the iron is mainly in ferrous state, is that of titration with permanganate after reduction by zinc to the ferrous state. For this purpose a measured quantity ot the solution, which should not contain more than about o*l or 0*2 grm. of iron, is trans- ferred to a flask, strongly acidified with pure sulphuric acid, and the neck of the flask is closed with a cork through which passes 252 LEATHER INDUSTRIES LABORATORY BOOK. a short tube drawn to a capillary opening to permit the escape of the hydrogen evolved, or preferably with two tubes, through one of which a slow current of carbon dioxide (or coal-gas) is passed into the flask to expel oxygen. The solution is warmed, and theri fragments of zinc (free from iron *) are dropped in, together with a small piece of platinum foil or a few drops of platinic chloride to set up electrolytic action, until the solution becomes colourless, and a small drop withdrawn on the end of a glass rod produces no redness with potassium thiocyanate solu- tion. In place of zinc, coarsely powdered magnesium may be used. The heating is continued till the zinc is entirely dissolved, and the solution is titrated with N/IO potassium permanganate solution (3*16 grm. per liter), till a faint pink coloration is pro- duced. Each c.c. of permanganate corresponds to 0*0056 grm. of Fe, or 0*008 of Fe 2 O 3 . Dilute permanganate solution does not keep well, but a N/i solution remains unchanged for a considerable time, and may be diluted as required. Permanga- nate solutions must always be used with a glass-tapped burette. If the strength of the permanganate is unknown, a check experi- ment is made by dissolving about o* I grm. of the fine soft wire used for binding flowers (after polishing ofT any oxide with emery cloth) in 10 c.c. of dilute sulphuric acid (i : 10) with the same precautions as in reducing by zinc. As the permanganate consumed is to the weight of iron, so will be that used for the unknown solution to the iron there present. In place of iron wire, pure ammonio-ferrous sulphate, FeSO 4 .(NH 4 ) 2 SO4.6H 2 O may be powdered and dried between blotting paper, and dis- solved in dilute sulphuric acid. 0*7 grm. of this salt contains o*l grm. Fe. Nitric acid, if present in the original solution, must be expelled by prolonged boiling during reduction. In presence of hydrochloric acid or chlorides, the cold solution must be diluted after reduction with 300 or 400 c.c. of boiled distilled water, which has been allowed to cool in a closed flask, and freely acidified with sulphuric acid and 3 or 4 grm. of manganese sulphate or magnesia sulphate must be added before titrating, but the results are less reliable than where sulphuric acid only is present. Gravimetric methods of separating iron from alumina are * If zinc quite free fiom iron cannot be obtained, a weighed quantity must be used in which the iron has been determined in a blank experiment and allowed for. ANALYSIS OF ALUM AND CHROME. 253 given -on p. 264, but the volumetric methods are generally more rapid and accurate. Iron is easily separated from chromium by fusing the mixture with sodium carbonate and adding ammonia, Fe 2 O 3 is precipitated and sodium chromate remains dissolved (loc. cit.). Estimation of Acids. Total sulphuric acid may be deter- mined in the usual way with barium chloride as described on p. 46, or by the benzidine sulphate method (p. 47). Hydro- chloric acid is titrated with argentic nitrate in presence of neutral potassic chromate and sodium acetate, or after neutralising with magnesia or calcium carbonate (p. 40). The free acid in solu- tions of aluminium sulphate or chloride may be approximately determined by titrating solutions of about decinormal strength with N/io sodium hydrate in presence of methyl orange, but to many people the change of colour is difficult to observe, as the indicator is affected by the acid reaction of the neutral salt. The recognition of the end point is assisted by comparison with a similar quantity of water with the same amount of indicator, of which the colour is fully turned by a drop or so of N/io acid. A more accurate method is that of Beilstein and Grosset (quoted by Sutton), which depends on the insolubility of alum and other sulphates in absolute alcohol. 1-2 grm. of the sample is dissolved in 5 c.c. of water, 5 c.c. of saturated neutral solution of ammonium sulphate is added, and the mixture is stirred for 15 minutes. 50 c.c. of absolute alcohol is now added, and after a short digestion, the mixture is filtered through a small filter, and washed with 50 c.c. more of absolute alcohol. The solution is distilled to recover most of the alcohol, and after evaporation on the water-bath to drive off the remainder, is titrated with N/io sodic hydrate or carbonate, and methyl orange. If preferred, the alcoholic solution may be titrated without evaporation, using caustic soda or lime water and phenolphthalein. Commercial sulphate of alumina frequently contains considerable traces of free acid. These are of little con- sequence for most purposes in leather manufacture, and may be easily neutralised by the addition of a little soda, which merely forms soda alum, or if in excess, makes the solution slightly basic. The use of basic solutions for tanning was patented some years ago by Mr. Bertram Hunt,* and they have some advantages * Eng. Pat. 15,607, 1884, now expired. 254 LEATHER INDUSTRIES LABORATORY BOOK. over those of normal salts. The amount of basicity may of course be found by determining the relative quantities of alumina and acid. Crystallised alums are generally practically neutral, while sulphate of alumina is sometimes acid, and sometimes basic, but in the latter case, the alumina is probably fully as available for tanning as in the normal salt. The total acidity of solutions of chlorides or sulphates of aluminium may be determined with very approximate accuracy on the principle of Hehner's method for " permanent hardness," though a portion of the acid is obstinately retained by the pre- cipitate (p. 32). In the case of sulphate of alumina, 5 grm. of the sample is dissolved in about 40 c.c. of water, and 50 c.c. of N/i sodium carbonate is added, and the mixture boiled for a short time. The liquid is then made up to 100 c.c., filtered, and 20 c.c., representing I grm. of the sample, and 10 c.c. sodium carbonate solution, is repeatedly titrated with N/i sulphuric or hydro- chloric acid and methyl orange. Deducting the acid used from 10 c.c., each c.c. of deficiency represents 0*049 g rm - f H 2 SO 4 , 0*0355 f Cl, an d in absence of other bases precipitated with sodium carbonate, 0*009 f Al. The method is not applicable to ammonia alum, or in presence of ammonia salts, on account of the volatility of the liberated ammonium carbonate, though possibly by using N/io solutions, and digesting for some time in the cold in a stoppered flask instead of boiling, correct results might be obtained. If the original substance contains free acid, this must be determined and deducted before calculating alumina. The writer, after repeated attempts, has failed in obtaining satis- factory results either from Bayer's method, or its modification quoted in Button's ' Volumetric Analysis.' * (Cp. also p. 265.) The atomic weight of Al is 27, and of A1 2 O 3 102. Potash alum, Al 2 (SO4)3.K 2 SO4.24Aq = 948 ; ammonia alum, A1(SO 4 ) 3 . (NH 4 ) a SO 4 .24Aq = 906. Crystallised aluminium sulphate is A1 2 (SO 4 ) 3 . i8Aq =666. As A1 2 is hexavalent in these salts, one- sixth of their atomic weights per liter produce normal solutions. Ckromitim Compoimds used in Tanning. Chromium (Cr = 52) forms four oxides. The lowest, chromous oxide, CrO, corresponds to ferrous oxide, and forms analogous salts, which absorb oxygen with such rapidity that they cart- only be pre- * Cp. Journ. Soc. Chem. Ind., 1890, p. 767 j 1891, pp. 202, 314. ANALYSIS OF ALUM AND CHROME. 255 served- with difficulty. Blue solutions of chromous salts may be obtained by treating acid solutions of chromic salts with excess of zinc or electrolytically by reduction with nascent hydrogen at the cathode. Their use for tanning has been patented by Berthon.* Such solutions might be used to reduce chromic acid in the two-bath process. The second oxide, chromic oxide, Cr 2 O 3 (molecular weight 152), is extremely similar to alumina in its general properties. It forms a hydrate Cr 2 (OH) 6 .6H 2 O, which, like alumina, has weak acid properties in presence of strong bases, and hence is soluble in potassic or sodic hydrate solution to a deep green liquid, from which, however, it is precipitated again on boiling. In ammonia it is slightly soluble, forming a pink compound, which also is decomposed on boiling. With acids it forms salts analogous to those of alumina and ferric oxide, which are often violet when crystallised (as chrome alum, K 2 SO4.Cr 2 (SO 4 )3.24Aq, mol. wt. 998), and usually form violet solutions, which become green on boiling, probably from dissociation and the formation of a basic salt.f It easily forms basic salts, which have powerful tanning properties, and are usually deep green or greenish-blue in solution. The more basic forms are very unstable, and are precipitated on boiling. The precipitated hydrate parts with water on boil- ing, and becomes much less soluble, both in alkalies and acids, and when ignited is converted into the oxide, which is practically insoluble in all reagents, unless it is first converted into chromic acid by oxidations at a high temperature with alkalies. The third oxide, CrO 3 , "chromic acid," or more strictly chromic anhydride (molecular wt. 100), forms ruby-red crystals^ which dissolve in water to an orange solution of chromic acid, CrO 4 H 2 , similar in constitution to sulphuric acid like which it forms two series of salts, neutral chromates, mostly of a lemon- yellow colour, and acid chromates, which are orange in solution. On crystallisation these solutions yield anhydrochromates, such CrO 3 K as potassium bichromate, of which the formula is O , or CrO 8 K * Fr. pat. 3534i8, April 17, 1905. f Some crystallised normal salts are bright green, and the nature of the change of colour is by no means certain. Probably chromium salts are much more complex than the formulae suggest. Chromium sulphates, for instance, often contain SO 4 in a form which is not precipitated by barium chloride till the complex acid is decom- posed, e.g. by boiling with an alkali. 256 LEATHER INDUSTRIES LABORATORY BOOK. Cr 2 K 2 O 7 (molecular weight 294), and which contains no water. The ammonium and sodium salts are analogous, but the latter crystallises with 2Aq, and is very deliquescent. Its molecular weight is 299, and hence it may usually be substituted for an equal quantity of potassium bichromate than which it is cheaper. Two parts by ^weight of bichromate yield approximately one part of chrome oxide Cr 2 O 3 . Free chromic acid is reduced to chromic oxide by most reducing agents, such as sulphurous, thiosulphuric and hydro- sulphuric acids, by ferrous salts, and by organic compounds such as alcohol, glucose and starch. If free chromic acid alone is present, or the reducing substance is insufficient in quantity for complete reduction, so-called "chromic chromate," or "chromium dioxide" (probably CrO 3 Cr 2 O 3 ), is formed, and the resulting pro- duct is brown, while in presence of excess of stronger acids, green or, in some cases, violet salts of chromic oxide are pro- duced. If the acid present is not sufficient to produce the normal salt, basic salts are formed which vary according to the amount of acid present. These salts have the property of pro- ducing completely insoluble chrome leather, unaltered even on boiling ; but when certain organic substances are used for reduc- tion, violet organic compounds are formed which are unsuitable for tanning, and act upon skin as if strongly acid. The most favourable compound for tanning contains approximately 3d to 2Cr, the formation of which requires 5 molecules of HC1 to I of potassium bichromate, and is easily made by adding the calcu- lated quantity of acid to a solution of bichromate, warming slightly, and adding glucose or sugar till the solution becomes a clear greenish-blue, carbonic acid being at the same time copiously evolved.* An equivalent quantity of sulphuric acid also pro- duces a similar solution of similar tanning properties, but giving a somewhat plumper and firmer leather. Basic solutions are also readily prepared by dissolving freshly precipitated chromic hydrate in solutions of sulphate or chloride of chromium, or by partially neutralising the acid of the normal salts with soda or sodium carbonate, f The degree of basicity most advantageous * Small quantities of organic matter often prevent precipitation of Cr 2 O 3 . Organic acids and aldehydes are among the organic products produced, and it is probably to the former that the injurious effects are due, while the action of aldehydes is advan- tageous to rapid tannage. The subject has not, however, been fully investigated. t Cp. Eng. Pats. 15,607, 1884, and 5491, 1886. ANALYSIS OF ALUM AND CHROME. 257 varies with the character of tannage required. The more acid solutious penetrate rapidly, but tan lightly, very basic salts precipitate in use, leaving a more acid solution. A formula given by the author,* which has been largely em- ployed in practical tanning, with excellent results, is 5 parts of concentrated sulphuric acid, 6 parts of bichromate (either of potash or soda), and 7 parts of good brown glucose or cane sugar. The bichromate is placed in a vessel, preferably stoneware or lead- lined, with say 12 parts of water, and the sulphuric acid gradually added with stirring so as to dissolve the bichromate by the heat produced. The glucose is now added in small portions to the hot liquor, which, if the quantity is large, may require time to cool, as the reaction is violent, and much gas and heat are evolved, while if the quantity is small, external heat (steaming) may be required to carry the reaction to its end, which is attained when the liquid is a bright green. Such a solution will contain about JO per cent, of Cr 2 O 3 , and may be diluted as required for use. The quantity of glucose is excessive, though it possibly tends to fuller leather, but about three parts is all which is actually oxi- dised. Certain samples of glucose, probably containing dextrine, produce the dull violet compounds referred to, and must be avoided. Excessive heat should be avoided, as it drives off the aldehydes formed, to which the rapid tanning properties of the solution are partly due. Potassium Bichromate (Cr 2 K 2 O 7 , molecular weight 294) is the usual starting-point in the formation of chromium compounds. The most important determination is generally that of the chromic acid present, on which its value depends. Estimation of Chromic Acid gravimetrically may be accom- plished by precipitation as lead or barium chromate, but most satisfactorily by reducing it to chromic oxide, by boiling its solution, acidified with hydrochloric or sulphuric acid, with a small excess of alcohol, till it becomes of a clear green colour, and then, after boiling off the excess of alcohol, precipitating with ammonia, igniting and weighing in precisely the same way as alumina (p. 250). Iron, alumina or chromic oxide, if present in the original solution, may be removed by precipitation with ammonia before reducing the chromic acid. (On the separation of these metals, cp. pp. 250, 365.) * Leather Trades Review, Jan. 12, 1897. 258 LEATHER INDUSTRIES LABORATORY BOOK. Volumetrically there are several rapid and accurate methods of estimation. A simple one, where but few analyses are to be performed, is to weigh out from a weighing bottle a small quan- tity of pure, dry ammonium-ferrous sulphate into a porcelain basin, add 5 or 10 c.c. of dilute sulphuric acid, and then run in the chromate solution to be tested from a burette, till a small drop withdrawn with a glass rod no longer gives any greenish shade when mixed on a white plate with a drop of freshly pre- pared solution of potassium ferricyanide (" red prussiate of potash"). 0*392 grm. of ammonium-ferrous sulphate contains 0*56, or \ of its weight of iron, and consequently corresponds to 0-052 of Cr, or 3'ioo of CrO 3 , from which the quantity of chromic acid may easily be calculated. A convenient modification of this method is to add to the weighed quantity of ammonium-ferrous sulphate a measured quantity of the chromic solution insufficient to wholly oxidise it, and to titrate back the excess of ferrous salt with standard per- manganate solution. 3 "92 grm. of the iron salt are just sufficient for I grm. of CrO 3 , and correspond to 10 c.c. of N/i perman- ganate. Hence, if 3*92 grm. are weighed out, and require 2*3 c.c. permanganate solution, the quantity of CrO 3 taken has been equal to f -j or 0*77 grm. The permanganate solution is added till a faint permanent pink is produced. A rapid and convenient method, where many estimations have to be made, and especially with dilute solutions such as those in use in the two-bath process of chrome tanning, is as follows : A measured quantity of the chrome solution, which should not contain more than about 0*1 grm. of potassium bi- chromate or its equivalent in chromic acid, is placed in a stop- pered bottle of 200 or 300 c.c. capacity, and 5 c.c. of concentrated hydrochloric acid, or an equivalent quantity of a dilute solution, and 10 c.c. of 10 per cent, solution of potassium iodide is added. The bottle is closed and well shaken, and allowed to stand a few minutes at least. It is then titrated with a N/io solution of pure sodium thiosulphate (24*8 grm. per liter) until the brown coloration of the liberated iodine has nearly disappeared, when about i c.c. of starch solution is added, and the titration con- tinued till the violet-blue of the starch iodide''has given place to the pale greenish tint of the chromium chloride. Each mole- ANALYSIS OF ALUM AND CHROME. 259 cule of CrO 3 liberates 3 atoms of I, and consequently each c.c. of N/io thiosulphate corresponds to 0*0049 grm. of potassium bichromate or 0*0033 f chromic acid. If pure thiosulphate is not at hand, a solution of about 25 grm. per liter of the purest attainable may be standardised against the iodine liberated by 20 c.c. of a solution of potassium bichromate decinormal as regards oxidising power, and containing 4*9 grm. per liter, which will correspond to 24*8 grm. per liter of sodium thio- sulphate. A similar method may be used to estimate the thiosulphate in "hypo liquors," by adding the liquor to be tested from a burette to the iodine solution obtained by treating 10 c.c. of the potassium iodide solution and 5 c.c. of hydro- chloric acid with 20 c.c. of N/io bichromate, which will corre- spond to 0^496 grm. of pure thiosulphate. No accurate results can be obtained by titrating thiosulphate solution direct with bichromate, as the oxidation is complex, and proceeds further than that with iodine. The iodide used must be free from iodate, which may be tested by acidifying a little with hydro- chloric acid, when, if iodate is present, iodine will be liberated colouring the solution brown, and giving the usual reaction with starch. The hydrochloric acid must contain no free chlorine, which may be similarly detected, and if necessary allowed for, by a blank test. Thiosulphate solutions should be kept in the dark, and do not remain unchanged for more than a few weeks (cp. p. 307). It is frequently desirable to determine not only the total quantity of chromic acid, but the relative amounts present as free acid and bichromate, or as bichromate and neutral chromate. It is obvious that all three of these cannot exist in the same solution, but it must not be forgotten that free chromic acid can coexist with other free acids, and that by any alkalimetrical process they will be estimated together. Neutral chromate reacts neutral with phenolphthalein, while bichromate and free chromic acid are acid to it. If, therefore, bichromate and neutral chromate are present, the former can be estimated by adding standard caustic potash or soda solution in presence of phenolphthalein till the latter is reddened. Every c.c. of deci- normal alkali will in this case correspond to O'OI47 grm. of potassium bichromate, 0*010 of " half-bound," or 0*005 of free chromic acid (CrO 3 ), as the question is one of acidity, and not S 2 260 LEATHER INDUSTRIES LABORATORY BOOK. of oxidising power. Hence free chromic acid will consume double as much soda as when present as bichromate. If a solution contains o 100 grm. of CrO 3 in whatever form, it will require 30 c.c. of N/io thiosulphate to reduce the iodine it evolves. If the chromic acid is free, it will also require 20 c.c. of N/io soda to make it neutral to phenolphthalein ; while, if it is present as bichromate, only 10 c.c., and as neutral chromate no soda will be required. We may, therefore, deduce the fol- lowing rule. If the N/io soda required to neutralise is less than one-third of the N/io thiosulphate, each c.c. of soda corresponds to O'OIO of CrO 3 as bichromate, or 0^0147 of potassium bi- chromate, and. the remainder of the chromic acid indicated by the thiosulphate will be present as neutral chromate. If the soda is more than one-third of the thiosulphate, each c.c. in excess of one-third corresponds to 0*010 grm. of free CrO 3 , and the remainder indicated by the thiosulphate is bichromate. If the soda exceeds two-thirds of the thiosulphate, the whole of the chromic acid is free, and the excess of soda over two-thirds cor- responds to some other acid. As the calculation is somewhat puzzling at first sight, some examples may be given. Suppose 10 c.c. of a chrome solution requires 15 c.c. of N/io thiosulphate and 7 c.c. of N/io sodium hydrate, we have (7 - ilj X O'OIO = 0-02 grm. of free CrO 3 , and 5- X O'OIO = 0*05 grm. of CrO 3 as bichromate, or 0*0737 of potassium bichromate. If only 4 c.c. of N/io sodium hydrate were required, we should have 0*04 grm. of CrO 3 present as bichromate and 0*01 grm. as neutral chromate, the 15 c.c. of thiosulphate employed showing that o 05 grm. of CrO 3 in all was present. It is also possible to estimate the free chromic acid directly, and the following methods are taken from a paper by Mr. Carlton Heal and the author, which was published in the Journ. Soc. Chem. Ind., 1895, p. 248. The first is a modification of one published by M'Culloch,* and depends on the fact that free * Chem. News, 1887, Iv. p. 2. ANALYSIS OF ALUM AND CHROME. 261 chromic acid is oxidised to blue perchromic acid by hydrogen peroxide, while bichromates are unaffected. Perchromic acid corresponds to the fourth oxide of chromium already mentioned, but is so unstable that little definite is known of its constitution. It is soluble in ether, forming a deep blue solution which retains its colour for a short time. The analysis is carried out as follows. A measured volume of the solution in which the free chromic acid is to be determined is introduced into a narrow stoppered bottle, or more conveniently, into a 4 oz. stoppered separating funnel, and rendered alkaline by addition of a known volume, say 5 c.c. of N/io sodium carbonate solution and 1-2 c.c. of " 10 volume " hydrogen peroxide, and about I inch depth of ether are added. Great care must be taken that both these reagents are thoroughly neutral, and especially the hydrogen peroxide, which is generally supplied in slightly acid solution, and in most cases requires a few drops of soda solution to make it neutral to litmus, or the amount of sodium carbonate neutralised must be deter- mined by a blank experiment. The solution is now titrated with N/io HC1, shaking well after each addition, till the ether takes a faint blue tinge, visible against a white background in a good light. Deducting the amount of acid used from the volume of soda solution, the remainder is the soda consumed in converting the free chromic acid into bichromate, and each c.c. of this corre- sponds to O'Oio of CrO 3 . If the separating funnel is employed, the neutralised solution can be run off, retaining the ether for another titration. Dreher employs the same reaction for the colorimetric estimation of chrome and chromic acid.* Another method invented by Mr. Carlton Heal depends on the precipitation of a basic copper chromate from copper sulphate by the least trace of neutral chromate in a boiling solution. To a measured quantity of the clear chromate solution, 2-4 drops of a 5 per cent, neutral solution of copper sulphate is added, and the liquid is titrated at boiling temperature with N/io caustic soda. Immediately the whole of the free chromic acid is con- verted into bichromate, and a trace of neutral chromate begins to be produced, a very noticeable brown precipitate is formed, which serves as an indicator. In clear solutions this method is equally accurate with that just described, but its practical use for chrome tanning solutions is limited by the fact that organic * Collegium, 1903, p. in. 262 LEATHER INDUSTRIES LABORATORY BOOK. matter dissolved from the skins produces a turbidity which resists all ordinary means of filtration. The determination of free chromic acid in used chrome liquors is, however, rendered of less import- ance by the fact ascertained by Mr. Heal and the author, that under ordinary circumstances all free chromic acid is absorbed by the skin, and only bichromate is left, which does not seem to be taken up in appreciable quantities.* Determination of Chromium in solutions of chrome salts, such as the basic liquors used in tanning, in the absence of organic matter, is of course easily performed gravimetrically like that of alumina by precipitation with ammonia, as has been already mentioned. Rough estimations may perhaps be made colori- metrically, but the colour of basic solutions varies so much that no great accuracy can be expected. Better results are obtained by oxidation with nitric acid, or hydrochloric acid and potassium chlorate or a persulphate, and comparison in alkaline solution as neutral chromate. Acid chromates are much influenced in colour by dilution and the presence of neutral salts. Its rapid volumetric estimation, however, is a matter somewhat difficult. The problem is to find an oxidising agent which is powerful enough to oxidise the whole of the chromium to chromic acid, and which can itself be completely removed from solution before titrating with iodide of potassium and thiosulphate. Hydrogen peroxide, which naturally suggests itself, although it raises some chromic acid to a still higher state of oxidation, acts also as a partial reducing agent, with evolution of oxygen. Sodium per- oxide in alkaline solution oxidises chrome salts rapidly to chromic acid, but must be completely destroyed by boiling before acidifi- cation, which demands perhaps half an hour, and considerable care is needed to secure complete oxidation, and to avoid low results.f The following method is satisfactory in absence of organic matter, in the presence of which it is liable to fail from the formation of green alkaline manganates which are not de- composed by boiling. A measured quantity of the chrome solution, which should not contain more than 0*25 grm. of * Cp. Journ. Soc. Chem. Ind., 1895. p. 248 ; 1900, p. 226. t Alden (Journ. A.L.C.A., 1906, p. 174; Collegium, 1906, p. 327) obtained very satisfactory results with sodium peroxide, and states that a few minutes' boiling is sufficient to destroy any excess. It is undoubtedly the most 'convenient reagent. For oxidising by fusion, Alden recommends the use of sodium peroxide in an iron crucible. ANALYSIS OF ALUM AND CHROME. 263 Cr 2 O 3 , is rendered decidedly alkaline with sodium hydrate and raised to a boil, and N/i potassium permanganate is added till the supernatant liquor remains pink after boiling for two or three minutes. A drop or two of alcohol is now added, and the boiling continued till the pink disappears, when the solution is cooled, made up to 250 c.c. and filtered, and the chromic acid is determined in 50 c.c. with potassium iodide and thiosulphate, as on p. 258. In most cases evaporation to dryness, and ignition with magnesia and sodium carbonate, as on p.265, is to be pre- ferred. Determination of Chromium and Alumina in Leather. General methods for the analysis of leathers and leather ashes are given in Section XXII., but as determinations of the mineral constituents of chrome leathers are frequently necessary, and form a somewhat special case, their execution may be here con- veniently described. The leather is incinerated in the usual way (p. 363), and the principal constituents of the ash are salts of alumina and chrome, frequently mixed with iron used as a dye-mordant or even as a tanning agent, and with traces of lime and alkaline salts from imperfect deliming and washing of the leather after chroming, reducing or neutralising, or from the soaps employed in fat- liquoring. As obtained from the ignition of leather all three bases are usually in a very insoluble form, and especially chrome oxide when strongly ignited is quite unattacked even by prolonged digestion with concentrated acids. Recourse must, therefore, be had to fusion with alkaline bases with which the chromium is readily oxidised to soluble chromates, and the alumina forms soluble aluminates, while ferric oxide remains undissolved. For qualitative detection of chromium, fusion of a trace of ash in a borax bead in a loop of platinum wire is a sufficiently delicate test, minute quantities producing an intensely green glass. In absence of chromium iron produces a pale green glass in the inner or reducing, and a yellow in the outer or oxidising flame of the Bunsen burner or blowpipe, but these reactions are quite covered by mere traces of chromium. Iron may be detected by the ordinary thiocyanate or ferrocyanide reactions in the solution obtained by digesting a portion of the ash with concentrated hydrochloric acid in a porcelain basin, the chromium remaining 264 LEATHER INDUSTRIES LABORATORY BOOK. mostly undissolved. Platinum should not be used, as it may be attacked by the chlorine which is liberated. There is no satis- factory qualitative test for alumina, but in absence of iron and chromium it may often be detected in the leather by moistening with an alcoholic infusion of logwood, which gives a violet coloration if alumina is present, but is useless in presence of iron and chrome, which produce still deeper colorations. For actual proof of the presence of alumina, it is necessary to resort to quantitative separation. For this purpose it is most convenient, after powdering the ash very finely, preferably in an agate mortar, to mix it inti- mately with three or four times its weight of a mixture of about equal parts of dry K 2 CO 3 and Na 2 CO 3 and ignite it in a platinum crucible gently at first but finally before the blowpipe, placing the crucible sloping so that air obtains free access to the mixture. When the process is complete the melt should be pure yellow or at least quite free from green tinge. The operation may be shortened by the addition of small portions of sodium peroxide towards the end of the operation, but this is not only somewhat injurious to the crucible, but the peroxide must be completely decomposed by boiling before the direct volumetric estimation of the chromic acid. The Cr now exists as sodium and potassium chromates, the alumina as aluminates, and the Fe as insoluble Fe 2 O 3 . The melt is cooled rapidly by placing the crucible on a cold iron plate, which makes the mass separate from the crucible, and is treated with hot water, when the Cr and Al pass into solution, and the undissolved* iron oxide may be col- lected on a filter, washed, ignited and weighed as Fe 2 O 3 . To the filtrate (with the washings) excess of ammonium chloride is added, or the liquid is first rendered acid with HC1 and boiled to expel CO 2 , and is then rendered slightly alkaline with ammonia and boiled to precipitate the A1 2 O 3 which is filtered off, ignited and weighed. The CrO 3 in the filtrate may now be estimated volu- metrically with potassium iodide and thiosulphate (p. 258), or reduced by acidifying and boiling with a little alcohol and finally precipitated as Cr 2 O 3 and estimated gravimetrically (p. 257). The same method of separation is applicable to precipitates con- * If any considerable portion remains undissolved, it is safest (e mix with a fresh portion of fusion-mixture, and repeat the fusion, so as to be sure that no Cr or Ai remains undissolved. ANALYSIS OF ALUM AND CHROME. 265 taining Fe, Cr and Al obtained in analysis by precipitation of the " third group " metals with ammonia. Where estimation of magnesia is not required, the fusion- mixture of potassium and sodium carbonates is conveniently re- placed by one of equal parts of MgO and Na 2 CO 3 , which does not fuse but merely sinters, so that it can be readily and frequently stirred with a platinum wire. If the quantity of mixture does not exceed I or 2 grm., the oxidation is readily completed in about 20 minutes over a Teclu burner. The mixed chromates are dis- solved with dilute HC1, and the chromic acid is estimated with potassium iodide and thiosulphate as on p. 258. The same method is conveniently applied to chrome-liquors, by evapora- ting a measured quantity to a small bulk in the crucible, adding the fusion-mixture, and after evaporation to dryness, and, if necessary pulverisation, completing the ignition as above. With the methods which have been given in this section, there is no difficulty in determining the proportions of chromic acid, chromate and bichromate in a used liquor containing chro- mates only, but the question is much complicated if normal and basic salts of chromium and aluminium must also be considered. The degree of basicity of chrome liquors is also difficult to deter- mine, especially where aluminium is also present. In this case the only accurate method is probably to deter- mine the whole of the bases and the whole of the acids, but results of very considerable accuracy are obtained by titration of the diluted boiling solution with N/i sodium hydrate in presence of a rather large quantity of phenolphthalein. None of the ordinary impurities contained in tanning liquors interfere.* Kopeckyf estimates acidity by boiling with magnesium carbonate, and determining magnesia in the filtrate, and from this calculating the acid. The method is somewhat tedious but appears otherwise satisfactory, though the formation of basic salts does not seem unlikely. Procter and McCandlish (loc. cit.) attempted to estimate acidity by titrating the carbonic acid liberated by boiling with sodium carbonate, but though theore- tical results were obtained when the CO 2 was liberated with acids, no good concordance with theory could be obtained with chrome salts, presumably from the formation of undecomposed * Procter and McCandlish, Journ. Soc. Chem. Ind., 1907, p. 458.] t Journ. A.L.C.A., 1906, p. 261. ^66 LEATHER INDUSTRIES LABORATORY BOOK. basic salts. Appelius and Schall have recently published a similar method in which the CO 2 is absorbed and weighed, and the experiments they quote show a satisfactory agreement with theory. Much chrome liquor is usually run away unexhausted, from the fear that for want of more accurate knowledge it cannot be so strengthened as to exactly reproduce its original character. This fear is usually much exaggerated, and the practical chemist finds no difficulty as a rule in using the same liquor repeatedly, but where this is not done the recovery of the chrome is worth atten- tion. Probably the cheapest way is to reduce the whole to chrome oxide with " hypo " liquors, and after precipitation with lime, to furnace the sludge with potash or soda and recover bichromate. A more practicable method for tanners is to precipitate with sodium carbonate, and to use the chromium hydrate in the produc- tion of basic liquors. Chromic acid may be precipitated by barium hydrate, or barium chloride with an equivalent of lime or soda, and the barium chromate so obtained, acidified with sulphuric acid, will give a solution of chromic acid which may be used in the two-bath process. SECTION XVIII. DETECTION AND ESTIMATION OF GLUCOSE. THE detection and estimation of sugars (glucose, etc.) is im- portant to the leather manufacturer in several directions. In leather, glucose or starch syrup is often employed as a weighting material, and its detection and the estimation of added weight are often desired. It is sometimes similarly added as an adul- terant to extracts, but bodies which behave like sugar are normally present in most tanning materials, where they play an important part as sources of acetic and lactic acid which are required in the liquors. It is obvious, that as traces of sugars are always found in tanning materials and therefore in leather, it is necessary, in order to prove adulteration, to know how much may be naturally present, and to determine quantitatively that more DETECTION AND ESTIMATION OF GLUCOSE. 267 exists than can be so accounted for. For some further informa- tion on sugar in tanning materials, see pp. 276, 360. As the quantities found are often very small, it becomes de- sirable to somewhat vary the methods adopted in ordinary sugar determination, so as to avoid the necessity of concentration, and of the use of inconveniently large quantities of material. The subject has been investigated by Sim and* and by v. Schroeder, Bartel and Schmitz-Dumont,f and the methods given are selected from those they describe. They are modifications of that originally discovered by Fehling. Preparation of FeJiling's Solution. This is best prepared in two separate solutions which are only mixed at the time of using. It is suitable either for Fehling's original volumetric method, or for the gravimetric process of Allihn and its modi- fication adopted by v. Schroeder. (1) 34*639 grm. of the purest crystallised copper sulphate is dissolved in distilled water, IQ c.c. of normal sulphuric acid is added, and the whole made up to 500 c.c. (2) 173 grm. of pure crystallised potassium sodium tartrate (Rochelle salt) and 125 grm. of caustic potash ''pure by alcohol," are dissolved in water and made up to 500 c.c. Simple Volumetric Method. Though of less accuracy than the gravimetric method (which should always be used where an absolute determination is required), the volumetric method is sometimes very convenient as affording an approximate estimate of the sugar present, or as a semi-qualitative test for adulteration. Solutions containing tannin must be detannised and decolorised by basic lead acetate solution as described below, but where great accuracy is not expected it is not always important to remove the excess with sulphate. The decolorised and neutralised sugar solution, which should contain about J per cent, of reducing sugar, is placed in a burette. 5 c.c. of each of the solutions I and 2 are placed in a small flask or boiling tube held in a clip in a retort stand ; 40 c.c. of water are added and a few pieces of broken pipe-stem to prevent bumping. The liquid is brought to a boil over a small flame, and the sugar solution is run in I or 2 c.c. at a time, the liquid being boiled up between each addition. When the blue colour * Zeit. fur angew. Chem., 1892, Heft 23. t Ding. Polyt. Journ., ccxciii. (1894), p. 229. 268 LEATHER INDUSTRIES LABORATORY BOOK. has nearly disappeared, the sugar solution may be added in smaller quantities, and the red cuprous oxide allowed to subside after boiling, until on looking through it at a window, or white surface, the supernatant liquor appears colourless, or faint yellow. The requisite amount being approximately found in this way, a second titration must be made, adding nearly the whole of the required sugar solution at once, so that the entire boiling does not last longer than two minutes. Approximately 10 c.c. of a J per cent, sugar solution will be required, and if less is needed, the solution should be diluted accordingly and the titration repeated. If, as is generally the case in the analysis of leather and tanning materials, the sugar solution is so weak that more than 10 c.c. is required to effect decolorisation, it is best to reduce proportionately the water otherwise added to the Fehling's solution. Considerable accuracy may be obtained by direct comparison with a standard sugar solution, diluting in both cases with water, so that the same total volume is employed, and heating in a final titration for exactly the same time. Under the conditions given, the 10 c.c. of mixed solution will be reduced by the following approximate quantities : 0*0500 grin, of glucose (dextrose, laevulose, or invert sugar). 0-0475 > sucrose (cane sugar), previously inverted. 0-0678 lactose (milk sugar). 0*0807 i maltose (malt sugar). These quantities are, however, dependent on the exact con- ditions under which the titration is performed, and vary slightly with the dilution of the solution and the length of time for which it is boiled. It is best, therefore, to determine the amount by comparison with an actual sugar solution, and as the exact character of the sugars to be estimated is unknown, it is con- venient to use sucrose (cane or beet sugar), which is always attainable in a state of great purity. Sucrose, however, is only reduced by Fehling's solution when it has been previously " inverted," or converted into glucoses, by the action of acid. These glucoses may for the present purpose be assumed to be identical in their action on the solution, though strictly this is not quite the case. A standard sugar solution may be prepared as follows : 0*95 grm. of good lump sugar is dissolved in about 75 c.c. of water, 2 c.c. of N/i HC1 is added, and the mixture is DETECTION AND ESTIMATION OF GLUCOSE. 269 heated in the boiling water-bath for 30 minutes. The acid is then neutralised with 2 c.c. of N/i NaOH or Na 2 CO 3 solution, and made up to 200 c.c. This solution should be equivalent to a 0-5 per cent, solution of glucose, and should decolorise a mix- ture of 5 c.c. of each of the two Fehling's solutions ; and if this is not exactly the case, a suitable factor must be calculated for the Fehling's solution under the prescribed conditions. Un- adulterated leathers rarely contain I per cent, and never more than 2 per cent, of sugar, except perhaps when tanned with myrobalans or pine-bark, or one or two other materials which are particularly rich in sugar, while it is hardly likely less than 5 per cent, would be added as an adulteration. Thus, if 10 grm. of the suspected leather be extracted with warm water and the solution after decolorising by addition of lead solution be made up to 250 c.c. and filtered, at least 16 c.c. will be required to decolorise the 10 c.c. mixed Fehling's solution, if the sample is not adulterated. Von Schroeder's Gravimetric Method. This is a modification of Allihn's method specially adapted for the small quantities of sugars naturally found in tanning materials and leathers ; and is much more accurate than the volumetric method just de- scribed. As the result is to a marked extent dependent on the concentration of solutions and the time of heating, the details given must be rigidly adhered to in order to get accurate results, and the special table of Koch and Ruhsam given by v. Schroeder must be employed. Preparation and Decolorisation of Solutions. It is necessary to remove all tannin and gallic acid as well as colouring matters. This is best accomplished with basic acetate of lead. The lead solution is prepared by rubbing together 300 grm. of pure lead acetate, 100 grm. of pure and finely powdered litharge, and 50 c.c. of water, and digesting the mass on the water-bath till it becomes white, water being added as required to replace evapo- ration. The mass is made up to i liter with distilled water, allowed to settle, and filtered. io c.c. of the lead solution is titrated with a strong solution of sodium sulphate till no further precipitate is produced, and 100 times the quantity of sulphate solution so employed is made up to i liter. The titration need not be very accurate, but it is better to have the sulphate a little in excess than the reverse. 270 LEATHER INDUSTRIES LABORATORY BOOK. Sodium sulphide may, if required, be used as an indicator by spotting on filter-paper, as on p. 56. The tanning infusion or leather extract (see pp. 276, 360) should contain at least I *5 grm. of solids per 100 c.c., and if the solutions employed for tannin analysis be used, it must be con- centrated to this extent by evaporation. It must not, at the most, contain more than i per cent, of sugar, but very small quan- tities can be satisfactorily estimated. To 200 c.c. of this solution, which need not be carefully filtered, 20 c.c. of the basic lead solu- tion is added, well shaken, and allowed to stand for 15 minutes and filtered through a dry filter. The filtrate should give no further precipitate with a drop of the lead solution. Should it do so, more both of this and of the sodium sulphate must be used, and allowed for in calculating the results. To 1 10 c.c. of the lead filtrate 10 c.c. of the sodium sulphate solution is added, and after the lead sulphate has completely settled, the liquid is again filtered through a dry filter. If the above-mentioned quantities have been used, 120 c.c. of this filtrate is equal to 100 c.c. of the original infusion, and the sugar found must there- fore be multiplied by I *2 to bring it to original value. If 30 c.c. of lead solution is required, 115 c.c. of the filtrate must be treated with 15 c.c. of sodium sulphate, and the result multiplied by I 3. Determination. 30 c.c. of each of the Fehling's solutions and 60 c.c. of water are put into a beaker of 200 c.c. capacity, and brought to a boil over a flame. The beaker is then placed in a boiling water-bath, 25 c.c. of the decolorised sugar solution is introduced and stirred up, and maintained at boiling tem- perature for exactly 30 minutes. If the sugar solution is very weak, 50 or 75 c.c. may be used, the water added to the Fehling's solution being correspondingly reduced, so that the total volume remains 145 c.c. At the expiration of 30 minutes the liquid is allowed to settle, and at once filtered through a weighed asbestos filter by aid of the filter-pump. The filter (Fig. 18) consists of a piece of com- bustion tube about 10 cm. long and 1-2 cm, interior diameter. The lower end is drawn out to about half its original diameter, while a small funnel is fitted into the top with a perforated rubber cork. A small plug of glass wool is placed in the conical part to prevent fibres of asbestos being carried through, and about 3 cm. of the tube is rilled with fibrous asbestos, of which DETECTION AND ESTIMATION OF GLUCOSE. 271 the lower part is pretty closely packed, while the upper part is quite loose. Packed in this way the filter is much less liable to choke, as a large part of the cuprous oxide is caught on the loosely packed layer. As some samples of asbestos are somewhat soluble in the solu- tions employed, it is well to wash the asbestos before using with hot 20 per cent, solution of caustic soda, and then with distilled water, and to make a blank experi- ment with a filter to ascertain that its weight remains constant. The precipitate is thoroughly washed with hot water, and then with alcohol, and finally with a little ether to quicken the drying, which may be accomplished in an air- bath in less than 15 minutes. The tube is then gently ignited with a slight current of air through it to destroy possible traces of organic matter, and the copper reduced in a stream of pure dry hydrogen, the tube being heated by a Bunsen burner. The reduction takes place at a low temperature, and it is not necessary to bring the tube actually in contact with the flame, and especially the glass wool must not be ignited. When the precipitate has taken the colour Qf metallic copper, it is allowed to cool in the stream of hydrogen. A little dry air is now sucked through to expel the hydrogen, and the tube is weighed ; and from the gain in weight the amount of glucose is found by means of the table on pp. 272 and 273. The same filter may be repeatedly used if the copper be removed with a few drops of strong nitric acid, and the filter well washed with water, alcohol and ether, dried and re-weighed. Under this treatment a good filter should not lose any appreciable weight. In place of using the asbestos tube and reducing to metallic Cu, small ordinary quantitative filter papers may be employed, preferably with a perforated platinum (or parchment-paper) cone in the funnel and with the aid of the vacuum-pump. As the Cu 2 O is very finely divided it is sometimes necessary to use a FIG. 18. 272 LEATHER INDUSTRIES LABORATORY BOOK. double filter. The precipitate should be first washed repeatedly by decantation to free it as far as possible from the alkaline solution, and afterwards on the filter till the latter no longer reacts with phenolphthalein. Paper absorbs and obstinately re- tains traces of copper, and Fehling's solution sometimes gives a slight precipitate when boiled alone, so that it is necessary to determine these errors by a blank experiment without sugar, and use the result as a correction. The paper is dried in a basin or crucible, slowly charred and heated for some time, first TABLE FOR THE DETERMINATION OF THE GLUCOSE IN TANNING MATERIALS FROM THE WEIGHT OF COPPER REDUCED, AFTER HEATING FEHLING'S SOLUTION WITH THE GLUCOSE SOLUTION FOR HALF AN HOUR. (R. Koch and R. Ruhsam.} Cu. Glucose. Cu. Glucose. Cu. Glucose. Cu. Glucose. Cu. Glucose. mgr. mgr. mgr. mgr. mgr. mgr. mgr. mgr. mgr. mgr. I '4 39 16-3 77 35' i "5 54-6 153 73'5 2 0-8 4 16-7 78 35'7 116 55'i 154 74-0 3 4 1-6 41 4 2 17-2 17-6 79 80 36-2 117 118 55'7 56-2 IP 74'5 5 2'0 43 18-0 81 37 '2 119 56-7 157 75'5 6 2*5 44 18-4 82 37'7 120 57'2 I 5 8 76*0 8 3'3 45 46 18-9 83 84 38-2 38; 7 121 122 ll'l '59 1 60 76-5 77-0 9 3]7 47 19-7 85 I2 3 58-7 161 77'5 10 48 20'2 86 39'8 124 59*2 162 78-0 ii 12 4'5 4*9 49 50 21*3 87 88 40-3 40-8 59-7 60-2 I 3 164 78-5 79-0 13 5'3 21-8 89 41 '3 127 60-7 165 79'5 14 5 ' 7 52 22 '3 90 41-8 128 6l'2 1 66 80-0 15 6* i 53 22-8 91 42-3 129 61-7 167 80-5 16 6*5 54 23^ 92 42-8 I 3 62-2 1 68 81-0 17 7-0 55 23'9 93 43'3 62-6 169 81-4 18 7'4 56 24*4 94 43*9 I 3 2 63-1 170 81-9 19 7-8 57 24-9 95 44'4 133 63-6 171 82-4 20 8'2 58 25'4 96 44 '9 134 64-1 172 82-9 21 8-6 59 97 45'4 135 64-6 173 83-4 22 9-0 60 26-4 98 45*9 136 65-1 174 : 83-9 23 9'4 61 26 '9 99 46-4 137 65-6 175 ! 84-4 24 9*9 62 27-4 100 46-9 138 66-1 176 84-9 25 10-3 63 28'0 101 47'5 139 66-6 177 85-4 26 10-7 64 28-5 102 48*0 140 67-1 178 85-9 27 II'I 65 29-0 I0 3 48-5 141 67-6 179 86 '4 28 n-6 66 29*5 I0 4 49-0 142 68-1 180 86-9 29 30 I2'0 I2- 4 68 30*0 30-5 % 49'5 50-0 143 144 68-6 69*1 181 87-4 182 87-9 31 I2'9 69 31-0 107 5 '5 145 69-6 183 88-4 32 I3-3 70 31-6 108 51-0 I 4 6 70-1 184 88-9 33 13*7 71 32-1 109 51-6 147 70-6 185 89-4 34 I4-I 72 32-6 no 52-1 I 4 8 71-1 186 89-9 35 I 4 '6 73 33;i in 52-6 149 187 90-4 36 I5-0 74 112 150 72''0 188 90-9 37 '5*4 75 34' i "3 53'6 151 72-5 189 91-3 38 *5'9 76 34-6 114 54*1 152 73'o 190 91*8 i 1 DETECTION AND ESTIMATION OF GLUCOSE. 273 TABLE FOR THE DETERMINATION OF GLUCOSE continued. Cu. Glucose. Cu. Glucose. Cu. Glucose. Cu. Glucose. Cu. Glucose. mgr. i mgr. mgr. mgr. mgr. mgr. mgr. mgr. mgr. mgr. 191 9 2 '3 249 122*1 306 I52-2 363 183*7 4 20 2l6'7 192 9 2-8 250 122*7 307 I52-8 364 184-2 4 2I 217-3 193 93-3 251 I23-2 308 153-3 365 184-8 422 217-9 194 93-8 252 123-7 39 153-9 366 I85-4 423 218-4 195 94; 3 253 124-2 3 IO 154-4 367 186*0 424 219*0 I 9 6 254 I24-8 3 11 I 55" 368 186*5 425 219*6 197 95'3 255 I25-3 312 I55-5 369 187-1 426 22O"2 I 9 8 95-8 2 5 6 I25-8 156-0 370 187-7 427 220-8 199 96-3 257 I26-3 314 156-5 371 188*3 428 221'4 2OO 96-8 258 I26-9 3i5 372 188*8 429 22I-9 201 97-3 259 I27-5 316 157-6 373 189-4 430 222*5 202 97-8 260 128-0 158-1 374 190*0 431 223-I 203 98-3 26l I28-5 318 158-7 375 190*6 432 223-7 204 98*8 262 I29-0 319 159-2 376 191*1 433 224-4 205 99'3 263 I29-5 320 159-8 377 191-7 434 225T 206 99*8 264 I30T 321 160-3 378 192*3 435 207 100*3 265 I30-6 322 160-9 379 192*8 43 6 226*4 208 100*8 266 I3I-I 323 i 161-4 38o 193-4 437 227*1 20 9 101-4 267 I3I-6 324 162-0 194-0 438 210 101-9 268 I32-2 325 162-5 382 194-6 439 228'5 211 102*4 26 9 I32-7 326 163-0 383 195-2 440 229*1 212 102*9 270 133; 2 327 163-6 384 I95-7 441 229*8 213 I0 3'5 271 328 164-1 385 196-3 442 230-5 214 104-0 272 I3 4 -2 329 164-7 386 196-9 443 231 *2 215 104-5 273 I34-7 330 165-2 387 197-5 444 231*8 216 105*0 274 I35-3 33 i 165-8 388 198*0 445 232'5 217 105-5 275 135-8 166-3 389 198-6 446 233-2 218 106*0 276 I36-3 333 166-9 390 199-2 447 219 106*6 277 136-8 334 167-4 199-8 448 234*5 220 107-1 278 I37-4 335 167-9 392 200-3 449 235-2 221 107*6 279 137-9 336 168-4 393 200-9 450 235-9 222 108*1 280 I38-4 337 169-0 394 201-5 451 236-6 223 108*7 281 I39-0 338 169-5 395 202-1 452 237-2 224 log' 2 282 I39-5 339 170-1 396 202 "J 453 237-9 225 109*7 283 I4O-O 340 170-6 397 203*3 454 238-6 226 IIO*2 284 I40-5 34i 171-2 398 203*8 455 239*3 227 110*7 285 I4IT 342 171-7 399 204-4 456 239-9 228 111*2 286 I4I-6 343 172-2 400 205*0 457 240*6 229 iii*8 287 142' I 344 172-8 401 205-6 458 241*3 2 3 112*3 288 I42*6 545 I73-3 402 206-2 459 242*0 2 3 I 112-8 289 143*2 346 i73'9 403 206-8 460 242*6 232 233 113-3 113-8 290 291 I43-7 I44-2 348 174-5 404 405 207*3 207*9 461 462 243-3 244*0 234 114-4 292 I44-7 349 175-6 406 208*5 463 244*7 235 114-9 293 I45-3 ! 350 176-2 407 209* I 464 245-3 2 3 6 115-4 294 145-8 176-8 408 209*7 465 246-0 237 115-9 295 I46-3 352 I77-3 409 210*3 466 238 116-4 296 146-9 353 177-9 410 210*8 467 247*4 239 117-0 297 147-4 354 178-5 411 211*4 468 248-0 240 117-5 298 I47-9 355 179*1 412 2I2*O 469 248-7 241 118-0 299 148-4 356 179*6 413 212*6 470 249-4 242 118*5 300 I49-0 357 180*2 414 213*2 250-I 243 119-0 301 H9-5 358 180*8 415 213*8 472 250-8 244 H9-5 302 I50-I 359 181*4 416 214*4 473 25I-4 245 120- I 303 I50-6 36o 181-9 417 214-9 474 252-I 2 4 6 120-6 304 I5I-I 182-5 418 2I5-5 475 247 121* I 305 362 183-1 419 2l6'I 476 253'5 248 121*6 274 LEATHER INDUSTRIES LABORATORY BOOK. to a dull, and afterwards to a bright red heat, and the precipitate is weighed as CuO, which must be reduced to Cu for comparison with the table by multiplication by . or practically by o 8. If complete oxidation is difficult, the precipitate may be moistened with concentrated nitric acid and again ignited. Hellon, Winter Blyth, and others, redissolve the copper oxide in dilute nitric acid, and deposit electrolytically on a platinum dish, a continuous current of i to I ampere current- density and 2*2 to 2*5 volts being used. The solution (looc.c.) should not contain more than 3 per cent, of nitric acid, and the temperature should be 20 to 30 C. When all the Cu has been deposited, which will require from two to five hours, and may be tested by addition of ammonia to the solution, the dish is washed with distilled water while the current is still passing, till all nitric acid is removed. The deposited copper is now washed further under the tap, rinsed with methylated alcohol and then with ether, dried in water oven, cooled, and weighed, the basin being afterwards freed from copper with nitric acid. An ordinary electric supply may be used if of continuous current, the resistance being regulated by lamps. The anode is a piece of platinum foil,, and the basin, forming the cathode, is connected to the negative wire, and is conveniently placed in a photo- graphic tray to receive the washings. It must be borne in mind that the glucoses which are em- ployed for weighting leather are by no means chemically pure r but contain variable quantities of maltose and dextrin, the colourless glucose syrup especially containing much larger quantities of dextrin than the ordinary solid glucoses. As Fehling's solution does not reduce dextrin when free from sugar, a mere determination of glucose alone gives but little information as to the actual amount of weighting material em- ployed. If, however, such glucoses containing dextrin are heated with dilute acid, the dextrin and maltose are converted into an equivalent weight of glucose, which can be directly estimated by Fehling's solution. Ninety-three parts of starch yield 100 parts of glucose as estimated by Fehling's solution,* and dextrin and maltose occupy an intermediate position, so that no great error is involved if it is assumed that the total * Noyes and others, Journ. Am. Ch. Soc., 1904, xxvi. 266 et seq. DETECTION AND ESTIMATION OF GLUCOSE. 275 weight of glucose found after conversion is equivalent to the original weight of dry matter added, since, of course, commercial glucose always contains impurities which are not converted into sugar, and it is practically impossible to hydrolyse starches and dextrins completely, the attainable maximum being of some 96 to 98 per cent. To the detannised solution containing glucose and dextrin, 10 per cent, by volume of hydrochloric acid of sp. gr. I 125 is added,* and the solution is raised to iooC., and heated on a water-bath for two hours, either in a closed flask or with a reflux condenser to prevent evaporation. 90 per cent, or more of the hydrochloric acid is neutralised with strong caustic soda solution, of which the necessary amount can be calculated, and the solution is made up to a known volume and either titrated against Fehling's solution as on p. 267, or determined gravimetrically as on p. 270. It is stated (loc. cit.) that the con- version proceeds better in solutions of 2 to 4 per cent, than in those of I to 2 per cent, but the time given should hydrolyse 97 to 98 per cent., even in a i to 2 per cent, solution. Simand t only employs 7*5 per cent, of hydrochloric acid, sp. gr. 1*125, instead of 10 per cent, as recommended above, but probably it is safer to use the larger quantity. In calculating the actual weighting of leather from the glucose found, it must be remem- bered that the commercial glucose is hygroscopic, and must retain at least as much (and probably more) moisture than the leather in which it is found. The average moisture of air-dried leathers is about 1 5 per cent., but varies with the weather. Ex- periment as to the actual weight given by known quantities ot commercial glucoses are desirable. The sugar naturally present in tanning materials is not merely important as influencing the calculation of weighting by glucose, but as furnishing the basis from which the natural acids of the liquors are derived by fermentation. The sugar is of more than one species, dextrose, levulose and less definitely known sorts being present in different materials, but probably the estimation as glucose by Fehling's solution gives a pretty accurate idea of their total percentage. The result of a large number of determinations by Von Schroeder J are given in * Loc. cit. t Zeits. Angew. Ch., Heft 22, 1892. t Gerberei-Chemie, Berlin, 1898, pp. 577-623. T 2 276 LEATHER INDUSTRIES LABORATORY BOOK. the following Table. These determinations were made by the gravimetric method (p. 270) after removal of the tannin by pre- cipitation with lead, and therefore only include the sugars existing as such ready formed in the infusion. In addition to these it is known that many tannins, like the glucoside colouring matters, contain sugars or sugar-yielding bodies, and like these can be hydrolysed by digestion with dilute acids. The author has, however, been unable to find in literature detailed observations SUGARS NATURALLY CONTAINED IN TANNING MATERIALS. (Von Schroeder.) Tannin? Matter. Sugars. Per cent, on Material. Per cent, on Tanning Matter. Oak bark (average 118 samples) ID'S 2'7 25-2 ,, inner flesh, tree 150 years old . I3'8 i'3 9'5 ,, outer crust ,, ,, 7'6 0-7 9 '2 ,, young tree bark (Tharand) 13-0 6-6 50-8 Oak wood. Trees over 100 years old 7*7 o*4 5'8 (Mitrowitz) ,, Q. sessilifiora. Young oak, 2*2 1*2 53'7 19 years ; wood only Evergreen oak. Q. ilex, bark I7-7 3'6 20-3 Garouille. Root-bark Q. coccifera, average 25-4 I'O 4-0 Pine bark, Abies excelsa, Lam., average . n-6 3*5 33'5 Willow barks (Russian), average . I3'4 4*5 33-6 Mimosa barks (Australian wattles), average 28-4 0-9 3'2 Aleppo pine (" Scorza rossa"), outer bark 20'6 ,2'0 9*9 Hemlock pine, Abies canadensis, old bark 12-3 7'I 5-8 Divi-divi pods, Ctzsalpinia coriaria 40-7 8-4 20-5 Algarobilla pods, C. brevifolia, average . 42-9 8'2 19-1 Myrobalans fruit of Terminalia chebula, 30-8 5*4 17-4 average Valonia sugar very variable, average 28-3 2-7 9'5 Sumach (Sicilian) .... 27-8 4'6 16-6 Canaigre, root of Rumex hymenosepalus . 30-1 4*3 I4'3 Chestnut wood, without bark 8'3 0-3 2-9 Quebracho wood, Loxopterygium Lorenzi 24'4 0-25 I'O Cube Gambier, Nauclea gambir . 47'2 Vf 3'9 Cutch, Acacia catechu, wood extract 39'9 o ; 5 I'J DETECTION AND ESTIMATION OF GLUCOSE. 277 either with regard to the quantities of these sugars, or the time, temperature and concentration of acid necessary for complete hydrolysis. If the heating with acid, such as is described for conversion of starch into glucose, were carried out before instead of after the detannisation with lead, any sugars liberated from the tannins would of course be estimated with those existing ready formed. The author hopes to investigate this subject further. Von Schroeder made a number of determinations of sugars in extracts, and in many cases the results show a higher percent- age than is found in the original material. It is not certain that such a result indicates adulteration with glucose, which is rarely practised, since it may be due to decomposition of tannin and liberation of glucoside sugar in the process of manufacture, the decomposition of cellulose, or the use of starchy materials in decolorising. SECTION XIX. ANALYSIS OF SOAPS. SOAPS are not only of direct importance to the leather manu- facturer, but their examination forms a good introduction to the study of oils and fats, and a brief account of the principal methods is therefore given. For more detailed information the reader is referred to Lewkowitsch's ' Oils, Fats and Waxes/ third edition, pp. 1074-1082. Sampling demands considerable care, as soaps dry rapidly, and the outer part of a bar has often lost much water before it reaches the chemist. In bar soaps, it is best to use the central portion only if the original character of the soap is to be deter- mined ; and in soft soaps, the sample should be drawn from the centre of the barrel. Weighings should be made rapidly, or in closed vessels, and methods of analysis involving shaving finely before weighing are to be avoided. Determination of Water. About J inch depth of sand, pre- viously ignited, is placed in a beaker of about 100 c.c. capacity, together with a short glass rod, and the whole is weighed ; 3 to 278 LEATHER INDUSTRIES LABORATORY BOOK. 5 grm. soap, cut in slices, is quickly weighed to this, and 25 c.c. good alcohol added. The beaker is placed on a water- bath, and stirred till the soap is dissolved and the alcohol is evaporated, and then dried to constant weight in the air-oven at 1 10 C. (Gladding.)* A more rapid determination for technical purposes may be made by placing 5 grm. of the soap in a large porcelain crucible on the sand-bath, heating with a small flame, and stirring con- tinuously with a rod weighed in with the crucible, and rough and jagged at the end to break up the lumps formed. The operation is complete when, after removing the flame, a cold glass plate placed on the crucible is no longer dewed, which should require about half an hour. Great care must be taken not to burn the soap, which is indicated by the smell. Results, if carefully done, are accurate to per cent.f In place of drying alone, a known weight of oil free from water or of known water- content may sometimes be added with advantage. Lewkowitsch points out that actual water determination is rarely important because of the uncertainty of sampling, and that it generally suffices to estimate it from the deficiency found on adding up the constituents determined by analysis. Where determination is needed, he recommends cutting the soap into fine shavings, and weighing in a porcelain dish with a glass rod with which to break up the flakes. The first part of the drying should be conducted at a low temperature insufficient to melt the soap. Soap as manufactured contains about 30 per cent, of water if much less than this is found, the soap has been intentionally or accidentally dried ; if more, water has been added, as is very frequent. Determination of Fatty Matter and Alkali. About 10 grm. of the soap is weighed into a 200 c.c. beaker and dissolved in about 100 c.c. of hot water on the water-bath with constant stirring. A few drops of methyl orange and 50 c.c. of N/i HC1 (or more if the acid is not in distinct excess) is gradually added, and the heating and stirring continued till the fatty matter collects in a clear oily layer on the top of the liquid. The glass rod is rinsed with hot water and the heating con- tinued till the fats have again collected in a mass, which is * Chem. Zeit., vii. 568. t Watson Smith, Jour. Soc. Dyers and Colorists, i. 31. ANALYSIS OF SOAPS. 279 then allowed to cool and solidify. The cake is detached from the beaker with a clean spatula, the acid liquid poured carefully into another beaker, and the cake of fat rinsed with cold water which is added to the acid solution. The fat cake is cautiously dried with filter paper and placed in a desiccator in a carefully weighed basin or clock-glass, any fat which can be scraped off the beaker being added to it ; and after it has dried for some time it is weighed. If the fatty matter is not solid when cold, or in any case if it is not to be subjected to further examination, it is con- venient to add 5 grm. of dry beeswax or stearic acid before cooling, and to deduct it from the weight. The beeswax is conveniently weighed in the same basin in which the fatty matter is to be weighed and added to the tare. Another rapid method susceptible of tolerable accuracy is to separate the fatty matter from the aqueous solution in a tap burette and wash it with hot distilled water. The burette is then heated in a deep beaker of boiling water, or in steam ; the volume of fatty matter in c.c. multiplied by 0*85, the average specific gravity of the fatty acids at 1 00 C, gives its approximate weight in grams. (For more accurate methods, cp. p. 314) The acid aqueous solution is titrated with methyl orange and N/i sodium hydrate or carbonate solution, the amount of which deducted from 50 c.c. gives the acid neutralised by the total alkali of the soap, which in hard soaps may be calculated as Na 2 O, and in soft soaps as K 2 O. If the acid solution is turbid it must be filtered before titration. An insoluble sediment at the bottom of the beaker consists of silica or " filling," such as china clay, insoluble in hydrochloric acid. Chalk or whiting is of course dissolved if present, and estimated as alkali. The " total alkali " includes that combined with the fatty acids as soap, together with any alkalies present as hydrates, carbonates, bor- ates, or silicates. Further information on these points is obtained by repeating the analysis by the alcohol method described below. A preliminary test may be made by moistening the freshly cut surface of the soap with an alcoholic solution of phenolphthalein. A pink coloration is produced by the presence of caustic alkalies, and also, if the soap is not too dry, by carbonates, silicates and borates. The fatty matter consists of the fatty acids of the soap, together with resin acids (p. 316) if resin has been used in the 280 LEATHER INDUSTRIES LABORATORY BOOK. manufacture, sometimes traces of fats which have escaped saponi- fication (or rarely, traces of free fatty acids), and unsaponifiable matter such as petroleum products purposely added, or choles- terol and higher alcohols, derived from the use of " recovered grease," or the addition of lanolin. These can be separated to a certain extent by the methods given under oil analysis, pp. 309 et seq. For this purpose it will be necessary, if wax has been used, to prepare a fresh quantity of fatty matter, which need not be done quantitatively ; but the liquid fats may be separated from the aqueous liquid by a tap funnel or burette, and a portion weighed out for analysis. Soluble fatty acids will dissolve to some extent in the acid liquid, from which the bulk may be recovered by saturation with common salt, but, except in the case of cocoa-nut and palm-nut oil soaps, they may generally be neglected. If a qualitative examination reveals no unsaponified or unsaponifiable matter, the fatty matter may usually be assumed to consist of fatty and resin acids. If required to add up to 100 in a complete analysis these must be calculated to fatty anhydrides, which may usually be done with sufficient accuracy by multiplying their weight by 0*965. The alcohol method may be employed as a check on the above, and gives considerable additional information. 4 or 5 grm. of the soap, which if much watered should be somewhat dried after weighing, is dissolved in 50 c.c. of hot absolute alcohol. Any insoluble residue, consisting of mineral and organic "fillings," together with carbonates, silicates, borates, and sulphates, is filtered off on a dried and weighed filter, well washed with hot alcohol, dried at 100 C. and weighed. Organic matter may be determined in this residue by ignition, and bases existing in combination with organic acids, or as carbonates, borates or silicates, by solution in standard HC1 and titrating back with soda, the indicator being methyl orange or lacmoid paper. The alcoholic filtrate is received in a narrow-necked flask to avoid carbonation by the air ; phenolphthalein is added and the solution titrated with N/i HC1, till the red colour is discharged, and the result calculated as caustic alkali. After distilling off the alcohol, and free dilution with water, methyl orange is added and the titration continued to redness, the additional acid used corresponding to the alkali present as soap. It'sometimes hap- pens that no redness is produced with phenolphthalein from the ANALYSIS OF SOAPS. 281 total absence of caustic alkali, and in this case free fatty acids may possibly be present. In this case the solution should be tested with a single drop of N/io NaOH. If this produces no redness, the addition may be continued till redness is produced, and the results calculated as " free fatty acids," in terms of oleic. For molecular weights of fatty acids, see p. 302. SECTION XX. OILS AND FA T S* GENERAL CHEMISTRY OF OILS AND FATS. To the chemist, oils and fats constitute a single class, merely differing in melting points and consequently in consistence many which are solid fats in temperate climates being liquid oils in the tropics. All the true fats and oils are salts or " esters " of certain organic acids with glycerin ; but the closely allied waxes, together with sperm oil and a few liquid waxes usually classed with oils, are similar esters in which members of the alcoholic series are combined with the acids, replacing the glycerin. There are also some hydrocarbons, such as paraffin wax and oil, vaseline, etc., which, though of quite different chemical character, are popularly classed with the oils and waxes ; and so-called " volatile oils" of very varied constitution which are usually the odorous constituents of plants. Glycerin, as might be surmised from its solubility in water and its very sweet taste, may be viewed either as one of the simplest type of sugars, or as a trihydric alcohol, its CH 2 (OH) formula being CH(OH) or C 3 H 5 (OH) 3 , and each of the three CH 2 (OH) hydroxyls is capable of combining with a molecule of organic acid with elimination of water, just as hydrochloric acid com- * Cp. Lewkowitsch, ' Oils, Fats and Waxes ' ; and Schmitz-Dumont, Ding. Polyt. Journ., 1895, ccxcvi. Heft 9-11. 282 LEATHER INDUSTRIES LABORATORY BOOK. bines with sodium hydrate, HC1 + NaOH = NaCl + OH 2 . The neutral fats are therefore triglycerides, each molecule of glycerin being linked to three acid groups, which may be all similar, though there is good reason to believe that mixed esters exist, such as oleo-distearin. (The names of neutral fats are terminated in " in " as " stearin," which is glycerin tri-stearate to distin- guish them from the free acids such as stearic acid.) Since all the true fats are glycerides, the differences between them depend entirely on the acids ; and these owe their acid nature to the CO. OH group which they contain, and their differences to the other groupings with which this is linked. All acids naturally present in fats are monobasic, or have one CO. OH group only. The acids of the ordinary hard fats belong to the series, often described as the "fatty series," of which formic and acetic acids are the simplest members. Formic acid is j co.OH ; acetic CO.'OH ; P r P ionic I CO. OH ; so ( C H rising by successive additions of CH 2 to palmitic acid j ~Q QTT, 1C H CO OH ' wn i cn are tn e principal acids of hard tallow stearines. (The so-called " distilled stearines " from recovered greases are not true stearins, but consist mainly of their free fatty acids unaccompanied by glycerin.) In such a " homo- logous " series, the boiling and melting points usually rise with the increasing chain of carbon atoms, so that while formic and acetic acids are volatile liquids, the acids higher than caprylic, C 7 H 15 CO.OH, are solid at ordinary temperatures, and their solubility in water diminishes as their melting points increase, caprylic acid being nearly insoluble in cold water. H From the common formula of this series, H C . . . C OH A o 'it will be observed that all the valencies or combining bonds of the atoms are employed and satisfied ; and that consequently such acids are " saturated " compounds, which can only form new com- pounds by "substitution," and not by "addition" of new ele- ments, and they are therefore stable substances not easily broken up. OILS AND FATS. 283 There is, however, a second series of acids which we may call H H I I the " oleic " series, in which the group C = C occurs some- where in the carbon chain, and in which the double link can be easily broken to form such compounds as CHOH CHOH or CHI CHI in place of CH = CH . Thus oleic acid is C 17 H 31 CO.OH, differing only from stearic acid by the 2H corresponding to the double link. In the laboratory, though not as yet in practical manufacture, it is easy to replace this 2H, converting oleic into stearic acid. The glycerides of acids of this series are all liquid at ordinary temperatures, and " olein " or glycerin trioleate forms a large part of most non-drying oils, and, with stearin and palmitin, of such fats as tallow and lard, which are the softer the more olein they contain. It is obvious that more than one double link may be contained in the carbon chain, and, from linseed and fish oils, acids have been isolated containing three or even more double links. The fats of such acids are usually liquid, and absorb oxygen readily from the air, becoming converted into gummy or resinous bodies. Hence oils containing them have the property of " drying " more or less rapidly in the air, becoming converted into solid or sticky masses. Many oils, such as cotton-seed, are known as " semi- drying oils," thickening and gumming in the air from the presence of acids less saturated than oleic. The total unsaturatedness of an oil is easily determined in the laboratory, by ascertaining the quantity of iodine necessary to satisfy the unsaturated bonds ; but in mixtures of acids of various series, such as constitute the natural oils, it is not usually possible to determine accurately whether the unsaturation is due to a large quantity of a slightly unsaturated acid, like oleic, or a much smaller proportion of a much more unsaturated acid, such as some of those of linseed or cod-liver oils. Some attempts in this direction are referred to in footnote, p. 331. Fatty acids also exist in which one or more of the H atoms in the chain are replaced by OH groups, and an acid of this type of some importance is ricinoleic, the acid of castor-oil, which is hydroxyoleic acid. Ricinolein is insoluble in petroleum spirit, and most other oil solvents, but soluble in alcohol, showing in this way the effect of the hydroxyl. When unsatu- 284 LEATHER INDUSTRIES LABORATORY BOOK.* rated oils, such as olein and the semi-drying oils, are oxidised by blowing air through the heated oil, oxygen is absorbed, and fats of oxy- and hydroxy-acids are produced in many ways resembling castor-oil, which such " blown-oils " are frequently used to adulterate. Somewhat similar reactions, but more pronounced from the greater unsaturation of the oils, take place in the manufacture of degras from marine oils, in chamois- ing, and in the conversion of linseed-oil into linoleum-mass by oxidation. The solid waxes, and liquid waxes such as sperm oils, con- sist, as has been stated, of fatty acids in the hard waxes, mostly high in the stearic acid series ; as, for instance, cerotic and melissic acids from beeswax, which contain 26 and 30 atoms of carbon respectively, combined with high members of the alcohol series. Ordinary alcohols consist of a hydrocarbon chain similar to the acids, but terminated simply by the group CH 2 . OH instead of CO. OH. They have, therefore, no acid properties, and when treated with alkalies are unaffected, and hence appear in waxes and sperm oils as " unsaponifiables " (p. 309), but must not be confounded with the petroleum products and rosin oil which are frequently used as oil adulterants, and which are also unsaponi- fiable. The structure of the petroleum products, such as benzoline, petrol, paraffin oils and waxes, may be still simpler than that of the alcohols ; consisting merely of CH 2 chains, terminated at each end by CH 3 , the lightest petroleum spirit having about six carbon atoms, and the hard waxes averaging about eighteen, like the solid fatty acids. There are, however, many complications, such as branched chains (which may also occur in fatty acids and alcohols), and unsaturated compounds in which double links occur, or the two ends of the chain may be united to form a ring-compound. The above sketch, while it may assist the less chemical reader to understand the analytical processes which are now to be described, is of course, exceedingly imperfect in detail, for which the reader is referred to larger works, and especially to that of Lewkowitsch.* * ' Oils, Fats, and Waxes ' (Macmillan, 1904), 3rd edition. OILS AND FATS. 285 A. DETERMINATION OF TOTAL FATTY MATTER, AND NON-FATS. It is in many cases sufficient for the purposes of the leather manufacturer to determine the total amount of fatty matter present in a leather sample or elsewhere ; and in others its extraction is a necessary preliminary to further examination. The most usual means of accomplishing this is by extraction with a solvent ; and for general purposes the most convenient apparatus, notwithstanding its fragility, is the ordinary form of Soxhlet tube. The arrangement of apparatus will easily be understood from the figure. A is a cylindrical glass vessel which contains the material to be ex- tracted, and which is fitted above to a vertical condenser B, so arranged that the liquefied solvent drips back on the material. A is fitted with a syphon C, which comes into action when the liquid rises high enough to cover the material, and returns it to the boiling flask D at intervals. A side tube, E, allows the vapour to pass up again to the con- denser. A slender tube fitted to the top of the condenser lessens the diffusion of uncondensed vapour into the laboratory. The whole is fitted together with good corks, and it is best to deprive these of fatty matters by boiling a little of the solvent in the apparatus for some time before actually taking it into use. With the solvents used for oils, a hot solution of gelatin and glycerin in a little water may be applied to the joints if they cannot otherwise be made tight. The flask D is best of a spheri- cal form, as in the writer's experience conical flasks have proved liable to fracture. It should be heated in a water-bath, preferably heated by steam or an electric resistance, and placed as far as pos- sible from open flames and inflammable surroundings in case of accident. The flask should not come in contact with the bottom FIG. 19. 286 LEATHER INDUSTRIES LABORATORY BOOK. of the water-bath, but may rest on a piece of perforated zinc or wire gauze, or be supported by a clamp. There is considerable choice of solvents, but those most usually employed are ordinary ethyl ether, petroleum ether, and carbon disulphide ; and of these the latter is generally to be preferred, as its vapours, though poisonous and unpleasant, are heavy and easily condensed, and therefore little liable to escape into the laboratory. If petroleum ether is used, it must be carefully fractionated to contain no portion boiling above 75 C., and the quantity used must not be too small, as otherwise it is very liable to cause accidents from the temperature rising in the flask much above that of the solvent in the extractor which suddenly syphons down into it. The fractionation is made easier by the addition of about 5 grm. per liter of vaseline or paraffin wax to the com- mercial ether before distillation (Schmitz-Dumont), and may be performed in an ordinary distillation flask with Liebig condenser, but is made much more perfect by the use of one of the various forms of dephlegmating column (see p. n). Carbon disulphide should be kept in the dark, and occasionally distilled to purify it from dissolved sulphur.* Probably now that it is attainable at a moderate price, carbon tetrachloride is to be preferred to any of the solvents named from its non-inflammability. It boils at 78 C. and dissolves all ordinary fats, but also some substances such as lecithin which are not soluble in petroleum ether. The material to be extracted must usually be finely divided. Leather is cut into shavings, while other substances (e.g. egg- yolk, degras) are often mixed with sand or other indifferent bodies to make the mass more permeable. It is generally de- sirable to dry before extraction. In many cases it is sufficient to put the material for extraction direct into the Soxhlet tube, a little cotton wool (previously freed from fat) or glass wool being pla:ed below and above it to prevent particles being carried over by the syphon. Finely powdered materials may be placed in a sort of thimble of filter paper, either specially made for the pur- pose (Schleicher and Schiill) or formed by folding over the end of a suitable test-tube. It is often convenient, instead of this, to use an inner tube formed by blowing a hole in the bottom of a large test-tube, which is closed with a tuft of cotton or glass wool. The flask D is carefully weighed, and supplied with enough * On other forms of apparatus, compare Lewkowitsch, p. 148, 3rd edition. OILS AND FATS. 287 of the solvent to fill the Soxhlet about once and a half, or twice ; the apparatus is put together and heated till it syphons regularly once in 8 or i o minutes, and this is allowed to continue till the material is judged to be exhausted. To test this, the flask may be removed and a second weighed flask substituted with fresh solvent. After completion of the extraction the solvent is dis- tilled off both flasks and recovered, and the flasks are dried in the water- or air-oven till constant, which is facilitated bypassing a gentle current of air or coal gas through them, and they are then cooled and weighed. If there is any gain in the second flask it is added to that of the first, but by noting the approximate number of syphonings required for each class of work, a second flask is usually unnecessary. Lewkowitsch uses a small tap on the outer limb of the syphon to draw samples and ascertain when the extraction is complete, and this is undoubtedly convenient,, but adds materially to the cost of the apparatus. Where rapid work is required, it is often advantageous, after distilling off most of the solvent, to transfer the remainder to a weighed basin,, rinsing the flask with a little more solvent, and evaporating to dryness on the water-bath in a draught closet. Absolute con- stancy is not to be expected, as on the one hand oxidation increases the weight, and on the other there may be slight loss from volatilisation. Where the matter to be analysed consists mostly of fats, it will generally be possible, after evaporating the water (see below),, to filter a sufficient part for further analysis through a weighed filter, which can then be completely extracted in the Soxhlet apparatus, and the insoluble matters determined. If the material is too thick to filter in this way, even in a warm water-oven, it may be thinned with a solvent sufficiently to enable it to be filtered, and this filtrate added to that from the extractor before distilling off the solvent. If the quantity of solid matter and the filter employed is small, it will often be sufficient to simply wash the filter with small quantities of petroleum ether, which in this case is the most convenient solvent, until a drop or two of the filtrate, evaporated on a watch-glass, no longer leaves a residue. The insoluble residue of fats, and especially of degras and other stuffing greases, may contain beside organic matter acci- dentally present, and salts and acids from the process of manu- 288 LEATHER INDUSTRIES LABORATORY BOOK. facture considerable quantities of soap, either added or formed by saponification of the fats, and mineral fillings, such as talc .and china clay. Determination of the ash will,' of course, give the mineral matter present. Soaps may be estimated from the amount of alkali found in the ash, or the fatty acids separated as in Section XIX. The methods of determining fats in egg-yolk and in degras (pp. 337 and '352), may be referred to as good examples of dealing with rather difficult cases. Determination of Water in Oils and Fats. If the oil or melted fat is quite clear and transparent, it seldom contains any water or other impurities which for technical purposes need be removed before analysis, but sometimes, on heating, albuminous and gelatinous matters which have apparently been dissolved in the oil, are precipitated and separate. To determine water, 2-3 grm. of the fat or oil is weighed into a platinum crucible without cover, and placed on a triangle in a somewhat slanting position. It is then gently and slowly heated by a small flame moved about beneath it. In presence of much water the liquid boils gently, but the last particles become superheated and escape with slight crackling and a tiny puff of smoke, indicating that the flame must be withdawn. If the crucible be rapidly cooled by placing it on a block of metal, the whole estimation can be performed in fifteen minutes.* Water may also be determined by drying on sand as de- scribed for egg-yolk, but the results are scarcely more accurate on account of oxidation, while the process is much more tedious. Where the quantity of water is not large, it is best simply to dry the fat at no-i2O C., or even at 100, in a small beaker with a stirring rod ^weighed in, stirring frequently till the weight is constant. Where the determination of water and insoluble matters is not necessary, a sample may generally be purified sufficiently for analysis by simply filtering through a dry paper in a warm place, such as a water-oven. Determination of Ash. This is, as a rule, only necessary where the presence of salts or soaps is suspected, as in degras. Some samples of oil and degras contain traces of iron, which darken leather and give the ash a reddish tint, and may be estimated by the colorimetric process described jon p. 40. * Fahrion, Zeits. fiir angew. Chem., 1891, p. 172. OILS AND FATS. 289 10 or 20 grm. of the fat is placed in a shallow platinum basin, which is slightly inclined, and the side only is heated by a small flame. The water first escapes without spurting, and the vapours then take fire and the fat burns quietly away (compare p. 288). Detection and Estimation of Free Mineral Acids. These (and especially sulphuric acid, which is extremely injurious to leather) are often present in degras and sod oil. If the sample has a strong acid reaction, 25 grm. are boiled with 200 c.c. of water, the aqueous layer separated by a separating funnel and made up to an even quantity, of which an aliquot part is titrated with sodium carbonate solution, using methyl orange as indicator to prevent the interference of soluble fatty acids. In another part of the solution, the nature of the acid should be qualita- tively determined. B. PHYSICAL EXAMINATION OF FATS. In default of marked chemical characteristics, the physical constants of fats are often of great value in their recognition and the detection of admixture, especially as in many cases they admit of easy determination on very small quantities. The approximate constants of many of the fatty matters which are most important to the leather manufacturer .are given in the tables on pp. 349-350. Specific Gravity. That of oils is conveniently taken in an ordinary pycnometer, special care being taken as to temperature, as oils expand rapidly with heat. The outside of the bottle may be cleaned if necessary with a little petroleum ether before weighing ; and after use it may be washed out with very weak caustic soda solution, then with distilled water, and carefully dried. The Sprengel tube is, however, to be preferred for accurate work, especially for small quantities. This is shown in Fig. 20, and consists in a U-tube of thin glass with two capillary ends, of which one, B, should be of about J-mm. and the other, A, of -mm. bore. On the wider, a mark is made at M with a writing diamond. The tube is filled by connecting A with a glass bulb and rubber tube, and sucking the oil by B till it flows into the bulb. The ends are then wiped, the oil is brought to the requisite temperature by immersing the tube in water, and U 290 LEATHER INDUSTRIES LABORATORY BOOK. adjusted to the mark by applying a piece of filter paper to the narrow end if oil is to be withdrawn, on a drop of oil on a glass rod if more is required. The tube is now weighed and the specific gravity calculated in the ordinary way ; its capacity having been previously deter- mined by weighing it first empty and then filled with water. Specific gravities are also conveniently taken by the plummet method, especi- ally with Mohr's or hydrostatic balance (cp. p. 8). Hydro- meters should only be used for very rough determinations. The specific gravity of solid fats may be determined as on p. 7 ; but this deter- mination has been largely superseded by that of the melted fats at 100 C, which is conveniently made in a Sprengel tube immersed in boiling water or steam in a beaker, or the neck of a flask,, covered with a watch-glass, so that the capillary ends only protrude. After boiling for twenty minutes, the volume of fat is adjusted, and the tube dried, cooled and weighed. Instead of the Sprengel tube, an ordinary small pycnometer may be used, suspended in steam with a wire sling. Any fat which is expelled is carefully removed from the hole in the stopper with a piece of blotting paper, -while the flask is still in the steam. The flask is filled with liquid fat at a temperature below 1 00 C. The plummet method with Mohr's balance may also be used by placing the cylinder in a water-bath so arranged as to protect the balance from steam. The specific gravities are usually compared to that of water at 15 C. or 60 F. The sp. gr. of oils in glass vessels varies about 0*00064 for i C. ; but in whale oil and liquefied fats the change is somewhat greater. Some chemists have determined the sp. gr. of oils at FIG. 20. OILS AND FATS. 291 100 C. as compared to water at the same temperature, which is easily done by weighing the tube filled with water at the same temperature as is used for the fat. Such determinations will be higher than those with water at 15 and fat at 100 in the pro- portion of I "04 : i. The sp. gr. of oils at 15 C., as compared to water at the same temperature, is about 0*00025 higher than when both are at 1 5 5 C. A method of general applicability to very minute quantities of either solid or liquid fats at ordinary temperatures, is to suspend globules or small fragments in a mixture of strong and very weak alcohol, of which the proportions are varied till the fat globules neither sink nor rise to the surface. The sp. gr. of the liquid is then precisely the same as that of the fat, and can be determined in any convenient way. As the sp. gr. of alcohol solutions of known strength have been accurately tabulated and determined, this can be done volumetrically. An alcohol 50 per cent, by volume is of gravity 0*9335 at 15 '5 C. (60 R), and consequently just heavier than the heaviest oils, while a 70 per cent, alcohol is sp. gr. 0*8892, so that mixtures of these cover all oil gravities. Small volumes of alcohol are sufficient, and can be recovered by distillation. Globules of fat can be conveniently formed by dropping into cold 60 per cent, alcohol, or even on a damp glass plate, but they must be allowed to remain on blotting paper 18-24 hours before use to recover their original density (cp. p. 7). Air-bubbles, either adhering or enclosed, must be carefully avoided. The Melting and Solidifying Points of fats and oils are often very characteristic, though from the fact that these bodies are usually mixtures of various constituents of different melting- points, the results are frequently somewhat indefinite. They are also in many cases considerably affected by the previous treat- ment of the sample. Thus, a tallow which has been melted at a high temperature has its melting-point considerably lowered, but regains its normal one on being remelted at a temperature little above that necessary for fusion and again cooled. In many cases the melting-point is varied by the proportion of free fatty acids present, and some sources of error are obviated by taking the melting or solidification point of the freed fatty acids in place of that of the original fat. Prior to making the determination, the fats must be carefully U 2 292 LEATHER INDUSTRIES LABORATORY BOOK. freed from water and solid impurities. This is generally suffi- ciently effected by keeping them gently melted for an hour or two, and filtering through dry paper. Fatty acids must be purified from alcohol and soluble acids by boiling for an hour with a large volume of water, and repeatedly washing with hot water. Fats do not regain their normal melting-point for at least twenty-four hours after solidification. The determination of the melting-points of solid fats is usually made as follows. Thin tubes are drawn out to an open capillary point of 0-5-0*3 mm. bore, and sealed at the capillary end. A minute angular fragment of the solid fat is dropped into the tube, so that it rests on the shoulder of the capillary part, and the tube is attached with an elastic band to a delicate thermometer, so that the fat is level with the bulb, and immersed in a good-sized beaker of water, previously boiled to expel air-bubbles, which is heated by a small flame so that the temperature rises very slowly as it approaches the melting temperature of the fat. The water is continuously stirred with the thermometer or a separate stirrer, and the point is noted at which the angular fragment loses shape and adheres to the side of the tube. Sometimes the melting is gradual, and after incipient fusion the fat is still thick and turbid. In this case the temperature at which the globule becomes clear must also be noted. The method is applicable to pasty fats, like bone grease and dubbin, consisting of mixtures of liquid and solid constituents, but the temperature of clear fusion only can be noted. In many cases, and especially in that of the fatty acids, the temperature of solidification is more definite than the melting- point, but is usually some degrees lower than that of complete fusion. If a solid fatty acid is melted and allowed to cool slowly while the temperature is observed, it will be noticed that the latter falls till the fatty acid begins to crystallise, and then rises some degrees from the latent heat of fusion which is liberated, remaining constant till solidification has taken place. With neutral fats this effect is less marked and definite, but at least a temperature can be observed which remains stationary for some little time during the solidification. As improved by Shukoff,* the observation is carried out as follows. A thin test-tube of about 3 cm. diameter is fixed by means * Chem. Zeit., 1901, p. 99. OILS AND FATS. 293 of a cork in a good sized wide-mouthed bottle, and a delicate thermometer capable of being read to 0*1 degree is fitted in it with a cork, so as to reach well down into the fat which is to be tested. The stem of the thermometer must be long enough for the graduation to be outside the cork at the point where it has to be read. The tube is next filled to a suit- able height with the melted fat, which must not be too hot for the graduation of the thermometer, which is now inserted, and the whole allowed to cool slowly. As soon as the fat begins to show signs of turbidity, the vessel is vigorously shaken, and the 60 55 50 45 40 10 20 30 40 SO 60 70 FIG. 21. PARAFFIN AND MONTAN WAX. 9O Mtrx temperature read from minute to minute. At the point of solidification, the temperature will either rise slightly, when the highest point reached is the required solidification temperature, or will at least remain constant for a short time at the required temperature. Of course, in fats and mixtures like dubbin, containing various constituents, there may be more than one solidification-point as this and the other constituent separates, and in such cases it is useful to continue the observations till the fat is fully solid, plotting them on a piece of curve-paper. A diagram obtained in this way by Graefe,* for a mixture of montan wax and soft paraffin is given as an example. Though the apparent melting-point by such methods as the capillary tube was 6 1 '2 C., this is only indicated by a slight kink in the curve * Chem. Zeit., 1904, p. 1144. 294 LEATHER INDUSTRIES LABORATORY BOOK. caused by the separation of most of the montan wax, and the real solidification-point of the remaining mixture is 42*6C., scarcely over that of the original paraffin. In this connection it may be noted that the true melting-point of such mixtures is always lower than the average of the constituents, and often lower than that of the one of lowest melting-point. As applied by Dalican to the examination of the fatty acids of tallow, a similar method, known as the " titer test," is largely used both in this country and on the Continent for commercial purposes, and is carried out as follows : Fifty grams of fat is saponified with 40 c.c. of potash solu- tion, sp. gr. 1*4 (or with 20 grm. of stick potash dissolved in 25 c.c. of water), and 40 c.c. of good alcohol, in a porcelain basin on the water-bath with constant stirring till the soap becomes pasty. It is then dissolved in a liter of water and boiled at least three-quarters of an hour to drive off the alcohol ; water is added if necessary, and sufficient sulphuric acid to de- compose the soap. The mixture is heated on the water-bath till the fatty acids separate in a clear oily layer, which is allowed to cool and solidify. The cake is perforated with a glass rod and the acid liquid poured off, and the cake is boiled repeatedly with fresh quantities of distilled water ; and finally dried, melted at a low temperature and filtered through paper into a porcelain dish, in which it is allowed to solidify and stand over-night in a desiccator. It is then carefully melted on the water-bath at as low a temperature as possible, and its solidification temperature taken by the method of Shukoff. Dalican gives a table, repro- duced in most books on fats, giving empirically the proportions of solid and liquid fatty acids, commercial stearic and oleic acids, corresponding to different solidifying-points. The solidifying-point, or point of commencing turbidity, can be taken like the melting-point where only small quantities are available, allowing the water in the beaker to cool slowly, and noting the temperature at which the fat becomes turbid. In the case of oils, freezing mixtures may be used. Index of Refraction. Where a refractometer is available, it is desirable to determine the refractive index, as this can be rapidly done on even a single drop of fat, and in some cases furnishes valuable information. Many refraction constants are given in the tables on pp. 349-350. For the principles of this OILS AND FATS. 295 measurement the reader is referred to elementary treatises on optics. The process differs somewhat with different instruments, and it is not necessary to describe it in detail, as working in- structions are always provided, Zeiss's Butyro-Refractometer, a modification of the instrument invented by the late Prof. Abbe, is most convenient, and allows of determination over a wide range of temperatures. FIG. 22. This instrument is shown in Fig. 22, and consists of two prisms of dense glass of acute angle, the face of the upper of which is shown at A, while the second is in the hinged part B. A drop or two of the oil to be examined is spread on the face of the lower prism, and the portion B is then turned on its hinge so that the oil forms a thin film between the faces of the two prisms. Light is now reflected by the mirror J through the 296 LEATHER INDUSTRIES LABORATORY BOOK. lower prism, and a portion passes through the oil into the upper prism, and thence to the eye-piece of the instrument, while another part striking the prism faces more obliquely, is " totally reflected " from the surface of the upper prism, and lost. The angle at which this takes place is dependent on the difference between the index of refraction of the oil and the glass ; the nearer the two approach, the more oblique being the angle at which the light can pass through. On looking through the eye- piece the field of view is therefore divided into a light and a dark portion, the boundary of the light being sometimes sharp, but often shaded into a bluish or reddish fringe. The field is crossed by a scale divided into 100 parts, and on this the position of the boundary is read. As the scale is an arbitrary one, and only applies to the butter-refractometer, while some other instru- ments give the index of refraction direct, it is often desirable to convert it into terms of the latter, which is a well known physical constant. For this purpose the following table is given, derived by interpolation from the numbers given by Zeiss. In order to insure the constancy of the indications of the instrument, a " normal fluid " is supplied by the makers, which gives a sharp and colourless boundary, which at I5C. corresponds exactly to 77 3 scale-divisions. By passing a current of water, warmed a degree or two above that required, through the nozzle D, and out at E, the temperature can be regulated to any required tem- perature up to 70 C, so that the refractive index of fats and fatty acids of high melting-point can be taken. For this purpose it is desirable to order a thermometer reading to 65 or 70 C., as that usually sent out with the instrument reads to 50 C. only. In addition to the refractive power, some information may be gained from the appearance of the boundary of the light part of the field. If it has a blue margin, it indicates that the dispersion of the liquid examined is higher than that of butter, for which the instrument is corrected, while an orange or red boundary indi- cates a lower dispersion. In cases where this margin is broad and undefined, a sodium flame (a Bunsen flame coloured yellow by salt, which may be introduced on a little spoon made of platinum foil, or by a tuft of asbestos moistened with salt solu- tion) may be used as a source of light, when the margin will be sharp, and the refraction measured will be that 'of the sodium or " D " line. A red or orange glass fitted into grooves provided OILS AND FATS. 297 TABLE FOR REDUCING SCALE OF BUTTER-REFRACTOMETER TO REFRACTIVE INDEX.* Scale. Index. Diff. Scale. Index. Diff. Scale. Index. Diff. Scale. Index. Diff. O I 4220 80 25 I'44I5 74 50 1-4593 66 75 i'4754 5& I 28 26 23 51 600 76 60 2 36 27 3 52 06 77 66 3 44 28 37 53 13 78 7i 4 52 29 45 54 19 79 77 5 1-4260 80 30 1-4452 72 55 1-4626 66 80 i'4783 58 6 68 31 59 . 56 33 81 89 7 76 32 67 57 39 82 95 8 84 33 74 58 46 83 800 9 92 34 81 59 52 84 06 10 1-4300 78 35 1-4488 72 60 1-4659 64 85 1-4812 56 ii 08 36 95 61 65 86 18 12 16 37 503 62 72 87 23 13 24 38 10 63 78 88 29 H 3i 39 17 64 85 89 34 15 i '4339 76 40 1-4524 70 65 1-4691 64 90 1-4840 56 16 47 4i 3i 66 97 9i 46 17 54 42 38 67 704 92 5i 18 62 43 45 68 10 93 57 19 69 44 52 69 17 94 62 20 1-4377 76 45 1-4559 68 70 1-4723 62 95 1-4868 54 21 85 46 66 7i 29 96 73 22 92 47 73 72 35 97 79 23 400 48 80 73 42 98 84 24 07 49 86 74 48 99 90 100 1-4895 below the lower prism may be substituted with convenience, but perhaps some loss of accuracy. Messrs, Zeiss also manufacture another and more complete form of the Abbe refractometer, which, instead of an arbitrary scale, has merely cross-wires in the eye-piece, to which the shadow-edge can be brought by rotating the prisms, the move- * The Refractive Index is given to the nearest decimal in the 4th place. The Differences represent the average value of the 4th and 5th places of decimals of the index for one scale-divison, for the succeeding five divisions. With reading to o-i of a scale-division, the index is uncertain to about I unit of the 4th place. igS LEATHER INDUSTRIES LABORATORY BOOK. ment being measured on a sector graduated directly to show refractive index. This instrument has the advantage of a much wider range than the butyro-refractometer, practically sufficient for the examination of all liquids, and is in addition, independent of the coloured fringes mentioned above, as it can be achroma- tised for liquids of any dispersion by a compensating prism fitted with a graduated circle, which enables the dispersion not merely to be compensated but measured. In some few cases this measurement can be utilised as a means of identifying oils.* It is frequently necessary to correct the refractive index for small differences of temperature from the standard adopted. Tolman and Munson f have found that the change in ordinary oils and fats is very regular, being about 0*000365 for i C. If the butyro-refractometer is used, the scale number must be cal- culated to refractive index before applying the correction. On examining the tables of oil constants, it will be noticed that in most cases the refractive index is roughly proportional to the gravity ; and of course, like the latter, variable with tem- perature. Some additional information is obtained by calculating a specific refractivity independent of gravity and temperature, which may be approximately done by Gladstone and Dale's formula, " ' = c, where c is the constant, n the refractive a index, and d the specific gravity ; and more accurately for the ;/ 2 I present purpose by the later formula of Lorenz, - 2 -3. For more detail as regards the application of this formula to oils, and a table for facilitating its calculation, the reader is referred to a paper of Procter's on " The Refractive Constant in Oil and Fat Analysis." % The main results are that gravity and refractive index are increased both by the presence or increased proportion of oxygen and of unsaturated linkages in an oil or fatty acid, but not in the same proportion, so that the refractive constant calculated by either of the formulae given shows an increase from unsaturated linkage, and a diminution from oxida- tion. Thus, castor oil on the one hand and linseed and fish oils * Cp. Procter and Holmes, Journ. Soc. Ch. Ind., 1905, p. 128. "The Oxida- tion of Oils," part I. t Journ. Amer. Ch. Soc., 1902, p. 754. \ Journ. Soc. Chem. Ind., 1898, p. 1021 ; and Collegium, 1904, p. 153. OILS AND FATS. 299 on the other, are among the highest both in gravity and refrac- tive index, but are easily distinguished by refractive constant ; that of castor being low (G and D 0-4987), from the presence of oxygen, while linseed and cod average about 0*518, from their unsaturated acids. On oxidation by " blowing " or exposure to air, as in chamoising, the unsaturated bonds are saturated by the absorption of oxygen, and in the case of a cod-liver oil examined the refractive constant fell to 0*4955. Thus the fall of this con- stant is a measure of oxidation. For further details on this point, the paper by Procter and Holmes above cited may be consulted. The refractive constant is a very valuable index of chemical con- stitution. It is very high in mineral oils, from the absence of oxygen, and in a less degree in sperm and other liquid (and solid) waxes, where the proportion of oxygen is low. Specific Temperature Reaction (Maumene, Thomson, and Ballantyne).* When oils are mixed with concentrated sulphuric acid, complicated reactions take place with evolution of consider- able heat. The amount of this is a measure of the chemical action, and depends both on the number of unsaturated bonds and that of OH-groups in the oil, while the "iodine value" (p. 306) depends on the former only. Hence, when an oil oxidises by exposure to the air, the heating with sulphuric acid is increased and the iodine number diminished. With close adherence to the prescribed method, the specific temperature reaction gives constant results, and is very rapidly performed, though it cannot claim the same amount of scientific accuracy as the iodine value. Both are in general terms an index of the proportion of drying oils and olein. A deep beaker about 2 inches wide is fitted into a nest of cotton-wool in another beaker or cardboard box, so as to lessen the loss of heat.f 50 c.c. of water is measured into the beaker, and its temperature carefully determined with a delicate ther- mometer. 100 c.c. of the strongest sulphuric acid of exactly the same temperature is now slowly run in from a pipette, while the mixture is constantly stirred with the thermometer, and the * Journ. Soc. Chem. Ind., 1891, p. 233. This reaction might with equal justice be classed as a chemical one, but as the physical effect only is observed, it may con- veniently retain its present place. Cp. Twitchell, Journ. Soc. Ch. Ind., 1897, p. 1002 ; whose method, however, has not proved capable of giving accurate quantitative results. t The Shukoff apparatus described on p. 292 may also be used. 300 LEATHER INDUSTRIES LABORATORY BOOK. highest temperature noted. If the acid is sufficiently strong for the purpose, the rise should amount to 42-44 C. The ex- periment is now repeated in precisely the same manner, with 50 grm. of oil in the same beaker, with the same acid and ther- mometer. The rise of temperature with the oil, divided by that with the water, and multiplied by 100 to avoid decimals, gives the " specific temperature." If any change whatever is made in the apparatus or acid used, the determination with water must be repeated. Equality of temperature is best at- tained by placing the vessels containing the acid and oil in a water-bath, which is generally at laboratory temperature, though with pasty fats, or where the reaction sets in with inconvenient slowness, it may be raised to 20, or even more. It is best to use a pipette for the acid, drawn out to a fine point so that it requires about a minute to empty ; and the temperature reached should not greatly exceed 60 C., or further reactions take place which render the results unreliable.* Since linseed, cod, and many other fish and seed oils, give a much higher rise, it is necessary to moderate the reaction, which is best accomplished by the addition of a proportion of heavy mineral oil, which gives but small rise of temperature, especially if it is previously purified by shaking with a little concentrated sulphuric acid and thorough washing. 20 grm. of linseed or fish oil to 30 grm. of mineral oil is generally a suitable proportion, and for semi-drying oils 30 to 20 grm. The heat evolved by the mineral oil cannot be determined directly, as the reaction is inconveniently slow, and its effect seems to vary with the proportion used. It must therefore be ascertained by experiment with the same propor- tion of a colza or rape oil, of which the rise of temperature has been carefully determined without mixture. If, for instance, the rise of temperature with 50 grm. colza amounts to 55, that for 20 grm. should equal 22, but a rise of 35 has been found for a mixture of 20 grm. colza and 30 grm. mineral oil. The rise due to the mineral oil is therefore 13, and this must be deducted in each case where that quantity of mineral oil has been used, and the remainder multiplied by 5 -f- 2 before calculating the specific temperature. The determination may be made with 25 grm. of oil and 5 c.c. of acid in a smaller beaker, and if divided b/a water deter- * Ellis, Journ. Soc. Chem. Ind., 1886, p. 150. OILS AND FATS. 301 minatiori made under the same conditions, should give the same specific temperature as the larger quantities. Olive or lard oil may be used instead of a mineral oil, and its evolution of heat may be directly determined, but a much larger proportion is required to produce the same lowering of temperature. Specific temperatures of various oils are given on p. 349. A similar temperature reaction, with the substitution of bromine for sulphuric acid,* has been employed by Hehner as a rapid substitute for the iodine value determination, but since the introduction of the more rapid volumetric methods of Wijs and Hanus (p. 308), these are generally to be preferred to either the bromine or the sulphuric acid reaction. C. QUANTITATIVE ANALYSIS OF FATS. In many cases where it is difficult or impossible actually to separate the various constituents of a fatty mixture, much information may be obtained from the determination of certain chemical constants or " values. " The most important of these are: the "acid value" or number of mgr. of KHO required to neutralise the free acids of I grm. of the fat ; the " saponification value," or mgr. of KHO required to form neutral soaps with the whole of the fatty acids of the same quantity ; the " ether value," obtained by subtracting the acid value from the saponi- fication value, and equivalent to the mgr. of KHO employed in decomposing and saponifying the neutral fats or glycerides ; and the "iodine value," indicating the percentage of iodine absorbed by the fat, which is a measure of its unsaturated " bonds." These values form the basis of much of the quanti- tative analysis of fats, but a few of their more important applica- tions may be mentioned here. The Acid Value, when high, either denotes that the fat is rancid and decomposed, or that it contains additions of artificially produced free fatty acids, as, for instance, those produced by distillation, or in the recovery of grease from soap solutions. It is stated by Gadd that cod oil with a high acid value is particularly prone to " spue " ; and by Wood, that such oils are * Analyst, xx. (1895), 146 ; cp. also Journ. Soc. Chem. Ind., 1897, 87, 193, 209. 302 LEATHER INDUSTRIES LABORATORY BOOK. generally though not always unsatisfactory for chamoising. Oils and fats with much free acid are unsuitable as lubricants, as they corrode metals. If the "mean molecular weight" or average combining equivalent of the free acids is known, their percentage is easily calculated from the acid value, by multiplying by this equivalent, and dividing by 561 (the eq. of KHO X 10) (see below). The free acid, however, is not always of the same character as that remaining as undecomposed glyceride. The Saponification Value, if known for any pure oil, affords the means of calculating approximately the amount of adultera- tion with mineral or rosin oils, which are unsaponifiable. Thus, pure cod oil has a saponification value of about 185. If a given sample only shows 166, we are justified in assuming that it con- tains about 10 per cent, of unsaponifiable matter. As, however, the saponification value varies a good deal even in genuine oils, it is always better, if the matter is of importance, to separate and weigh the unsaponifiable part (pp. 279, 309). It should be noted that distilled oils and greases and wool-fats contain naturally large proportions of unsaponifiable matter. The saponification value is a useful means of distinguishing different classes of oils. From it, the total fatty acids of the sample can be calculated if their molecular weight is known. That of stearic acid is 284, of palmitic 256, of oleic 282. The mean molecular weight of the mixed acids of tallow is about 275, and of cod-liver oil, according to Dieterich, about the same. The mean molecular weight of the free acids, insoluble in water, is easily determined from the saponification (or acid) value of a small quantity of the washed acids recovered after saponification (p. 306). If M = mean mole- cular weight, and K = saponification value, M = ^ . In is. the case of fatty acids, the acid value should be identical with the saponification value unless lactones (anhydrides of the hydroxy-fatty acids) are present, which is probably never the case in natural fats, but may occur where they have been treated with sulphuric acid or other dehydrating agents, or distilled at high temperatures. In this case, the fats have what is called a " permanent ether value," as the lactones are saponified by hot potash solution, but not by cold aqueous potash ; and again partially take their anhydride form, if the soap so formed be OILS AND FATS. 303. decomposed by acids (cp. p. 317). According to Ruhsam, they are found in degras.* The Ether Value gives, as has been stated, the amount of potash corresponding to neutral glycerides of the fatty acids. As most natural fats consist exclusively of glycerides with more or less free acid, the percentage of glycerin may be found by multiplying the ether value by 92, the molecular weight of glycerin, and dividing by 561 X 3 = 1683. Similarly, the fatty acids are found by multiplying by their molecular weight, and dividing by 561. The sum of the fatty acids and glycerin will be more than 100 per cent, by rather over 6 per cent., on account of the hydration of the acids and glycerin, as will be seen by the following equation, R being the radical of any fatty acid, R.OH:- C 3 H 5 3 .R 3 + 30H 2 = 3 R.OH + C 3 H 8 O 3 Neutral fat Water Fatty acid Glycerine The Iodine Value is a measure of the acids with unsaturated carbon atoms, and of course where only one such acid is present,, as oleic in tallow or lard, its amount can be calculated from the value. The theoretical iodine value for oleic acid, with which the experimental closely agrees, is 90' 07, and for olein, its trigly ceride, 86 2. The percentages may therefore be found by the following formulae, given by Benedikt, I being the iodine value : Percentage of oleic acid = = i 1601 X I. 90-07 r , 100 + I Percentage of olein = ^- = I 1 102 X 1. oO ' 2 Where several unsaturated acids are present, their quantities cannot be calculated in this way ; but the iodine value will still enable us to calculate approximately the proportion of two constituents for which it is known. Thus a dubbin composed of equal parts of a cod oil of 160, and a beef tallow of 40 iodine value, would have itself an iodine value of 100. We may also safely conclude that any oil with an iodine value above 86 must contain glycerides with more unsaturated bonds than oleic acid, and hence probably of a drying character.. * 3 Jahresbericht der Deutschen Gerberschule, Freiberg, p. n. 304 LEATHER INDUSTRIES LABORATORY BOOK. The converse, however, would not necessarily be true, on account of the probable presence of saturated fats, such as stearin and palmitin. Determination of Acid- Value. In a wide-necked conical flask of 250-300 c.c. capacity put 50 c.c. of good alcohol, and I c.c. (or 20 drops) of a I per cent, alcoholic phenolphthalein solution ; and add a drop or two of N/io KHO or NaHO till it is faintly pink. Weigh the oil or fat in a small beaker with a glass rod, transfer about 3 grm. (or more if the oil is known to be nearly neutral) to the flask, and determine the weight taken by re- weighing the beaker. Add N/ioKHO or NaHO from a burette very cautiously to the flask till the pink colour no longer disappears at once on vigorous shaking, and note the quantity used. If solid fats are titrated, the flask must be gently warmed, so that the fat is kept just fused during the operation. The acid value is obtained by multiplying the number of c.c. of N/iO alkali by $'6i, and dividing by the weight in grams of oil or fat. In cases where the quantity of oil available is very small, it is often convenient to use standardised alcoholic potash, in which case the same quantity of fat can be used for determination of saponification value (see p. 306). Determination of Saponification- PW#(K6ttstorfer). A solu- tion of alcoholic potash is made by dissolving about 8 grm. of potash "pure by alcohol "in 10 c.c. of water, and making up to 250 c.c. with good nearly absolute alcohol. Methylated spirit may be used if it is previously purified by shaking with powdered permanganate till strongly coloured, and after standing some time to deposit the hydrated manganese peroxide, is distilled off potassium carbonate, the first and last portions being rejected for commoner purposes. Another method, recommended by J. Carter Bell,* is to place 500 c.c. of strong methylated alcohol in a flask of 1000 c.c. capacity, and to dissolve in it 25 grm. of stick potash. 250 grm. of lard or other saponifiable fat is now added, and the mixture heated on the water-bath with vertical condenser till saponification takes place, when the condenser is turned downwards and about 450 c.c. of alcohol distilled off. Alcohol sufficiently pure for use 'gives a colourless solution of potash, which on keeping (in the dark), or on boiling, only turns * Journ. Soc. Chem. Ind., 1893, P- 2 3 6 - OILS AND FATS. 305 yellow and not brown. Ordinary commercial methylated spirit contains a proportion of light petroleum spirit, which if suffi- ciently volatile does not interfere with the use of the alcohol even for the separation of unsaponifiable matter. Fractions of petroleum of higher boiling-point might probably be removed by rectifying the spirit off paraffin or vaseline (p. 286), or very likely off lard. Mr. T. Mann purifies methylated spirit by the use of meta- phenylenediamine hydrochloride, of which 10 grams in fine powder is added per liter. The mixture is shaken up from time to time during two or three days, and then distilled, preferably with the aid of a fractionating column, about nine-tenths of the spirit being allowed to come over. Alcoholic soda of N/2 strength made up with methylated spirit purified in this way does not discolour on boiling, and is still quite fit for use after keeping a month in daylight. In fact it is only very slightly inferior to that made with the best rectified spirit. The alcoholic solution of potash is allowed to stand for twenty-four hours, to deposit potassium carbonate, and is then transferred (if necessary filtering through glass-wool to obtain a clear solution) into a bottle fitted with an india-rubber cork, through which is passed a 25-c.c. pipette closed at the upper end with a piece of rubber tube with a pinchcock. It should be kept in a dark and not too cold place. A quantity of oil or fat not exceeding 2 grm. is transferred as before into a flask of 150-200 c.c. capacity, which should be of Jena glass (or other hard glass not easily attacked by alkali), and 25 c.c. of the alcoholic potash is added, allowing the pipette always to deliver the same number of drops after it had ceased to run in a stream, so as to secure an absolutely constant quantity. A second 25 c.c. is run in the same way into a similar flask as a blank test, both flasks are closed with corks fitted with vertical glass tubes of perhaps 50 cm. long and 0-4 cm. bore, which should be cut or broken slanting at the lower end to facilitate dripping. The flasks are now placed in a shallow water-bath already boiling, and simmered for half an hour, the contents of each flask being frequently shaken with a rotatory motion. I c.c. of i per cent, phenolphthalein solution is now added, and the liquids carefully titrated back with accurately N/2 HC1, great care being taken not to overstep the point of neutrality. x 306 LEATHER INDUSTRIES LABORATORY BOOK. The difference in the two results represents the potash consumed in saponifying the fatty acids, both free and combined. Each c.c. of the N/2 HC1 corresponds to (56-1 -f- 2) mgr. of KHO, which multiplied by the number of c.c. of acid used, and divided by the weight of the fat in grams, gives the saponification value. In some cases, where only a small quantity of the fat is available, it is desirable to determine both the acid and saponi- fication values on one weighed quantity of fat. This is easily done by using alcoholic potash of known strength for titrating the acid value, and then adding the usual 25 c.c. of alcoholic potash and saponifying. In this case the potash consumed in saponifying will give the ether value direct. If 34 grm. of fat be used, with 50 c.c. of alcoholic potash for the saponification, it is possible to shake out the unsaponifiable matter as on p. 309, and afterwards to separate the fatty acids and examine them on a single portion of fat. Determination of Iodine Value. The original method of Hiibl has been superseded in most cases by the modifications of Wijs and Hanus, but is given here since many of the published data have been obtained by it, the points in which it has been varied by Wijs and Hanus being noted subsequently. Hiibl 's Method. The following solutions are required : 1. Iodine Solution (Iodine and Mercuric Chloride). 25 grm. of pure iodine are dissolved in 500 c.c. of pure " absolute " alcohol. Methylated spirit must not be used. 30 grm. of pure mercuric chloride are separately dissolved in a similar quantity of the same alcohol, and, if necessary, filtered. The solutions are best kept separately, but should be mixed in equal parts at least twelve hours before use, as considerable loss of free iodine takes place soon after mixing, but afterwards much more slowly. Some operators, however,* prefer to mix only at the moment of using ; and correct results may be obtained in this way if a blank test is made under precisely the same conditions to establish the strength of the iodine solution. The mixed solution is best kept in a bottle fitted with a 25 c.c. pipette as described for alcoholic potash. 2. Sodium Thiosulphate Solution. About 25 grm. per liter. If pure thiosulphate is at hand, a solution of 24*8 grm. per liter will be accurately decinormal. As, however, it does not keep * Fahrion, Chem. Zeit., 1892, p. 862. OILS AND FATS. 307 permanently, the strength must be determined from time to time. This is most conveniently done with standard solution of potassium iodate (to be had pure of Merck), but potassium bichromate may also be used (p. 258). ' One molecule of either of these salts liberates 6 atoms of iodine from potassium iodide in presence of HC1. Therefore 3- 566 grm. of KIO 3 or 4-91 grm. of Cr 2 K 2 O 7 made up to I liter, is decinormal as regards iodine of which it is equal to 12*7 grm. per liter, and 20 c.c. of either of these solutions allowed to stand for half an hour in a stoppered bottle with 10 c.c. of 10 per cent, potassium iodide solution and 5 c.c. of strong hydrochloric acid will liberate 0^254 grm. iodine, and equal 20 c.c. of decinormal thiosulphate. 3. Potassium iodide solution is made by dissolving 100 grm. in a liter of water. Commercial iodide not unfrequently contains iodate, which liberates iodine in presence of HC1. If this is the case, the amount of iodine liberated must be determined and allowed for. Hydrochloric acid also occasionally contains traces of free chlorine, to which the same remarks apply. 4. Chloroform should be pure, and not absorb any iodine ; but if the strength of the iodine solution is always ascertained by a blank test, in which the same quantities of potassium iodide and chloroform are used as in the actual determination, any errors due to these sources will be cancelled. 5. Starch solution is best prepared fresh for each analysis by mixing J grm. of starch into 100 c.c. of cold water, and raising to a boil with constant stirring. The clear solution is used. The determination is carried out as follows: 0*15 to O' 1 8 grm. of drying or fish oil, 0*3 to 0*4 grm. of a non-drying oil, or 0*8 to I grm. of a solid fat is weighed accurately into a bottle of 500 to 800 c.c. capacity, having a well-ground stopper. 10 c.c. of chloroform is added, and 25 c.c. of iodine solution measured in with the same precautions as in measuring the alco- holic potash for the saponification value. If the mixture, after shaking with the iodine solution, is not clear, more chloroform must be added. If after a short time its deep brown colour shows signs of becoming pale, another 25 c.c. of the iodine solu- tion must be added. The bottle is allowed to stand four to six hours protected from the light (some chemists allow 18 hours), after which 10 c.c. of potassium iodide solution and about 300 c.c. of water are run in. Should a red precipitate of mercuric iodide X 2 3 o8 LEATHER INDUSTRIES LABORATORY BOOK. be formed, it indicates an insufficiency of potassium iodide, of which more must be added. The mixture is now titrated with thiosulphate solution, which is run into the bottle and well shaken after each addition till the chloroform layer is only faintly coloured. A few drops of starch solution are now introduced, and the titra- tion continued very cautiously till the blue colour becomes permanently discharged. A blank test should be made at the same time with the same quantity of reagents, but without the fat, and similarly titrated, and the difference between the two will give the thiosulphate solution equivalent to the iodine absorbed by the fat, from which the iodine value is calculated by multiplying the corresponding weight of iodine by 100, and dividing by the weight of fat employed. If the reagents are all perfectly satisfactory, it is sufficient to titrate 25 c.c. of the iodine solution alone at the same time to obtain the original strength, but it is safer to carry out the blank experiment. Methods of Wijs and Hantis. These methods are simpler, more rapid, and much cheaper, and the solutions are much more permanent than that of Hiibl. The most convenient way of making the Wijs solution is to dissolve 5 grm. of iodine trichlo- ride in 500 c.c. of glacial acetic acid, and to add 5*4 grm. of iodine, which dissolves to a clear port wine colour. The Hanfts solution is similarly made by dissolving 6*6 grm. of iodine in 500 c.c. of glacial acetic acid, and adding i 5 c.c. (about 4 ' 2 grm.) of bromine to the cooled solutions. Both these solutions keep very well. Both are used in the same way and give approxi- mately equal results, those with Wijs, on the whole, being slightly the highest, while the Hanus solution is intermediate between that and the original Hiibl. For this reason, and because the solution is most readily prepared, the Hanus method has mostly been used in the author's laboratory. The process is carried out as follows : A suitable quantity of oil, varying from o* I grm., for fish or drying oils, to 0*4 grm. in the case of non-drying oils, is weighed into a stoppered bottle, and dissolved in 10 c.c. of pure chloro- form, and 25 c.c. of the iodising solution (either Wijs' or Hands') is run in, accurately measured, from a burette or pipette. It is found best to let the mixture stand for i hour before titrating, though for non-drying oils, considerably less' time would be sufficient. After standing the necessary time, 10 c.c. of 10 OILS AND FATS. 309 per cent potassium iodide solution is added, and the iodine liberated is titrated with N/io thiosulphate solution as described above, employing starch towards the end of the reaction. The first disappearance of the blue colour is to be taken as the end, though it may afterwards reappear through oxidation or dissocia- tion, and it is therefore desirable that the titration should be done as rapidly as possible, in order to avoid the change taking place during the process, and so introducing an error. The following table gives the results of comparison between the three methods which have been mentioned. Name of Oil. Hiibl. Hanfls. Wijs. Pale seal . .. -A ;'.-' .' JIT'S I20'7 J2 4 -8 Whale ... . . I2O"2 I2O'6 123-7 Olive. .... 82-3 8l- 9 83-4 Castor (cold drawn) . . ^ _ 82-6 84-4 85-6 Dogfish liver .... i34'7 I4I-2 I43-2 Arachis (earthnut) 91-8 91-6 93-4 Arctic shark .... 89-0 93'8 9 8- 9 Pale rape . . , 103-0 101-9 102 'I Coast cod ..... i48'5 i47'5 I54-6 Newfoundland cod 144-8 150-0 154-7 Linseed 174-8 I74-5 I77-3 Cottonseed . . .' . . 108-5 107-0 ^ IIO'O Determination of Unsaponifiable Matter. In the cases of oils and fats which are not likely to contain wool-fats, distilled "olein," or sperm or shark oil, 10 grm. are weighed into a flask, 50 c.c. of alcohol and 5 grm. of stick potash dissolved in 5 c.c. of water are added, and the flask is boiled for one hour with inverted condenser ; 50 c.c. of water are now added, and the solution is allowed to cool. It is then poured into a stoppered separating funnel of 150-200 c.c. capacity, and 30-50 c.c. of petroleum ether (boiling below 75 C.) are added, and gently mixed with the soap solution with a rotatory motion, avoiding at first violent shaking. If, on allowing the mixture to rest, the ether layer readily separates, somewhat more agitation may be given, but care must be taken to avoid the formation of a permanent emulsion. The mixture is now allowed to rest and separate (if 310 LEATHER INDUSTRIES LABORATORY BOOK. necessary for some hours), and the soap solution is run back into the flask in which it was saponified, where, to save time, it may be tightly corked and shaken with another portion of ether, and this time somewhat more vigorously, as the bulk of un- saponifiable matter is already removed, and allowed to rest while the ether in the separating funnel is washed with water. Care must be taken that, before washing, the whole of the soap solu- tion is removed, it being much less injurious in this stage to run off a drop of ether with the soap than to retain even a trace of the latter. 50 or 100 c.c. of water are now added to the contents of the funnel which are even more cautiously shaken than in the first instance, allowed to separate, and the water run off into a large flask or beaker, taking care in this case to avoid running away any of the ether. This washing is repeated twice, but the second and third time, the mixture being free from all but a trace of soap, may be vigorously shaken. Should an emulsion be formed which separates with difficulty, the addition of a few cubic centimetres of alcohol will often break it up. The washed ether is now run into a dry flask and well corked. The soap solution is poured back into the separating funnel ; a further por- tion of petroleum ether is added, and the process just described is repeated for a second and third time, the washed and separated ether being added to that in the c6rked flask. After the whole of the petroleum ether solution has remained in the dry flask long enough to deposit particles of water which it retains, it is poured into a weighed flask, the first flask being rinsed with a little fresh petroleum ether, which is added to the rest. The bulk of the ether is now distilled off and recovered, and the un- saponifiable matter dried to approximate constancy in the water oven, and weighed. Volatile unsaponifiable matter is, of course, lost, and in the case of mineral oils it is not well to push the drying too far. Wool-fat, spermaceti, waxes, and sperm, dolphin, shark-liver, and some other marine animal oils consist wholly or in part of compounds of fatty acids with higher alcohols instead of glycerin, and are usually only imperfectly saponified by the ordinary treatment with alcoholic potash. Complete saponification can be effected by digesting with double-normal alcoholic potash under pressure at 105 to 1 10 C., or by treatment with sodium ethoxide (prepared by dissolving 5 grm. metallic sodium in OILS AND FATS. 311 IOO c.c. of absolute alcohol), with subsequent digestion on the water-bath. The most generally practical method is to heat with 2 N alcoholic potash in a copper flask or tube with sciew stopper, or in a small soda-water bottle with wired cork. The latter should be enclosed in several thicknesses of canvas in case of explosion, and placed vertically in an iron pan with saturated salt solution, which is very slowly heated to its boiling point (about 109 C.), and maintained at this temperature for one hour. If an accurately measured quantity of potash solution be used, the saponification value can be simultaneously deter- mined.* After complete saponification, the higher alcohols remain as " unsaponifiable matter," and can be shaken out in the usual way. Sperm oil and spermaceti yield about 40 per cent, of these alcohols, and wool-fats less. In general, the "unsaponifiable matter" in wool-fats and recovered greases is determined simply by saponification in the ordinary way, and of course includes the undecomposed waxes. For many purposes this is sufficient, since the proportion sapo- nified at the ordinary pressure seems practically constant, amounting to 40 or 45 per cent, of the total fats. Where diffi- cultly saponifiable matters are suspected, it is desirable to con- tinue the saponification for two hours, and, in the case of marine oils, to shake out with ethyl-ether instead of petroleum, since it has been observed by Lewkowitschf that the higher alcohols are not easily taken up by petroleum ether. Where ethyl-ether is used, the greater part of the alcohol must be boiled off before shaking up, and the soap solution made up with water, but a little alcohol may be subsequently added if needed to break up emulsions. It will frequently happen with fats of the class named that a flocculent layer of difficultly soluble soaps of the higher acids will form below the ethereal layer. These must be retained with the soap solution. Flocks of similar soaps are often suspended in the ether layer, and will generally separate with the particles of water in the dry flask, but must otherwise be re- moved by filtration, and the filter washed with ether. Ethyl-ether dissolves larger proportions of neutral soaps than petroleum ether. Gawalowski \ states that neutral soaps are quite insoluble in petroleum ether, and recommends approximate neutralisation * Cp. Herbig, Ding. Polyt. Journ., 1894, ccxcii. p. 66. f Journ. Soc. Chem. Ind., 1896, p. 14 \ Zeit. Anal. Chem., xxvi., p. 330- 312 LEATHER INDUSTRIES LABORATORY BOOK. with calcium chloride, and addition of sodium hydrogen car- bonate with subsequent boiling. Good results may also be obtained by carbonating the caustic alkali by passing CO 2 through the solution, or by neutralisation to phenolphthalein with dilute acid, and the subsequent addition of a small quantity of sodium carbonate to restore alkalinity and combine with any trace of freed fatty acids. Soap may be detected in the " un- saponifiable " by incinerating, when it will be indicated by the presence of alkali in the ash, from which its amount may be approximately calculated. In place of shaking out, unsaponifiables may be extracted from the dried and powdered soap in the Soxhlet apparatus, and Baldracco * recommends this method for the analysis of degras. Details of this procedure will be found on p. 342. It is often of importance to determine the nature of the unsaponifiable matter. Apart from the presence of wool-fat, mineral oil, and less frequently rosin oil, added by way of adult- eration, are by far the most common. They can generally be recognised by their liquid consistence, colour and smell. Rosin oil is easily distinguished from mineral oil by its much greater sp. gr. and higher refractive index. Mineral oil ranges from 0*850 to 0*920 sp. gr. and I *477^ to l '498o ref. ind. : while in rosin oil, these are from 0*960 to 0*990 and from 1*5274 to I '5415 respectively. Rosin oils are miscible with acetone in all proportions, but mineral oils require several times their volume for solution. Both classes of oil are generally fluorescent, but a feeble green " bloom " or fluorescence is often noticed in animal and vegetable oils of undoubted purity, while that of mineral oils is frequently masked by addition of nitronaphthalene and other nitro-compounds. In this case it is restored by washing the oil with concentrated sulphuric acid. The ordinary colour tests for rosin oil are useless in presence of liver oils, or wool-fat, since very similar colours are produced by the cholesterol and lipo- chromes present in these fats, but in most cases resin acids may be detected among the fatty acids by the tests given on p. 316. The presence of large quantities of cholesterol is characteristic of wool-fat, but as this body, together with lipochromes, occurs in liver oils, the same remarks apply, and its qualitative detection is not of much value. When the more or less solid consistence, * Collegium, 1904, p. 333. OILS AND FATS. 313. the smell, or other causes lead to suspicion of the presence of wool-fats, the most satisfactory method is to resaponify the " unsaponifiables " obtained in the ordinary way, either under pressure, or by the aid of sodium ethoxide (p. 310). An additional amount of fatty acids and corresponding diminution of unsaponi- fiables, indicates with certainty the presence either of wool-fat, or of liquid waxes from some other source, and probably with further experience, it will be found that the amount of wool-fat can be inferred from the quantity. If the unsaponifiable matter is small in quantity, and more or less crystalline, it probably consists mainly of cholesterols and isocholesterol, or, in the case of vegetable oils, of the closely allied phytosterols. These bodies maybe detected by dissolving the unsaponifiable matter in a little chloroform, and shaking with an equal volume of concentrated sulphuric acid. Cholesterol and phytosterol colour the chloroform layer red or purple. Iso- cholesterol causes a strong green fluorescence in the sulphuric acid, and gives a yellowish tint to the chloroform layer which often masks the purple or pink due to cholesterol. Somewhat similar colorations to that due to cholesterol are also given by the " lipochromes " present in liver oils, and reddish colorations by resin acids (see p. 316) ; but the latter should be removed by saponification. Phytosterol may be distinguished from cho- lesterol by difference of melting-point and crystalline form of the two bodies and their acetates, under microscopic examination,, and their recognition is important in the detection of mixtures of animal and vegetable oils, but for details the reader must be referred to the books of Lewkowitsch and others. The high alcohols present in wool-fat and other solid and liquid waxes may also be present among the unsaponifiable matters, as well as undecomposed waxes, if the saponification has not been done under pressure. It is not easy to separate these from the cholesterols, and the reader must be referred for the best methods to Lewkowitsch (' Oils, Fats and Waxes'). Paraffin and ceresin may also be present among the unsaponi- fiable solids, and may be separated from cholesterol and other alcohols by digestion for two hours at 1 00 C. with acetic anhydride or less accurately by washing with ordinary alcohol, in both of which the paraffins are practically insoluble. The method is also applicable to liquid paraffins. If cholesterols or other 314 LEATHER INDUSTRIES LABORATORY BOOK. alcohols are present in large quantity, their acetylised products usually separate from the acetic anhydride as it cools, or in any case by pouring the solution into cold water. Fatty Acids. The soap solution, after removal of unsaponi- fiables and boiling off the alcohol, may be decomposed with dilute sulphuric acid, and the fatty acids separated as on p. 278. Care must be taken, in boiling off the alcohol, that the vapour which at first contains considerable petroleum ether, does not catch fire, and the operation is best performed on the water-bath ; or a distillation flask and condenser is used, and the alcohol re- covered. If it is desired to determine the acids quantitatively, it is convenient to wash them with hot water in a separating funnel, the wash waters being passed through a weighed filter of close texture, on which the fatty acids are finally collected. The wet filter is now placed on the'top of a small tared beaker and dried in the water-oven to approximate constancy, when the greater portion of the fat will filter through, and is ready for further examination. The acids thus obtained are, of course only those insoluble in water, any soluble (volatile) acids present being re- tained in the wash water. The percentage of insoluble acids is known as the ' Hehner value." For the determination of the volatile and soluble acids (" Reichert-Meissl value ") see Lew- Icowitsch. Separation of oxidised fatty Acids {Degr as- Former"). Lin- seed and fish oils yield, on oxidation, products of which the fatty acids are insoluble in petroleum ether, but soluble in alcohol. The determination of these oxidised acids in linseed varnishes and japans, affords a criterion of the amount of oxidation they have undergone, while in d^gras, in which they are formed during the chamoising process, their presence, according to Simand, is the cause of its readiness to form an emulsion with water and to a certain extent a measure of its genuineness and quality, and they are hence styled by him " degras-former " (Degrasbildner). Their importance in this respect has, however, been much over-rated, as they can be formed in any quantity by the artificial oxidation of fish-oils. They may be determined in the soap solution, after extraction of the unsaponifiables, in the following manner. After boiling off the alcohol, the soap is dissolved in hot water and the solution is transferred to a large separating funnel and decomposed with slight excess of dilute sulphuric or hydro- OILS AND FATS. 315 chloric acid, which should be roughly standardised against the quantity of potash used in saponification. After cooling, the mixture is vigorously shaken for five minutes with TOO c.c. of petroleum ether, and the funnel set aside for several hours till both the ethereal and aqueous layers have become clear. The aqueous solution is then run off, when the oxidised acids will usually adhere to the sides of the funnel, and the ethereal layer can be poured from the top of it without disturbing them. They are washed several times with petroleum ether, and dissolved in hot alcohol, which is filtered, evaporated to dryness in a porcelain basin (which should not be more than half filled, as the solution readily creeps over the sides) and finally dried for one hour at 100-105 C. and weighed. The ethereal solution, after shaking a few times with water to remove mineral acid, may be distilled off to recover the remaining fatty acids for further examination. Where only determination of oxidised fatty acids is required, as is frequently the case with degras, the process may be simplified as follows (Maschke). 5-7 grm. of the degras are weighed into a porcelain basin ; O' 5-1*0 grm. solid caustic soda, or rather more caustic potash, is added, together with 50 c.c. of good alcohol, and if necessary a few drops of water to com- plete the solution of the alkali, and the whole is slowly evaporated nearly to dryness on the water-bath with occasional stirring. The soap is then taken up with water, and decomposed with acid, and after separating with petroleum ether as above described, the alcoholic solution of oxidised acids is evaporated, and weighed in the same basin in which the saponification took place. The method is not to be recommended for the determination of small traces of oxidised acids in easily oxidisable oils, which would be liable to further oxidation in the open saponification. According to Simand, genuine degras should contain not less than 1520 per cent, of these oxidised acids calculated on the dry weight. It may be noted (cp. p. 299) that as the oxidised acids increase, the iodine value falls ; and in all probability the acetyl value, being a measure of the hydroxyacids, correspond- ingly rises. The resinous exudation on leather known as "spue- ing," " gumming," or " frying," is due to the production of these oxidised acids on the surfaces of the leather, and therefore fish oils of very high iodine value, such as menhaden oil, are specially liable to produce it, but it does not seem to be favoured by the presence of ready formed oxidised acids as in degras. 316 LEATHER INDUSTRIES LABORATORY BOOK. Most artificial degras is made with oxidised or " blown " oils, and it is probable that such oils may find other useful applications in currying. There is no known reason why a degras formed by direct oxidation should be in any way inferior to the actual product of chamoising. Determination of Resins. Rosin, or ordinary pine resin, is a usual constituent of soaps, and is frequently found in stuffing greases and degras, either as an intentional addition, or as a constituent of rosin oil. Rosin consists practically of the anhy- drides of abietic, and perhaps other acids, and hence is found after saponification among the fatty acids. It may often be recog- nised by taste and smell where the ordinary colour reactions are inapplicable. It may be detected among the fatty acids by the Liebermann-Storch reaction (in absence of liver oils, which con- tain lipochromes which give a violet coloration with sulphuric acid). Cholesterols, which give a similar colour-reaction, must first be removed with other unsaponifiable matters by shaking out the soap solution, as on p. 309. The separated fatty acids are dissolved in a small quantity of acetic anhydride. The solu- tion is then cooled, and a few drops of sulphuric acid of S.G. 1*53 (about 55 c.c. of concentrated S.G. 1-84 acid made up to IOO c.c.) is gently run in. In presence of rosin a transient reddish violet coloration is produced. In absence of cholesterols the test may also be used to detect rosin oil, and even when both have been removed by shaking out, the fatty acids may show traces of rosin. Rosin is probably best estimated by the volumetric modifica- tion of Twitchell's method,* though the results must only be taken as approximate to the quantity of crude commercial rosin used. It is carried out as follows : 2-3 grm. of the mixed acids are weighed accurately, and dissolved in a flask in 30 c.c. of absolute alcohol. The flask is cooled by immersion in cold water and dry HC1 gas is passed in till it ceases to be absorbed, which occurs generally in about three-quarters of an hour. The flask is allowed to stand for one hour and then diluted with 150 c.c. of water, and boiled till the aqueous layer becomes clear. After cooling, the contents of the flask are transferred to a separating funnel, the flask being repeatedly washed with ether which is * See Benedikt and Lewkowitsck, p. 195 : Wilson, Journ. Socr. Chem. Ind., 1891, P- 952. OILS AND FATS. 317 united with the rest. The acid layer is now run off, and the ethereal is either washed with repeated changes of water and vigorous shaking till the aqueous layer remains perfectly neutral, or decinormal alkali is cautiously added till it remains neutral to methyl orange. After running off as much of the water as possible, 50 c.c. of alcohol is added, and the solution of resin acids is titrated with standard alkali in presence of phenolphthalein as in determining the "acid value" (p. 304). Each c.c. of normal alkali is approximately equivalent to 0*346 grm. of rosin. In the determination of other resins an equivalent must be found experimentally. The process is based on the fact that fatty acids combine with alcohol in presence of dry HC1 to form neutral ethylic esters, while the rosin acids are practically un- altered. Lactones are the so-called " inner " or cyclical anhydrides of hydroxyfatty acids. When a hydroxyl is substituted in a suitable position for one of the hydrogen atoms in the hydrocarbon nucleus of a fatty acid, it becomes capable of combining with the hydroxyl of the carboxyl group of the same molecule (just as the hydroxyl of another alcohol might unite with it to form an ordinary ethereal salt or " ester "), with elimination of a molecule of water derived from the hydrogen of the carboxyl group and the alcoholic hydroxyl. Such a lactone may be likened to a snake with its tail in its mouth, and is devoid of acid properties, behaving like an ethereal salt. It is saponified by caustic alkalis in the ordinary way, the original hydroxyacid being re-formed as an alkali salt ; but as soon as it is liberated again from the base, it loses one molecule of water, and re- combines with itself to form a lactone. Fatty acids con- taining lactones have thus a '' permanent saponification value " and " permanent ether value," from which the amount of lactone may be calculated. In some cases, however, this re- formation is only partial, and by repeated saponifications the amount of lactone is often reduced or converted into permanent free acid. It is not every hydroxyfatty acid which is capable of forming a lactone. The combination takes place most readily when the hydroxyl occupies the 1896, p. 90. 334 LEATHER INDUSTRIES LABORATORY BOOK. to red and brown. Old oils only give the red coloration (which is also given by palm oil), and not the violet. These colour reac- tions are due to the presence of lipochromes and cholesterol. Similar results are obtained by using chloroform as a solvent as in the cholesterol reaction described on p. 313. Rosin and rosin oil also give red coloration with sulphuric acid. . . . These reactions might be valuable for the detection of liver oils in other oils, but are useless where liver oil is known to be the principal consti- tuent. Mineral and rosin oils render oils containing them quite unsuitable for chamoising, but are not injurious to tanned leather, and in many cases principally objectionable on account of their lesser pecuniary value. Although other marine oils than cod are little used in England for leather manufacture except as adulterants of cod oil, they are considerably employed on the Continent, and apparently with advantage, and these pages would not be complete without a short notice of them, and a few words must also be said on the hydrocarbons which are used as adulterants, and occasionally as intentional additions to stuffing greases or for oiling off levants. Other Fish-Liver Oils. It has been mentioned that in " coast- cod " the oil extracted from the livers of other fish is generally included. Of these fish the principal are the ling (Molva vul- garis}y haddock {Merluccius aeglefinus), coal-fish * (Gadus mer- langus\ and hake (Merluccius communis\ and not unfrequently that of the Arctic shark f (Somniosus microcephalus), and of the dog fish (ScylUum canicula and catalus\ which also belongs to the shark family. Some constants of these oils will be found on p. 349. Shark and dog-fish liver oils contain considerable quantities of unsaponifiable liquid waxes analogous to those of sperm and dolphin oils. This lowers the saponification value, that of shark oil being about 1 60, and its iodine value is also low, being about 115. The unsaponifiable alcohols of the waxes are not easily shaken out with petroleum spirit, but by shaking with ordinary ether, about 10 per cent, may be removed from shark oil. Exact constants are not available for dog-fish oil, but it is probably closely similar to that of the Arctic shark. Boegh and Thorsen J give a method of recognising shark-liver oil through * Sometimes called " Black Jack " ; German and Norse, Sai or Sei j Fr. Merlan. t Formerly Scymus niictocephalus, or borealis or Lcemargus borealis, the " bask- ing " shark, Norse Baakjarring ; Danish, Bavkal\ Ger. EisTial. \ Collegium, 1904, p. 73. OILS AND FATS. 335 the greater insolubility of its soap in hot water, which is no doubt due to the presence of the liquid waxes alluded to. Their process consists in saponifying 10 grams of the oil in a flask with 50 c.c. of alcohol and IO c.c. of a solution of pure sodium hydrate, containing by titration approximately 0*36 grm. of sodium hydrate per cubic centimeter. The operation is carried out on the water-bath, if necessary with a reflux condenser, and the solution is finally evaporated to dry ness. 50 c.c. of boiling distilled water is then added, while the flask is heated on the water-bath, and frequently shaken, and successive additions of 5 c.c. of distilled water are made until the soap forms a perfectly clear liquid, or no further solution takes place. The total quantity of water used is noted, and if it exceeds 70 c.c., it may be assumed that shark oil is present. Shark oil alone will not give a clear solu- tion, and a mixture of equal parts of this and cod or coal-fish liver oils requires over 200 c.c. of boiling water to do so. It is probable that similar results would be obtained with mixtures of other oils containing liquid waxes, but most of these are too expensive to be used as adulterants of ordinary fish oils. Shark liver oil is said to be particularly prone to cause the resinous exudation on leather known as "spueing," and from its large percentage of unsaponifiables is quite unsuitable for chamoising. The other liver oils named are very similar to cod in their general characteristics. Skate liver oil and haddock oil seem to be remarkable for high iodine values, a sample of the former repeatedly examined in the author's laboratory having an iodine value of 187. It is not known whether these oils are in anyway less suitable than that of cod for purposes of leather dressing, but some differences in their rate of oxidation were observed by Messrs. Procter and Holmes in their experiments on " oil oxida- tion."* Fish Oils. Under this head may be classed all those oils which are extracted, not from the liver, but from the body or intestines of several fish. Lewkowitsch points out that those fish (e.g. cod) which have very oily livers, rarely have much oil in other parts of the body, while in many others, the largest yield is from the body and intestines. The principal fish oils are menhaden, porgie, or " Straits " oil from Alosa menhaden, a fish somewhat larger than a herring, which is caught in large * Journ. Soc. Chem. Ind. 1905, p. 1287. 336 LEATHER INDUSTRIES LABORATORY BOOK. quantities off the New England coast ; sardine oil, which is obtained in the preparation of sardines for food ; Japanese sardine oil or Japanese fish oil, which is extracted from the sardine, either by boiling or by allowing it to rot in heaps, and subsequently pressing the mass ; herring oil, of which consider- able quantities are occasionally made in England from damaged herrings, and which is generally mixed with other fish oil from the refuse of various fish, and not unfrequently added to " coast cod/' Most of these oils are very similar in their chemical characteristics to genuine cod oil, but from their mixed character and uncertain origin no very definite constants can be given. The specific temperature reaction of menhaden oil is very high, and is given by Thomson and Ballantyne as 306. Its drying power is so considerable that it is sometimes used as a paint oil for outside work, and it is probable that leather curried with it would be particularly liable to " spue." It is frequently used for adulterating Newfoundland cod. It is quite suitable for oiling sole leather in warm weather, but gives trouble in cold from its high percentage of solid acids, which might no doubt be removed by filtration at a lower temperature. The saponification values of these fish oils are generally higher than that of unoxidised cod oil, averaging about 190. Sardine oils may often be detected by their taste and smell of sardines, but appear to lose it on oxida- tion by exposure to the air. Whale Oil. This is extracted from the blubber of various species of whale of the genus Balcena and Balcenoptera ; and varies considerably in quality, not only according to the species but the method of extraction, Nos. o and I, steam-boiled at low temperatures, being pale in colour, with little fishy odour, and almost neutral, while 3 and 4 extracted under pressure and often after some putrefaction are darker and of more fishy smell, and with high acid values (25-100). The "foots" contain much palmitin, and the oil contains olein and semi-drying oils like fish oils, but its chemical constitution has been little investigated, and probably varies considerably with the species from which it is obtained. It sometimes contains traces of spermaceti, but the total unsaponifiables do not exceed 2 or 3 per cent. Its saponi- fication value is 188 or a little over. Its iodine value and specific temperature reaction seem to vary considerably. It is much used on the Continent for chamoising, and consequently is one OILS AND FATS. 337 of the constituents of degras. It is probably perfectly suitable for use in currying. Seal Oil. Extracted from the blubber of various species of seal (Phoca vitulina, A. grcenlandica, P. lagura, P. Caspica}. It is frequently mixed with fish oils, as in the Swedish " Dreikronen- thran " (Three Crown Oil). It considerably resembles both whale and fish oils, and cannot be detected in mixture with these by any of the ordinary analytical reactions. Its saponification value seems generally to be about 192, though it sometimes falls as low as 185 in reputed pure samples. . . . Its iodine value is high, generally ranging from 125 to 160, and its specific tem- perature reaction 212-229, according to Thomson and Ballantyne. As the specific temperature reaction of cod oil is considerably higher, this may afford a means of inferring the presence of seal oil in mixtures. Seal oil itself is rarely adulterated, except with rosin or mineral oils which may be detected as unsaponifiables ; of which the pure oil usually contains only about J per cent. It is used largely on the Continent for chamoising and currying, and seems quite suitable for these purposes, especially in mixture with the more oxidisable oils such as cod and menhaden, of which it moderates the oxidation. Dtgras and Sod-Oil* Genuine degras is the residual oil wrung or pressed out of chamois-leather in the French method of manufacture, in which the skins are fulled with whale, seal, fish, or cod oil ; or a mixture of these, and oxidised by exposure to the air in warm rooms. A portion of the oxidation products are fixed in the leather, and together with the acrylaldehyde evolved produce an oil tannage, while the unoxidised products, and a considerable proportion of oxidised acids referred to on p. 314 are pressed out after dipping the skins in hot water. In some cases skins are worked over and over again merely for the production of degras, which may therefore be regarded as essentially an oxidation product of marine oils. In fact, a good deal of artificial degras is now manufactured by " blowing " or directly oxidising the oil under suitable conditions ; and appar- ently is equally suitable for currying. In the genuine manufac- * Interesting work on degras has been published, among others, by Eitner, Der Gerber 1890, 85 et seq. ; 1893, 245, 255 ; Simand, ibid., 1890, 243 et seq. ; Ruhsam, Ding. Polyt. Jour., 1893, 285, 233; Fahrion, Zeits. fur angew. Chem., 1891, 172; 1893, 140 ; Chem. Zeit., 1893, 434 and 684 ; Deutsche Gerber ZeiL, 1892, 33 et seq. Z 338 LEATHER INDUSTRIES LABORATORY BOOK. ture, the oil from the first wringing or pressing is called "premiere torse" and all the oil actually recovered from the skins is " motllon" Its value to the currier consists in its ready miscibility and emulsification with water, of which genuine degras invariably contains from 10 to 20 or even 30 per cent. This leads to its very complete distribution in a finely divided condition on the leather fibre, which gives softness without a greasy feel. The oxidised fatty acids (" degras-former ") are also said to be im- portant as producing a sort of retanning effect. For methods of determination, see p. 314. Sod-oil formerly differed from degras, because, in the earlier English methods of chamois manufacture, the skins are allowed to heat in pile, and a more intense oxidation is thus obtained, with so much thickening of the oil that it can no longer be removed by pressing, but must be washed out with an alkaline lye, from which it is separated by addition of sulphuric acid. It is thus liable to contain, beside much water, considerable traces of acid, or of alkaline soap, or even of both in imperfect mixture. It is fre- quently purified and deprived of water by heating in a steam pan, when the water is either evaporated or settles to the bottom with many of the other impurities. Such an oil is useful as a consti- tuent of currying greases, but it is no substitute for the genuine degras for the special purposes for which the latter is employed. Now, however, many English chamoisers treat their goods very similarly to the French, and remove a large part of the oil by pressing only, and there seems no reason why such oils should be inferior to the best French d6gras. As much of their value depends on their emulsion with water, it is a mistake to " evaporate " such sod-oils. The pure molllon is never sold unmixed, but always receives additions of untreated oils, and usually of tallow, which must hardly be regarded as adulterants. More questionable but fre- quent admixtures are wool-fat or recovered grease, and mineral oils. From what has been said, it will be seen that the analysis of so complicated a mixture as a commercial de"gras presents no common difficulty, and that a complete determination of its various constituents is generally impossible. Estimation of water is always important. *This may be done as in the case of egg yolk, p. 351, by drying mixed with sand OILS AND FATS. 339 in the air oven at about 110 C, but in presence of unsaturated fatty acids considerable oxidation may occur, and it is probably as accurate and much quicker, to heat 2-3 grm. in an uncovered and somewhat inclined platinum crucible over a small flame, till a slight crackling, accompanied by tiny puffs of smoke, indicates that the whole of the water is expelled. According to Fahrion,* the difference between the two methods does not on the average exceed o 2 per cent. If the crucible be rapidly cooled on an iron block, the entire operation need not take more than 15 minutes. In cases where much solid matter is present or from any cause such as the presence of soaps, the degras does not readily liquefy, it is preferable to add it to a known weight of oil or fat from which the moisture has been already expelled. In place of heating till smoke is produced, the melted fat may be stirred with a small (high temperature) thermometer weighed in with it, retaining the temperature within say 150 C. A useful comparative idea of the water present in degras and other fat emulsions may be obtained by shaking 50 c.c. of the material in a 100 c.c. cylinder with 25 or 50 c.c. of ordinary petroleum lamp oil, and allowing the mixture to stand in a warm place until it has separated into two layers ; deducting the volume of petroleum oil added ; the upper layer represents the the oil soluble in petroleum, while the lower consists of water, lime- and alkali-soaps, and oxidised fatty acids insoluble in petroleum. Several methods have been proposed for the determination of the specific gravity of degras, of which the best seems to be that of Galowski, f who employs an Erlenmeyer flask of which the neck is ground to fit a flat glass plate, though with care an ordinary conical flask may be used, filled accurately level with the top. The flask is weighed empty (= T), and filled with water (= W + T), carefully dried and again filled two-thirds full with the well mixed degras, and again weighed (= T + D). It is now kept for an hour or two at a temperature of 6o-8o C. ; water is added to about three-quarters of the total volume of the flask, and heated to gentle boiling till all air is expelled, which is often present in considerable quantity in the form of small bubbles. The flask is then cooled, filled with water to the * Zeits. f. angew. Chem., 1891, p. 172. + Gerber Zeitung, 1902, No. 53 ; 1903, No. I ; Collegium, 1903, p. 5. Z 2 340 LEATHER INDUSTRIES LABORATORY BOOK. brim, and weighed (= W l + T -j- D). The calculation, precisely similar to that on p. 7, is as follows : sp.gr. = :o the central \^/ being that taking the quinonoid >COONA / form. While the above formulae develop the relations of the phthaleins to triphenylmethane, their relation to the pyronine group \/ \^\^- is equally obvious. Phenolphthalein easily I parts with a molecule of water from the two hydroxyls, and 000 then becomes c o. Phenolphthalein is formed by the con- densation of ordinary phenol with phthalic anhydride c 6 H 4 / \o. If other phenols are substituted, other phthaleins are obtained. Resorcinol, for instance, yields fluorescein, which is the parent substance of most of the eosins, which are its iodine and bromine derivatives. The rhodamines are formed by the substitution of dialkyl (dimethyl, diethyl, diphenyl) amidophenols for the simple phenol employed in phenolphthalein. Another group, the acridine dyestuffs, including phosphine, benzoflavine and acridine orange are somewhat similarly related, N /\/i\/\ but contain an N in the position I i , 1 J . The safranines (in- I eluding mauve, the earliest of the aniline colours) are basic dyestuffs, in which three benzene rings are linked together by a N I * Trans. Chem. Soc., 1904, p. 398. 382 LEATHER INDUSTRIES LABORATORY BOOK. More important are the alizarine dyestufifs, derived from CO anthraquinone, Q Y Y |, itself a derivative of anthracene. Ali- co CO OH zarine, ( Y ] | OH , is identical with the natural colouring matter CO of madder, the cultivation of which it has entirely superseded. Alizarine (Turkey red) is a mordant dyestuff, giving deep red with alumina, and violets and darker tints with other mordants ; and its colours are remarkable for their fastness to light and washing. Many other anthraquinone derivatives, so-called " aliza- rine reds," "blues," "violets," and "oranges," possess similar characteristics of fastness and polygenetic dyeing. The above brief summary must suffice. Some important groups and many important individuals have been unnoticed, but even a slight sketch of the whole subject would exceed the limits which can be allotted to it in this volume. Indigo, the most important of the natural dyestufifs, is an aniline derivative, and is now produced artificially on an extensive scale. The yellow natural dyestufifs associated with tannins have been briefly noticed on p. 141. Haematein, the colouring matter of logwood, is of a somewhat analogous character to these. No simple and satisfactory tests are known by which the different groups described can be distinguished from each other. Most colouring matters can be decolorised by reduction with nascent hydrogen produced in acid or ammoniacal solutions by the addition of zinc dust, but the cause of decolorisation varies in different cases, the dyestuff being completely broken up in azo, nitro, and nitroso colours, while in many other cases, so-called leuco-derivatives are formed which either reoxidise with more or less rapidity on exposure to air, with return of colour, as in azine, and acridine colours, or only do so by treatment with an oxidising agent such as chromic acid, as in the triphenylmethane group. The return of colour, however, varies much in ease and rapidity, and in some dyes new colours are formed on reoxidisation. Green * says : " In the application of the zinc-dust reaction I have found that whilst the derivatives of triphenylmethane after re- duction are reoxidised by air only with great difficulty, the azines, * Journ. Soc. Dyers, etc., iii. p. 66. DYES AND DYESTUFFS. 383 oxazines, thiazines, and acridine colours all reappear with great rapidity on exposure of the reduced solutions to the atmosphere. " On the other hand, the triphenylmethane colours are dis- tinguished from those which do not yield leuco compounds, but are completely broken up, by the colour of the reduced solution being restored by chromic acid. The quinoline and primuline colours are distinguished from other dyestuffs by the great diffi- culty with which they are reduced, primuline for instance remaining unaltered after long boiling with zinc-dust and ammonia. " The reduction with zinc-dust requires to be performed with some care. It is usually best to add a little zinc-dust to the hot solution of the colouring matter contained in a test tube, to agitate, and then add dilute hydrochloric acid drop by drop until decolor- ised. An excess of acid must be carefully avoided. In other cases, more especially when the colour-acid is very insoluble, the reduction must be performed with zinc-dust and ammonia ; but it is usually safest to repeat the tests with both methods of reduction. " Whilst it is important that the colouring matter should be completely reduced, care should also be taken that the reduction is not carried too far, for some colouring matters (e.g. the indu- lines) which usually form oxidisable leuco compounds are there- by completely altered. When reduced, the solution is poured off from the greater part of the zinc-dust on to a piece of white filter-paper, and exposed to the air ; filtration is unnecessary, as a little zinc upon the paper has no injurious effect. If the colour does not return in a minute or two, the paper is touched with a glass rod dipped in a I per cent, solution of chromic acid (with addition of sulphuric acid for acid colours). Warming gently for a few seconds over a flame facilitates the reactions, and should also be employed when the reduction has been performed with zinc-dust and ammonia, in order to remove the excess of the latter before adding chromic acid, which would otherwise be neutralised. " In the case of acid colouring matters, the paper, after spot- ting with chromic acid, should be held over a bottle of strong ammonia, for some acid colours (e.g. the cosines) do not show their true colour whilst acid. On reduction of those cosines which contain iodine, a brown spot of the latter is obtained with 384 LEATHER INDUSTRIES LABORATORY BOOK. chromic acid, but this disappears on exposure to ammonia vapour, and the original colour of the eosine (or a rather yellower shade) returns. As already pointed out by Weingartner, great care must be taken that one is not led astray by secondary colours formed by oxidation of the diamines and amidophenols obtained by reduction of azo-colours." A large number of dyes sold under special names, or supplied to produce particular tints, are merely mixtures of two or more well-known colours. Browns, for instance, frequently consist of Bismarck or some other common brown shaded by the addition of small quantities of benzaldehyde green or methyl violet. Such mixtures are often useful, though they not unfrequently dye un- evenly from the different affinity of their constituent colours for the leather, and skilful dyers will usually prefer to mix for themselves. They are not easy to separate chemically, but may frequently be detected by sprinkling or blowing the powder very sparingly on a piece of wet filter paper, when the separate grains will give specks of different colour. In place of wet paper, the surface of water in a beaker may be sprinkled, and the threads of coloured solution watched from below. In some cases, where the mixed colours are too similar in appearance to be detected in this way, they are revealed by sprinkling in a shallow basin on concentrated sulphuric acid, in which many dyes dissolve with very different colours to those of their solutions in water (Spiller, Allen). Another method, which is applicable to solutions of mixed dyes, is to place a drop on a sheet of filter paper, when, in case of mixtures, concentric circles of different colour will frequently be formed, because of the different diffusibility of the colours. Solid dyes are generally best dissolved in a small quantity of alcohol, and the solution diluted with its own volume of water. Goppelsroeder* suspends strips of Swedish filter paper with their lower ends dipping in small beakers containing the dye- liquors, and thus allows a more prolonged and complete diffusion. It must be remembered that in many cases the dyes, as manu- factured, are not pure chemical products, but contain small pro- portions of other colouring matters produced at the same time by secondary reactions. Thus magenta often contains traces of phosphines. * Journ. Soc. Dyers, etc., iv. p. 5. DYES AND DYESTUFFS. 385 Mixtures may often be detected by dye-tests, if the affinity of their constituents for the leather or textile fabric is at all unequal, since the first portions dyed in a bath will not only be darker in colour, but of different tint to the later ones, because those colours which have least affinity are left in larger propor- tion in the bath. The use of the spectroscope will often detect differences in colouring matters of which the tint to the eye is identical. The dye is dissolved in water or alcohol, and diluted in a test-tube or small rectangular bottle till the absorption-bands appear to the greatest advantage when the light passing through the solu- tion from the sky or a lamp is examined. A small pocket spectroscope is well adapted for the purpose, but it should be provided with a prism to permit the direct comparison of two solutions. The advantage of rectangular bottles over test-tubes is not only that the light passes better through the flat sides, but that two different thicknesses of solution are available by turning the bottle, and the appearance of the spectrum is often greatly influenced by the thickness or concentration of the solu- tion. A very elaborate series of tables for the identification of artificial colouring matters by their absorption spectra has been published by Formanek,* but its use requires the employment of a spectroscope fitted with apparatus for measuring the position of the bands. The analysis of curriers' inks and other blackings for leather is a problem which not unfrequently occurs to the leather trades chemist. The colouring matters of these inks are usually log- wood or tannins (frequently with fustic or quercitron to destroy the bluish shade of the blacks) in combination with iron, and often with small quantities of - chrome and copper salts, the chrome being added in the form of bichromate to oxidise the logwood and iron. Aniline colours may also be present. The separation of these different dyestuffs may sometimes be effected by the capillary method with strips of filter paper which has been described above, but is very difficult by any strictly chemical process. Logwood usually gives a red solution or residual colour when discharged by hydrochloric acid, while pure tannin blacks become yellow or colourless. By treatment with * Spektralanalytscher Nachweis kiinstlicher organischer Farhstofie. T. Formanek, Berlin, 1900. 2 C 386 LEATHER INDUSTRIES LABORATORY BOOK. caustic soda, a portion of the logwood colouring matter may be brought into solution in a colourless form, which turns violet on acidification. Gallic acid and the colouring matter of log- wood can be separated in some cases by shaking the ink out with ether after acidification. If the ether be shaken with water and a little ammonia, logwood gives a pinkish violet coloration. Some aniline colours can be shaken out with amyl alcohol. As a rule, the most useful information is obtained by exami- nation of the ash obtained on incineration. Chromium is not likely to be present without logwood. Copper and iron may be used in either tannin or logwood inks. Sulphuric acid should be determined by igniting the dried residue of the ink with addition of a little sodium carbonate, and subsequent solution of the residue in hydrochloric acid and precipitation with barium, chloride. Its proportion to the bases present should be calcu- lated, as organic salts of iron are less likely to injure the leather than ferrous or ferric sulphates. Thickening matters likely to be present are sugar, gums and dextrine, and occasionally fatty substances. American blacks often contain tragacanth, which may be recoghised by its large round starch cells, which only colour slowly with iodine unless the solution is very strong,, probably on account of a protective covering of mucilage. Gums and dextrine may be precipitated by addition of considerable quantities of strong alcohol, while the sugars will remain in solution, and may be recognised by their reaction with Fehling's solution (p. 267). Another class of blackings are those used on boots, in which the black consists mainly of carbon, produced by the carbonisa- tion of some form of] sugar with sulphuric acid. Bone black is frequently added to the mixture to neutralise the excess of acid, and hence calcium sulphate, phosphate, and carbonate may be found on analysis. Sugar is usually present, with some quantity of oil or fat. Pastes for coloured boots contain waxes and oils, with soaps or turpentine and aniline browns or yellows. " Quick- blacks," for shoe manufacturers are alkaline solutions of shellac and other resins or waxes with logwood or aniline blacks. The blacks used on waxed leathers consist of lampblack with oil or soap ; while in the sizes, glue, tallow, 'flour, gum-traga- canth and soaps are frequent constituents. Carbon blacks are DYES AND DYESTUFFS. 387 perfectly insoluble in acids or alkalis and are only destroyed by heat and powerful oxidisers. They may, therefore, always be removed by filtration after bringing the other constituents into solution by digestion with acids, alkalis and hydrolysing agents. The insoluble matter may then be collected on a tared filter and weighed, and the insoluble inorganic constituents estimated by ignition. The estimation of fats and fatty acids is frequently difficult from the presence of mucilages which cause emulsifica- tion. These may generally be hydrolysed by prolonged digestion with weak acids, or sometimes by short treatment with strong sulphuric acid, and the fats may then be removed by shaking out with petroleum ether or some other suitable solvent. It must not be forgotten that unsaturated fats are acted upon by strong sulphuric acid (p. 299). Unsaponified fats or fatty acids may often be recognised as highly refracting globules under the microscope. No general methods for the analysis of such mix- tures can be given, and their investigation will usually test the ingenuity and resources of the chemist, while experience of the substances likely to be met with will be found to play an impor- tant part. Much additional information on the analysis and detection of colouring matters will be found in Allen's ' Commercial Organic Analysis/ already quoted. In the case of mordant colours, useful hints may often be obtained by the examination of the ash for metallic bases (pp. 363-366). Where the object of the chemist is simply to determine the dyeing value of some colour, the nature of which is known, in some cases recourse to chemical analysis may be had, though actual dye-trials will usually prove most satisfactory. Colours are often " diluted " with dextrine, sugar, salt and other indifferent matters, or salt may occur as an impurity of manufacture, being frequently used to precipitate dyes from their solutions. When colours are really identical, their relative strength may be determined by the colorimeter, or more simply by matching a dilute solution of the dye to be tested by adding a solution of a standard dye of known purity from a burette to a similar volume of water in a Nessler tube on a white surface. When the colours are equal, each tube contains the same weight of actual dye, from which the relative strengths can be calculated. If, however, the dyes are not really identical in character, such 2 c 2 388 LEATHER INDUSTRIES LABORATORY BOOK. tests are often deceptive. The tint of a dye may be matched and registered by the tintometer. (See p. 239.) The most convenient material for dye-tests on leather is generally sumach-tanned skiver, though any other material may be used which it is intended to dye. The method of testing adopted should as far as possible be arranged on the same lines as the method actually to be adopted in dyeing. Those prac- tically in use are " staining," in which the colour is applied to the surface with a sponge or brush ; dyeing in trays, in which the goods are moved by hand ; and dyeing in paddles or drums. Staining is not merely a question of painting with a coloured liquid, but it is necessary that the dye should have an affinity for the fibre. As the goods are not washed after staining, the use of sulphuric acid in any form in the staining liquor is quite inadmissible, and it is usually preferable to employ basic dyes, which not only have a great affinity for the leather, and great covering power, but can be applied without acid. If " acid " dyes are used, it is generally necessary to employ some acid to combine with the base and set free the colour-acid, and for this purpose formic acid is most suitable. Solutions of aniline colours of about 1-2 grm. per liter according to the strength and solu- bility of the dye are usually employed, and two or more coatings are given. The leather should be damped before the application of the dye, and the coating should not be repeated until the leather is dry on the surface. Samples are conveniently stained on a sheet of glass. Alumed kid and chrome leather are often dyed on the table by a process very similar to staining, but mordant-colours such as the dye-woods are very frequently used. The table is generally of lead, and slightly higher in the middle, so that the surplus dye runs off to a channel round the edges. The skins are placed on the table quite wet, and spread with a sleeker, and the necessary mordants and dyes are applied with brushes, washing the skin with water between the applications to remove loose dye which might be precipitated. Dyeing in the tray is done in two different methods. The usual old- fashioned English method employs a tray usually sufficient to hold from 2-3 dozen skins, and in order to economise dye and avoid dyeing the flesh side of the skins, they, are frequently " paired " flesh to flesh, or " pleated " by doubling down the back, and pressing the surfaces together on the table with a sleeker. DYES AND DYESTUFFS. 389 By careful handling, these surfaces may be maintained in con- tact throughout the dyeing process. Comparatively weak dye- baths are used, and acid colours are generally preferred, though basic may also be employed ; and the skins are turned in the bath, often for three-quarters of an hour or more, till the bath is as completely exhausted as possible, the colour being usually added in two or three portions so as to begin in a weak bath. In the usual Continental method of dyeing, two small trays only are used, each of them capable of holding a single skin or pair of skins which are paired or pleated. The bath usually holds about 6 liters of water for small skins, and basic dyes are preferred from their rapidity of action. The dye is usually kept in a con- centrated solution and diluted with hot water to the required volume, a temperature of 4O-45 C. being suitable. The skins are dyed singly or in pairs, and each is entered in a nearly exhausted bath, and finished in a fresh concentrated one in which it is turned till dyed to shade ; the dyeing usually occupy- ing only a few minutes. An undyed skin is entered in the second bath, through which two have already been passed, and after two or three minutes' turning goes into that from which the previous skin has just been dyed out. The exhausted bath is now emptied, and a fresh one prepared in which the skin is finished, the second bath now becoming the weakest, in which a fresh skin or pair of skins is entered. Thus, though only two trays are used, each skin passes through three successive baths, and the dye liquor is fairly exhausted. In paddle and drum dyeing the proceeding is much the same as that in the old English method of tray dyeing, but in the paddle a larger volume of liquor has to be used, and in both cases the flesh as well as the grain side of the skin is dyed, thus involving a larger expenditure of colour, though effecting a considerable saving in labour. Paddles are usually sufficiently large to dye several dozen skins, and the colour is added in successive portions, the exhausted dye-bath being usually run away. For experimental dyeing, basins of about 500 c.c. capacity, broad and shallow beakers, or porcelain or enamelled photo- graphic trays are conveniently employed, and must be maintained at a temperature of 4O-45 C. by placing them in a water-bath. Paddle and drum experiments are conveniently made in the churn described for colour tests on p. 240, a larger volume of 390 LEATHER INDUSTRIES LABORATORY BOOK. liquor being employed when it is desired to imitate the action of the paddle, and the jars being if necessary wrapped in flannel or felt to retain the heat. With acid dyes, if sulphuric acid is used, a quantity equal to half the weight of the dye is invariably sufficient. In dyeing in the basins or trays, goods must be con- tinuously turned over, either with the fingers or with a glass rod, or after the manner of the Japanese student, with two glass rods between the fingers like chopsticks. When the dyeing is completed the pieces are taken out, rinsed with water, and squeezed and smoothed out with a brass or glass sleeker on a glass plate, and are nailed or pinned out on a board to dry. By comparing a set of samples dyed in this way it is easy to judge of the relative strength and tint of the colours, and a further idea may sometimes be obtained by dyeing a number of pieces successively, allowing longer time as the bath gets weaker. If the colour is a mixture of two or more dyes, of which the attrac- tion for the leather is not exactly equal, the tint of the later pieces will vary from that of the earlier ones, although the depth of colour may be equal. The colour of dyed leather is made darker and fuller by finishing ; and the effect of this may to some extent be judged by burnishing lines on the very slightly darr^ leather with a smooth piece of wood, or the handle of a knife or toothbrush, or a sample may be glazed and finished if the appliances are at hand. The dry samples are invariably paler and duller in tone than when wet, but this is considerably recovered in the finishing process. If the dye has been allowed to precipitate in the bath, portions of it will adhere to the surface as ; powder which will rub off, and if its attraction has been too great and the dyeing too rapid, it will trequently occur that when sleeked out the leather " gapes," as the dyer calls it, showing undyed portions in the less exposed minute creases of the grain. Colours which are not quite exact to shade are usually corrected by the finisher, who mixes dyes with his seasonings. In calculating from dye-tests, I grm. per sq. meter is equiva- lent to 30*5 grm., or slightly over I oz. per 100 sq. feet, and I grm. per liter to I oz. per cubic foot of dye-liquor, or I Ib. per 100 gallons. SECTION XXIV. THE USE OF THE MICROSCOPE.. AMONG the means of investigation at the disposal of the leather chemist, the microscope takes a most important place. It is only by its use that the minute structure of the skin can be studied, and it is constantly requisite for the examination of surface defects, such as " weak " grain and other forms of bacterial damage. It has also become very important as a means of recognising adulteration in tanning materials, and especially in sumachs. It is essential in the study of bacteriology and fer- mentations, which play so important a part in leather manufacture. While for many of the purposes named comparatively low powers are sufficient, for bacteriology the highest powers and the greatest accuracy of adjustment both of focus and lighting are essential. For preliminary examination of leather and many other things, a good achromatic pocket-lens of the Coddington type, and of i-in. to J-in. focus is almost essential, and often forms an effective substitute for the lower powers of the microscope. While it is possible to do very useful work with a some- what inferior instrument, it is advisable in purchasing a stand to secure that in really essential points it is of the best, although sometimes refinements such as the mechanical stage may be dispensed with. Good instruments of comparatively moderate price can now be obtained both of English and Continental opticians, but I5-2O/. may be put down as about the cost of a really efficient microscope to cover all the uses mentioned. Out of this, however, 5/. at least will be required for an oil immersion objective for higher bacteriological work, which can generally be dispensed with at first unless bacteriology is made a. speciality. If more cannot be afforded, it is better at least to obtain a solid and simple stand with a suitable sub-stage fitting for an Abbe condenser, a substantial stage, and say i-inch and ^-inch objectives. As high powers are often required in the examination both of hide sections and of ferments, which are the principal objects of investigation in a tannery, it is of the first importance that 392 LEATHER INDUSTRIES LABORATORY BOOK. the fine adjustment should be perfectly steady, without vibration or backlash. This is seldom or never the case with cheap microscopes, in which the fine adjustment is made by a screw at the side of the tube moving the nose by means of a lever. A much more satisfactory arrangement is that in which the whole body of the microscope is raised or lowered by a screw in a pillar at the back of the stand on which it slides. A rack for coarse adjustment is very useful but not absolutely essential. If a sliding tube only is provided, it must be tight enough not to slip, but must move easily up and .down with a sort of screw- ing movement. A mechanical stage is not necessary, though often convenient, but for many purposes one of black glass or ebonite will fill all requirements. If a simple rotating disc with holes of various sizes is used for regulating the light, it should be as nearly level with the surface of the stage as possible, and when examined with a low power should appear in the centre of the field ; but an Abbe illuminator or some similar condenser with iris diaphragm is much more convenient, and, for work on the minuter ferments, is essential. This can be fitted into the tube below the stage of the cheaper microscopes, but is much better on a separate substage with focusing and centring adjustments. A frequent defect in cheap microscopes is that the mirror for substage illumination does not bring the rays of the lamp to a focus exactly on the object, but frequently some inches above it. This may be to a great extent overcome by the use of a bull's- eye condenser between the lamp and the microscope. Another defect is that sometimes the centre of the mirror is not in a line with that of the microscope body. It is wise, in purchasing a microscope, to select first a good and substantial stand, to which additional fittings and objectives can be added as required. Most good stands are now fitted with the standard screw of the Royal Microscopical Society.* The optical part of the compound microscope consists of the " objective," which is a compound lens screwing into the lower end of the microscope tube, within which it forms an enlarged * Among English opticians, Swift and Watson may be remarked as producing very suitable microscopes of moderate price, while on the Continent, Zeiss, Reichert, and Leitz have a well deserved reputation. Figs. 23 and 24 represent good English and Continental types, which now differ but little. THE USE OF THE MICROSCOPE. 393 image of the object, which is still further magnified by the " eye- piece," a second combination of two lenses which slides into the top of the tube. Of these, the objective is by far the more expensive and important, for no amount of magnification by the eyepiece will make anything visible which is not clearly defined in the image which the objective projects. It is also found that the " resolving power," or capability of rendering minute details, depends not merely on the magnifying power of the microscope, but on the "angular aperture " of the objective, or width of the cone of rays which it collects. The magnifying power of the objective depends simply on its focal length, with which it varies inversely. Thus objectives are known by their focal length : a "i -inch" with "A" eyepiece magnifying about 50 times, a J-inch 300 times, and a ^ about 600 times. With the " C " eyepiece the magnifications are about double those with the " A." The Continental opticians number instead of lettering their eyepieces, and their magnifying powers neither agree with the English nor with each other. It would be much better if the eyepieces as well as the objectives were described by their "equivalent focal length," that is, the focal length of a single lens of the same magnifying power. Described in this way an " A " eyepiece should be of 2 inches or 50 mm., and a " C " of i inch or 25 mm. equivalent focus. The magnifications given are those with a ID-inch tube. Continental, and many English microscopes are now made with a 6-inch tube, and therefore magnify less. They are mostly fitted with a "drawtube," by pulling out which the tube can be lengthened and the magnifica- tion increased. This is often very useful, but it must be remem- bered that an objective performs best with the length of tube for which it is adjusted; and that a lower power, if of larger aperture or greater optical perfection, will often show more than a much higher power if of inferior quality. It must not be supposed that an objective focuses at a distance from the object equal to its nominal focal length. This is measured from the " optical centre " of the combination, and in all high-power lenses the front lens is brought as close to the object as is practicable, in order to collect the largest possible cone of rays, and thus increase both the light and the resolving power. Frequently the " working distance " is so short as to necessitate the use of the thinnest possible cover-glasses, and 394 LEATHER INDUSTRIES LABORATORY BOOK. the most careful focusing to avoid damage either to the object or the objective. All objects which are examined under high powers must be covered with glass, and the objectives are con- FIG. 23. THE USE OF THE MICROSCOPE. 395 structed -in view of this, the presence of the glass involving a correction which varies with its thickness. Some objectives are FIG. 24. provided with a collar which varies this correction, but most are now made for use with a cover-glass of one thickness, generally 396 LEATHER INDUSTRIES LABORATORY BOOK. the thinnest to be had, and this is the most convenient arrange- ment. Where a variable correction is provided, it is used by focusing on some test-object, such as a diatom, and then alter- ing the correction collar, at the same time keeping the focus right with the fine adjustment till the best definition is obtained. This is very troublesome, and is best made once for all. Slight variations of thickness of cover-glass may be corrected by the drawtube the thicker the glass, the shorter the tube re- quired. The objectives which have so far been described are "dry," that is, a space of air exists between the cover-glass and the front lens. By a well-known law of optics, a ray cannot pass out of water or glass into air, at more than a certain angle with the surface, but is " totally reflected " back. This effect may be seen by looking upwards at the surface of water through the side of a tumbler, when it will be found impossible to see through it if the angle exceeds a certain amount. For glass this is about 41, and hence it is impossible to obtain from an object in balsam under glass, a cone of rays of which the sides are inclined to each other at an angle of more than 82, all rays beyond this being reflected downwards. If, however, a drop of water be interposed between the glass and the front surface of the objec- tive, this effect is greatly diminished, and if for water a liquid be substituted such as cedar oil, which has practically the same refractive power as glass, it disappears entirely, and all the rays can reach the objective which it is capable of collecting. Objectives constructed to be used in this way are known as " immersion objectives," and those using oil of equal refractive index to the glass are styled " oil " or " homogeneous " immersions. Either oil- or water-immersions can be made of wider angle than corresponds to an air-angle of 180, and the oil-immer- sion ^ has proved almost indispensable in bacteriological studies. The most useful objectives for the leather manufacturer are the i-inch and -inch dry, to which may be added the ^ oil- immersion for bacteriological work. The latter is only essential for actual study of bacteria, as most ordinary forms can be seen, and in many cases their species distinguished, by a good J of wide angle. Where the j^ is also provided, a of some- THE USE OF THE MICROSCOPE. 397 what smaller angle will suffice, and will be found cheaper and more convenient in use on account of its greater working distance. Illumination is one of the most important points in practical microscopy. With powers of not less than half-inch focus, objects may generally be examined by light thrown upon them from above by a bull's-eye condenser, or by good daylight. In this case they need not be transparent or specially prepared ; and the plan is often convenient for a mere surface examination. In examining bodies illuminated in this way, prominences fre- quently appear as hollows and vice versa, by a sort of optical illusion, which, once established, is very difficult to overcome. By remembering the direction of the light, and that this appears reversed in the microscope, it is easy to decide the truth. For all finer work and higher powers, and most generally with the low powers also, it is necessary to render the object transparent, and to examine it by light transmitted from the mirror below the stage. Good daylight is least trying to the eyes. Where artificial light must be used, that of a small paraffin, incandescent gas, or electric lamp is best ; and a blue chimney, or blue glass inter- posed between the stage and mirror, or lamp and microscope, spares the sight, and makes it easier to distinguish colours. The light should be sufficient, but not too dazzling. Work should never be prolonged after the least strain is felt, nor should the microscope be used for some little time after a meal. It is well to accustom oneself to keep both eyes open while observing. In observing in daylight it is usually best to use the plane side of the mirror, but with the lamp, the concave side must be employed and its distance from the lamp and below the stage so chosen that the light is brought to a focus on the object. If, as is sometimes the case, the focus of the mirror is too long to allow of this, it may be shortened by interposing the bull's-eye condenser between the lamp and the microscope, or by a con- densing lens fitted below the stage. In most cases the centre of the mirror should be in the line of the centre of the tube, but some objects, and especially diatoms, are best seen by oblique 398 LEATHER INDUSTRIES LABORATORY BOOK. light. The use of various apertures of the revolving diaphragm will assist in the making out of structure, the smaller apertures practically narrowing the angle of the lens, and increasing its penetration (the power of distinguishing objects slightly out of focus) while at the same time the resolving power is somewhat diminished. If the microscope is furnished with the Abbe illuminator with iris diaphragm, all other means of using transmitted light will be found to be more or less makeshifts, accomplishing with difficulty what it will do with ease, and with much greater per- fection ; and, especially, the power given by the iris of varying the illumination gradually from bright to dark, or wide to narrow angle is most useful, though it is only with high powers, and particularly for bacteriological work that it is indispensable. With the Abbe illuminator or any other substage condenser, the plane side of the mirror should be used. Focusing. It will be obvious from what has been said, that with the higher powers the greatest care is needed to avoid screwing the objective down on the cover, and so breaking one or both of them. One way to avoid this is to screw down as close as possible to begin with, and then focus upwards. Another plan, when the object on the slide is small, is to keep con- tinuously moving the slide gently with the fingers while looking into the tube. It is then easy to notice when the dust and small particles on the slide come into focus, and if the point should happen to be overstepped the contact will generally be felt before serious damage is done. While observing, the fingers should be kept on the focusing adjustment, which should be continuously slightly varied, so as to look into the object in its various planes of depth, and to learn their relations. In this way many points of structure become clear, which could not have been distinguished while only one plane was visible. It must always be remembered that with high-powered and wide-angled objectives only a very thin plane can be seen at once, to which the thinnest section which can be cut has a relatively great depth, so that the picture may be very different as we focus on the upper or the lower layers. A curious point which may be noticed is, that under these circumstances we can apparently see through opaque objects if they are of small size. A hair, for instance, on the upper surface may become THE USE OF THE MICROSCOPE. 399 quite invisible if we focus on something below it, of which a perfect image is formed by the outer rays of the cone of light,, which pass round its sides. By varying the focus we can distinguish between air-bubbles and oil-globules, both of which will be early encountered by the microscopist, and are so far puzzling to the beginner that, in a well-known Continental work on tanning the former are clearly described and figured as a new form of albumen ! Both form regular circular objects, which, in certain positions of the focus are surrounded with a clear dark outline. On screwing 1 the o objective downwards this broadens, and the centre brightens in the case of the air-bubble, while in the oil-globules the edge becomes lighter and the dark ring moves inwards, and becomes hazy and indistinct and the centre shaded. When the objec- tive is raised the reverse changes take place in the two objects. When both oil- and air-globules are present, it will be noted that the dark margins of the former are much narrower than those of the latter, and that oil-globules often act as small lenses, in which images of outside objects, such as the window bars may be dis- tinguished. Oil- and fat-globules may be observed in most glues and gelatines when dissolved or swollen in water, and are one of the principal causes of their opacity. Their size and number is often worth observing in fat liquors, egg substitutes, and suchlike preparations, as an index of the quantity of oil present, and the perfection of the emulsification. Preparation and Mounting of Objects. Very few objects can be examined under the microscope without some sort of prepara- tion. To study the grain of leather under low powers, a small piece may be laid on a glass slip, with its surface as level as possible, and examined under a i -inch objective by reflected light of which the direction should be varied by moving either the light or the object, so as to see it under as many aspects as possible. In the same way a thin slice showing the cut or section of the leather may be examined, when the fibre bundles and their finer and closer arrangement in the grain layer, the dark glossy ends of the cut bundles, and sometimes crystals and concretions of catechin (the " whites "), of glucose or barytes, and other weighting material may be detected. In a similar way ground tanning materials may often be examined with advantage, a small quantity being dusted on a slip, and compared with a 400 LEATHER INDUSTRIES LABORATORY BOOK. sample of known quality, when adulteration with other vegetable matters or with sand may be detected.* All objects which are to be examined with high powers require the use of transmitted light. The easiest to deal with are those which are sufficiently small or soft to form a thin layer under a cover-glass without being cut into sections. Generally such objects are examined in water or glycerine, but if naturally transparent, the dry state is sometimes advantageous, e.g. starches and crystals. Take a little arrowroot or potato starch, and dust so small a quantity on a glass slip as to leave a little whitish cloud ; place it on the stage of the microscope, which should be horizontal, and cover with a thin glass. It may now be examined by transmitted light under the \ or \ objective. The starch granules will be seen as flattened ovals. By careful focusing and lighting, the little spot known as the hilum will be seen near one end, while the surface of the granule is marked with rings, of which the hilum is the centre. Most other starches are smaller than those named, and frequently different in form. While potato starch may reach a length of 180/1, (thousandths of a millimeter), rice starch forms small polygonal grains, which do not exceed 6 ^ in length. Many starches and especially those of the bean tribe, have radiating cracks round the hilum. A good plate showing the forms of the most important starches may be found in Allen's ' Com- mercial Analysis,' vol. i. Starches are very widely distributed in the vegetable kingdom, in fruits, roots, and even the barks of trees. Those of some of the tanning materials are very charac- teristic (e.g. Angica bark, canaigre, etc.). Useful hints may often be obtained by detection and identification of starch grains, as to the farinaceous materials used in the tawing and finishing of leather. Such observations are best confirmed by direct comparison with genuine samples of the starches suspected. So far I have spoken only of the dry examination of the starch. Let it now be mixed with a little water and a cover- glass gently laid on with forceps, so as to avoid the formation of bubbles. It will probably be found more difficult to see the rounded form of the granules, and if glycerine be substituted for * For many of these purposes, the pocket-lens may be advantageously employed ; its portability and greater depth of focus doing much to compensate for its inferior magnifying power. THE USE OF THE MICROSCOPE. 401 water, this will be still more so, while the granules will appear flat and almost entirely transparent. This illustrates a general principle in microscopy, to the cause of which attention must be drawn. If a glass rod be looked at in air, its form will be clear, and any markings on its surface distinctly visible. Let it now be plunged successively in water, glycerine, clove oil or turpen- tine, and bisulphide of carbon. It will become less and less visible until, in the last, it will almost disappear. If the experi- ment be repeated with a tuft of glass wool, or with glass powder, it will be found quite white and opaque in the air, while in the carbon bisulphide it will be transparent, so that print can be read distinctly through it. A little consideration will make the cause of this clear. The glass in itself is perfectly transparent, but at its surface the light is refracted and reflected in all direc- tions, instead of passing into and through it. If it is brought into a liquid which has practically the same index of refraction as glass, it becomes for optical purposes like a solid block, through which the light passes without interference. The minute structures which are observed under the microscope may mostly be regarded as transparent, and when brought into media which, like glycerine, clove oil and balsam, are of nearly the same refractive power as themselves, they are cleared up in the same way as the glass powder. This is often extremely useful in enabling us to see into structures under the microscope, but, as we have seen, it renders their outside form less visible. To get over this difficulty we resort to processes of staining, which make the structure visible by differences of colour. Place a small drop of a solution of iodine in iodide of potassium in contact with one side of the cover-glass which is over the starch, and at the opposite side lay a little slip of blotting paper, and watch the result under the microscope. The iodine solution will be drawn under the glass, and will stain the starch granules deep indigo blue. This is a most useful method of recognising starch. Many of the coal-tar colours dissolved in water or alcohol are useful stains for vegetable matters, and bacteria, while for animal sections polychromatic colours such as carmine and logwood are more generally useful, as their dyeing is more selective, and enables different kinds of tissue to be more readily distinguished. If the microscope is provided with polarising apparatus the 2 D 402 LEATHER INDUSTRIES LABORATORY BOOK. starch granules observed with crossed Nicol prisms appear light on a dark ground, with a characteristic bright cross, which changes from light to dark on rotating the analyser. The polar- iscope is a very useful addition to the chemist's microscope, but for details books on microscopy must be consulted. Fibres, powders and pasty materials of various kinds may be examined in water, like the starch, and may be stained in Various ways if necessary. For many materials a useful and easy mode of preparation consists in teasing with a pair of needles, or even rubbing up in a mortar. Such needles are sold by opticians, or may be made by taking a common sewing needle, breaking off short with a pair of pliers, forcing the point into the end of a piece of wooden penholder to ^^ make a hole, and then pushing in the r/jL broken end with the pliers. A minute fragment of hide or leather may be teased in this way in a drop of water or gly- ?<-j. I m cerine, on a glass slip, and the fibres, if , j= . 5.5 r> desired, stained with logwood or an aniline I.HEMViioT "'E.SA-LL. FIG. 2 5 .-Muscular colour ( see P- 423), when they may be Fibre. examined under the highest powers. Com- parison of hide fibres treated with different depilatories or tanning processes is both interesting and useful. A shred of lean meat, preferably cooked, will be found an inter- esting object for comparison, the muscle-bundles being some- what similar to those of the hide ; but the finest fibres will show under the J objective the cross-striping which is charac- teristic of all voluntary muscles. In order to get a clear idea of general structure, it is in almost all cases necessary to make sections. Vegetable sections are easier to cut than animal, and a raw potato, carrot or radish is a good object to practise on. The potato will cut better if first cut into pieces of perhaps a quarter of an inch square, and hardened by soaking first in dilute and then in strong methy- lated spirit for a day or two, which withdraws the superfluous water. The strip is now taken between the fihger and thumb, as shown in Fig. 26, and with a sharp razor, flooded with dilute THE USE OF THE MICROSCOPE. 403 alcohol, and resting on the tip of the forefinger, a thin shaving is taken off the top of the strip, using either a drawing or pushing cut, as may be found most convenient to the individual worker. The section, when cut, is removed carefully by a camel's-hair brush dipped in dilute spirits, rather by floating than pushing, and is placed in a watch-glass or small porcelain basin in spirits or water. Some- times water may be .substituted for alcohol in cutting, but the latter is more generally useful. It will ' FlG 26 be found at first that few of the sections are thin enough, and it is best to cut a good number and, rejecting the thicker, select the thinnest only for use. These may now be examined in water under a cover-glass, and, if necessary, stained with iodine solution, which will colour the starch blue, as before, and the protoplasm of the cells yellow. A thin section, even if torn, is often more useful than a perfect one of greater thickness. Pieces of liver or kidney, hardened in alcohol, are also very good objects for practice, and if desired may be stained and permanently mounted like hide sections (see p. 423). Small or thin pieces are conveniently held between pieces of cork, elder pith, or even slices of carrot, or embedded in paraffin wax or a mixture of stearine and naphthaline, by filling a little square mould of paper with the melted mixture, and then, as it begins to set, but not before, holding the piece of tissue in it with a needle or forceps till it solidifies. The block may afterwards be trimmed into convenient form for cutting. It will be found impossible by freehand cutting to get large, even sections like those of the shops, which are cut with the microtome, but some part will probably be found thin enough for examination, and it is almost impossible to cut sections of skin exactly in the right direction with the microtome, though it is very useful for tanning materials, etc. Descriptions of the modes of using it will be found in books on practical microscopy, to which the reader is referred. The Cathcart microtome is a good and cheap form, which is available both for freezing and embedding. 2 D 2 404 LEATHER INDUSTRIES LABORATORY BOOK. Apart from the study of structure, the microscope will be found useful in the identification of minute quantities of substance by chemical reactions. Crystals may be recognised by their form, and, where polarising apparatus is available, by their action on polarised light, which often gives rise to gorgeous chromatic effects. Under a low power, minute portions of chemical re- agents may be added, either on an open slip, or, like the iodine FIG. 27. Section of oak twig stained with ferric salt, showing distribution of tannin (Bastin) ; c. corky layer ; t. tannin cells ; St. stone-cells ; Ca. cambium ; Mr. medullary ray ; p. pith. solution with starch, under a cover-glass. Iron may, for instance be recognised by its action with tannin or ferrocyanides, car- bonates by the production of gas-bubbles in presence of acids (best seen under a cover-glass), tannin in plant cells, etc., by its reaction with iron, fat by blackening with osmic acid, and so on. Further details of microscopic work will be given in connec- tion with the special cases in which it is applied, but much THE USE OF THE MICROSCOPE. 405 additional information on staining and preparation will be found in Lee's * ' Microtomist's Vade-mecum,' Zimmermann'st ' Micro- technique,' and works on special branches such as botany, bacteriology, and histology. It is important that the student should be provided with means of measuring and of drawing to scale the objects which he observes with the microscope. One of the simplest appliances for the purpose is a small disc of glass, ruled in squares, which is slipped into the eyepiece of the microscope. By using the ordinary square-ruled paper employed by engineers for sectional drawings, any one with a little experience in freehand drawing can make an accurate representation of what he sees, at least as regards outlines, and if a stage micrometer (a glass slide ruled in hundredths of a millimeter or thousandths of an inch) be substituted for the object, it also can be sketched in, and will then form a scale by which the size of the object can be directly measured on the sketch. If the microscope is provided, as it should be, with a graduated draw-tube by which the magnifica- tion can be varied, it is easy to bring the image given by any objective and eyepiece to such a size that some even number of divisions of the stage-micrometer coincide with one square of the eyepiece glass ; and if this is noted once for all, and the sketch is made with the same length of draw-tube, no further measure- ment is required. Some microscopes have an eyepiece scale instead of a square ruling, and this can be used in the same way for measuring, but is not convenient for drawing. Drawings may also be made by the camera lucida, which consists either of a prism or transparent mirror, fitting on the eyepiece in such a way that, by placing the microscope horizon- tally on a box or other stand, so as to raise it about 10 inches above the table, and looking down through the camera at right angles to the body of the microscope, the reflected image of the object can be seen at the same time as a sheet of paper laid or fastened on the table on which it can be drawn. Instruments are also constructed in which the horizontal position of the microscope is unnecessary. The focus of the microscope must * Lee (A. B.) 'The Microtomist's Vade-mecum.' Handbook of Methods of Micro- scopic Anatomy. Sixth Edition, 1905. (Churchill.) t Zimmermann (A.) * Botanical Microtechnique.' A Handbook of Methods for the Preparation, etc., and Microscopical Investigation of Vegetable Structures. Translated by J. E. Humphrey, 1896. (Archibald Constable and Co.). 406 LEATHER INDUSTRIES LABORATORY BOOK. be so adjusted that lines on the paper can be seen clearly at the same time as the object, and that slight movement of the eye causes no movement of the one over the other. The light must also be so regulated by the iris or otherwise that both can be seen at once, which is not always easy. By using square-ruled paper, it is obvious that objects can be drawn to scale and measured as before described. A simple camera lucida may be improvised from a thin cover-glass fixed at an angle of 45 with the eyepiece, with cork and cardboard. The dimensions of microscopic objects are now generally given in JJL, or thousandths of a millimeter ; 0*001 of an inch is approximately 0*025 millimeters, or 25 /z. Drawing is now to a considerable extent superseded by photographic methods, which are not only more accurate, but effect a great saving of time where many representations are required, and involve no very costly apparatus, nor special skill, unless very high powers are employed. Very good work has been done with an ordinary ^ -plate camera lengthened with an improvised cardboard tube. It is desirable that both micro- scope and camera should be screwed or clamped on to a single base-board, which may also carry the illuminating lamp, for whrch an incandescent burner for gas, or a Nernst projection lamp for electricity is well adapted. It is almost essential if a long camera is used to devise some means of adjusting the focus from the screen end, either by moving the latter, or by mechanically operating the focusing screw, which in some microscopes has a groove round which an india-rubber band can be placed to act as a belt. It is also very convenient to be able to adjust the posi- tion of an object by a mechanical stage operated from the screen. Powers of moderate angle are usually most convenient for ordinary work, and where any "depth of focus" is needed, a somewhat low power and a long camera gives the best results. Unless a special projection eyepiece is available, it is usually best to use the objective only, which then takes the place of the ordinary camera lens. Different objectives of equal quality for ordinary work sometimes differ widely in their photographic properties, and as they are not specially achromatised for ultra- violet rays, some correction of focus is often required, usually in the direction of shortened focus. If a photograph shows less definition than the visual image, a first approximation may be THE USE OF THE MICROSCOPE. 407 made hy throwing the visual image out of focus to approximately the same degree of indistinctness ; and the correction, once found, can be noted for future use. The employment of isochro- matic (and backed) plates tend to lessen the chromatic error. Colour screens may also be used. One of the most important applications of microscopic exam- ination for the tanner is in the detection of adulteration in sumachs and other finely ground tanning materials. In the case of sumach especially, adulteration has been in the past rather the rule than the exception, and as some of the adulterants are not only inferior in tanning strength, but contain catechol tannins which redden on exposure to light, their detection is of considerable im- portance. In the case of leaf sumach, it would not be difficult to pick out the fragments of rJius coriaria by direct examination under a low power with reflected light, as the hairs on the leaf are very charac- teristic, but in the fine powder this is impossible, while it is too opaque for examination with higher powers and transmitted light. The first and perhaps the most complete work on the structure and adul- terants of sumach was done by Andreasch,* when during his last illness he was obliged to winter in Sicily ; and will well repay study, but it is too detailed for useful abstraction, and has in some respects been superseded by later methods. A valuable observation was made by Lamb and Harrison f which affords the most convenient method of detecting the usual adulterants. While the internal structure of most leaves is completely destroyed by a short heating with strong nitric acid, the upper and lower cuticles are much more resistant, and often have very marked and characteristic structures. The cuticles * Sicilianische Sumach und seine Verfalschung, Wien, 1898. t Journ. Soc. Dyers and Colorists, March 1899. FIG. 28. Rhus coriaria. 4o8 LEATHER INDUSTRIES LABORATORY BOOK. themselves vary much in resistance to the reagent, that of sumach itself being delicate and easily destroyed, while that of the common adulterant pistacia lentiscus (lentisque, pistacio) is extremely strong, and remains almost uninjured after the sumach is totally destroyed. Probably the best mode of procedure is that given by Priestman,* who heats the samples with strong nitric acid in test-tubes, in a water-bath containing about 750 c.c. of water, over a single Bunsen burner, beginning cold, and raising the bath to a definite temperature. After the heating is com- pleted, the acid is neutralised with ammonia or caustic soda, and the sediment is allowed to settle, and repeated portions (7 or 8) are examined under a cover-glass. For direct examination, staining is not usually necessary, since most cuticles, and especially pistacio, are coloured distinctly yellow by the nitric acid, but for photographic purposes staining with safranine or acid red, and the use of isochromatic plates is advisable. For permanent mounting, the stained and washed cuticles may be dehydrated with alcohol, and mounted in balsam or dammar in the usual way (p. 423), though some delicate cuticles photograph better if dehydration is omitted, and glycerine jelly is used as a mountant. The photographs on plates I., II. and III. were taken with a Zeiss B objective, which was also frequently used for direct examination. The following table summarises the effects of heating for various of the more common sumach substitutes and adulterants. Cuticles separate at Totally dissolved at Rhus coriaria leaves 60-65 in I0 mins. 98 in 15 mins. from cold. stems 75 10 95-98 15 Tamarix stems 80 10 90 15 Pistacia lentiscus 65 IO 9 8 20 Coriaria myrtifolia 75 10 98 20 Rhus metopium .;' 70 15 i 98 2 5 Rhus glabra .. 75 15 Insol. Insol Colpoon compressa Ailaiithus glandulosa 80 75 15 15 98 ,,25 , Insol. Insol > j > RJius coriaria. Both upper and lower cuticles separate at 70-80 C. without manipulation, but are very delicate and trans- parent, and the hairs, which are extremely characteristic, and quite unlike anything found in the adulterants, can hardly be seen with- * Journ. Soc. Chem. Ind., p. 231, 1905. Collegium, pp. 184, 199, 1905. PLATE i. Rhus coriaria, upper. X 84. 3. Rhus coriaria, lower. X 84.' 5. Pistacia lentiscus, upper. X 84. 2. Rhus coriaria, upper. X 84. 4. Rhus coriaria, stem cuticle. X 6. Pistacia lentiscus, lower. X 84. [To face page 408. PLATE 7. TamariX) cuticle. X 84. 8. Tamarix, cuticle and fibre core. X 42, , TamariX) core fibres. ^Jx 84. lo. Arbutus unedo, substance and cuticle. X 42. 1 1. Ceratonia siliqua, upper. x 84. 12, Vitis venifera, upper. X 84. PLATE III. 13. Coriaria myrtifolia, upper, x 14. Coriaria myrtifolia, lower, x 84. 15. Ailanthus glandiilosa, upper. X 84. 16. Ailanthus glandulosa> upper. X 84. 17. Colpoon compressa, upper, x 84. 18. Colpoon compressa, lower. X 84. THE USE OF THE MICROSCOPE. 409. out staining. The hairs take the dye much more readily than the cuticles themselves. (Plate I., Figs, i, 2, 3. Stem cuticle,. Fig. 4-) Rhus glabra, an American sumach, has small club-shaped hairs, and minute stomata on its lower cuticle, which is extremely thin and transparent, and very difficult to see or photograph. The upper cuticle is strong, and resists boiling nitric acid for a considerable time. It has a well defined cell-structure without stomata, but with frequent vein- marks. (Fig. 29.) RJms metopium, another American sumach, has no hairs, and strongly resembles pistacio in the form of its cells. Pistacia lentiscus (Schinia, lentisque) was formerly the prin- cipal adulterant, but has proved so easy to detect that its use is diminishing. Its cuticles are strong, and resist acid well, their cell structure is characteristic, and they dye deep yellow with nitric acid. The reaction with nitric acid is usually more violent than with pure sumach. The cuticles separate well, and will stand the action of boiling nitric acid a few minutes longer than real sumach (Plate I., Figs. 5, 6). Only the lower cuticle has stomata, which are numerous. The cuticles are free from hairs, and but little distorted over the veins. Coriaria myrtifolia (" stinco," redoul, French sumach) has a well defined cell-structure in both upper and lower cuticles, the cell-walls are defined, and the cuticles are easily separated by warming to 65 or 70 C. in pure nitric acid. The structure somewhat resembles pistacio, but there is a curious waviness in the well defined walls which makes it easy to distinguish. The upper cuticle has no stomata, and the cells are elongated and distorted over the veins, a further distinction from pistacio. The lower cuticle has numerous well defined stomata. (Plate III., Figs. 13, 14.) KlG. 29. Rhus glabra. 410 LEATHER INDUSTRIES LABORATORY BOOK. Tainarix Africana (tamarisk), is a shrub with very minute leaves, so that the powder used in adulteration consists mainly of stem. The stem-cuticle has a well-defined structure of cells elongated in the direction of growth, with stomata and no hairs. The cuticle begins to separate at 70-75 C, but the woody fibre is not destroyed below 80-85 C., and as the cuticle itself is soluble a little over 90 it is well to maintain the heat for some little time below 75 rather than use a higher temperature for a shorter time. The true sumach stems behave in a very similar Avay, and their woody tissue much resembles that of Tainarix, but the cuticle of the lattter is easily distinguished by the ab- sence of hairs. Ground Vine Leaves. The cuticles are both thin, difficult to separate from the leaf-substance, or to observe under the micro- scope, and their structure is not very characteristic. The vine leaf contains scarcely any tannin, but injures the colour of the sumach in tanning. Ceratonia siliqua (carob or locust bean, caroubier). The structure of the leaf-cuticles so much resembles that of pistacio that it would probably be impossible to swear to one or the other if present in small quantities. This adulterant seems to be increasingly employed. Arbutus unedo (arbutus, arbousier) seems less used than formerly, but in structure so closely resembles Tamarix that definite distinction is often impossible. Plates of several other adulterants are given, but for further details the original papers must be consulted. It is always wiser if possible, to confirm identifications by direct comparison with known specimens, rather than trust to photographs merely. Beside microscopic examination, the chapter on the qualitative recognition of tannins by chemical tests (Sect. IX.) may be con- sulted, but on the whole chemical tests have not proved very reliable, and testing for catechol tannins by bromine water seems the most important. It is impossible for the chemist to recog- nise all the various vegetable materials which may be added to sumach ; sumach stems and inferior sumachs can hardly be con- sidered as adulterants, but may largely reduce the tanning value, and it is always possible to make use of exhausted and re-dried sumach leaves, which of course carmot be detected microscopically. If a sumach shows 28 per cent, of tanning THE MICROSCOPIC STRUCTURE OF SKIN. 411 matter or over by the I.T.A. test, and the absence of catechol tannins by bromine water, there is not much fear of adultera- tion, as few or no tanning materials of that strength are cheaper than genuine sumach. Dr. Hellon states that the following treatment is very effec- tive in giving a beautifully transparent residue for microscopical examination, and possesses the further advantage over the strong nitric acid process that there is no gelatinising of the vegetable matter and consequently the residues settle so rapidly from the various disintegrating liquors that the treatment may be com- pleted without setting down the test tube out of the hand. 1. A decigram of the sample is boiled for ^ minute with say 7 c.c. 5 per cent. H 2 SO 4 , and the clear liquid poured away. 2. The residue is then boiled for \ minute with 7 c.c. 5 per cent. NaHO solution, and the liquid poured away. 3. The residue is boiled for J minute with 7 c.c. HNO 3 sp. gr. I * 2, and the clear liquid poured away. 4. The residue is boiled as before with another portion of HNO 3 , and during the boiling a centigram of KC1O 3 is added from time to time until the substance is absolutely white. The clear liquid is poured away, and the residue washed free from acid with warm water, transferred to the microscope slide, covered and examined. This method certainly gives beautiful preparations, though in the Author's laboratory it has not superseded that of Priest- man. SECTION XXV. THE MICROSCOPIC STRUCTURE OF SKIN. THOUGH a great variety of skins of the Mammalia are em- ployed in tanning, they all possess the same general structure, and an anatomical description of that of the ox will apply almost equally to those of the sheep, goat and calf, although from varieties in thickness and closeness of texture, their practical uses differ widely. The differences between these and fish, lizard, alligator and serpent skins are much more marked. 412 LEATHER INDUSTRIES LABORATORY BOOK. In its natural condition, the skin is not merely a covering for the animal, but at the same time an organ of sense, and of secretion, and hence its structure is somewhat complicated. It FlG. 30. Vertical Section of Calf-skin, magnified 50 diam. : a, epithelial layer or epidermis, consisting of horny layer above, and rete malpighi below ; b, pars papillarif, and c, pars reticularis of corium \ d, hairs ; e, sebaceous or fat-glands ; f t sudoriferous or sweat-glands ; g> opening of ducts of sweat-glands ; A, erectores pili muscles, for erecting the hair. FIG. 31. Development of young Hair. FIG. 32. Section of Epidermis. consists of two principal layers, the epidermis (epithelium, cuticle) and the corium (derma, cutis, or true skin). Ttiese are totally distinct, not only in structure and functions, but in their origin. THE MICROSCOPIC STRUCTURE OF SKIN. 413 In the egg of birds and the ovum of higher animals, the living germ consists of a single cell, which, as soon as fertilised, begins to multiply by repeated division. The mass of cells thus formed early differentiates into three distinct layers, from the upper of which the epithelium arises, while the true skin, together with the bones and cartilages, is derived from the middle one. This distinction of origin corresponds with a wide difference of both anatomical and chemical characteristics. The epidermis is very thin as compared with the true skin which it covers, and is entirely removed preparatory to tanning ; it nevertheless possesses important functions. It is shown in Fig. 30 at a, and more highly magnified in Fig. 32. Its inner mucous layer b, the rete malpighi, which rests upon the true skin c f is soft, and com- posed of living nucleated cells, which multiply by division and form cell-walls of keratin. They are elongated in the deeper layers, and gradually become flattened as they approach the surface, where they dry up, and form the horny layer a. This last is being constantly worn away, and thrown off as dead scales of skin ; and as constantly renewed from below, by the multipli- cation of the cells. It is from this epithelial layer that the hair, as well as the sweat- and fat-glands, are developed. It will be seen in Fig. 30 that each hair is surrounded by a sheath, which is continuous with the epidermis. In embryonic development, a small knob of cells forms on the under side of the epidermis, opposite to a knot of a capillary blood-vessels in the corium, and this enlarges and sinks deeper into the true skin, while the root of the young hair is formed within it surrounding the capillaries, from which it derives nourishment and which forms the " hair papilla" ; this is shown in Fig. 31, # b. Smaller projections also form on the stalk of the knob, and in due time produce the sebaceous glands. The process of development of the sudo- riferous glands is very similar to that of the hairs. There is a great analogy between this process and that of the ordinary renewal of hair in the adult animal. At aP, Fig. 30, is seen an old and worn-out hair. It is shrunken and elongated, and is almost ready to fall out. It will be noticed that its sheath or follicle projects somewhat below the hair to the right. This is the first stage in the production of a young hair, and is quite analogous to the knob of epithelium which forms the starting- point of a hair in embryo. At d^ the process is further advanced, LEATHER INDUSTRIES LABORATORY BOOK. the young hair growing into the old sheath. At d 3 it is complete, the old having fallen out, and the young one taken its place. The hair itself is covered with a layer of overlapping scales, like the slates on a roof, but of irregular form. These give it a serrated outline at the sides, strongly developed in wool and some furs, to which it gives the property of felting. Within these scales, which are called the " hair cuticle," is a fibrous substance which forms the body of the hair ; and sometimes, but not FIG. 33. Hair Bulbs and Sudouferous Glands. FIG. 34. The same more highly magnified. always, there is also a central and cellular pith, which under the microscope frequently appears black and opaque, from the optical effect of imprisoned air. On boiling or long 'soaking in water, alcohol or turpentine, the air-spaces become saturated with the liquid, and then appear transparent. The fibrous part of the hair is made up of long spindle-shaped cells, and contains the pigment which gives the hair its colour. The hair of the deer differs from that of most other animals in being almost wholly formed of polygonal cells, 'which, in white hairs, are usually filled with air. At its base the hair swells THE MICROSCOPIC STRUCTURE OF SKIN. 415; into a bulb, which is hollow, and rests on the hair papilla //, Fig. 35 (y. s.), which has blood-vessels and nerves, and supplies nourishment to the hair. The hair-bulb is composed of round soft cells, which multiply rapidly, and press upward through the hair-sheath becoming elongated and hardened, and forming the hair. In dark hairs, both the cells of the hair and its sheath are strongly pigmented, but the hair much the more so, and hence the bulb has usually a distinct dark form. The dark-haired portions of a hide from which the hair has been removed by liming still remain coloured, from the pigmented cells of the air-sheaths^ which can only be got out completely by bating and scudding. The cells outside the bulb, shown at f in Fig. 35, pass upwards as they grow, and form a distinct coating around the hair, which is called the " inner root-sheath." This again consists of two separate layers, of which the inner is "Huxley's," the outer " Henle's." They arise from the same cells in the base of the hair ; but in the inner layer these remain polygonal and nucleated, while in the outer they become spindle-shaped and without nuclei. The inner root-sheath does not extend to the surface of the skin, but dies away below the sebaceous glands. Fig. 35 represents an ox-hair root, mag. 2OOdiam. : a, fibrous substance of hair ; b, hair cuticle ; c t inner root-sheath ; d, outer root-sheath ; , dermic coat of hair-sheath ; f, origin of inner-sheath ; g, bulb ; h, hair-papilla. Outside the inner root-sheath is a layer of nucleated cells, continuous with those of the epidermis, and of the same character. This is the " outer root-sheath," and is shown at d, Fig. 35. This, together with the whole of the epidermis, is separated from the corium by an exceedingly fine membrane, called the hyaline or glassy layer. It is this which forms the very thin buff- coloured " grain " surface of tanned leather, which evidently is of different structure from the rest of the corium, since, if it gets scraped off before tanning, the exposed portion of the pars papillaris remains nearly white, instead of colouring. The whole FIG. 35. 41 6 LEATHER INDUSTRIES LABORATORY BOOK. of the hair-sheath is enclosed in a coating of elastic and connec- tive-tissue fibres, which are supplied with nerves and blood- vessels, and form part of the corium. Near the opening of the hair-sheaths to the surface of the skin, the ducts of the seba- ceous or fat glands (e, Fig. 30) pass into them, and secrete a sort of oil to lubricate the hair. The glands themselves are formed of large nucleated cells, arranged somewhat like a bunch of grapes ; one is shown highly magnified in Fig. 36. The upper and more central cells are most highly charged with fat, which is shown by the darker shading. FlG. 36. &\ See * Journ. Soc. Ch. Ind.,' 1890, p. 27 ; 1893, p. 422 ; ' Brit. Ass. Rep./ 1893, p. 723. The principal ferment of bran drench. Oval cells I '25 /JL X 0*75 //,, frequently found in pairs, dumb-bells or chains. Pure cultures do not attack hide, and only ferment the starch after it has been converted into glucose by cerealin, a zymase or unorganised ferment similar to diastase, and naturally present in the bran. The principal products of the fermentation are lactic and acetic acids, with traces of formic and butyric acids, hydrogen sulphide and tri- methylamine, and considerable evolution of hydrogen and carbon dioxide. In one drench analysed, o* 24 grm. acetic and O'79 g rm - lactic acid per liter were present, and an artificial solution of 0*5 acetic and I *o lactic acid (sp. gr. I *2io) per 1000 produced in i -2 hours an effect on skin quite similar to that of an ordi- nary drench in 12-16 hours, and when tanned gave a satisfactory leather. (Cp. p. 94). Wood * has isolated a second bacterium (B. furfuris ft), very similar in appearance to the first, and taking part with it in the fermentation of ordinary drenches, which it somewhat modifies. No doubt in different tanneries differences exist in the character of the ferments present. When the skins themselves are attacked the effect is generally due to putrefactive bacteria brought in from the puers, etc., but these do not usually produce any effect till the glucose, which forms the nutriment of B. furfuris, is exhausted. In hot weather a butyric fermentation sometimes supervenes, when the skins become " glassy " and swollen, and are rapidly destroyed (see p. 434). B. furfuris a and /3 belong to the group of Coli bacilli, and are capable of fermenting milk with production of lactic acid.f B.mirabiliSy Proteus mirabilis (Hauser). Mobile rods 2-3. /A * 'Jour. Soc. Ch. Ind.' 1897, p. 510. f On Manufacture of Lactic Acid, 'Jour. Soc. Ch. Ind.,' 1897, p. 516. 2 F 2 436 LEATHER INDUSTRIES LABORATORY BOOK. X O'6/u,. Agent of putrefaction of animal substances. Aerobic, liquefies gelatine rapidly in presence of oxygen. Colonies on plates produce long tortuous prolongations. B. vulgaris, Proteus vulgaris (Hauser). Almost identical with the above. Mobile rods 1*25^ x 0*8 //,. Compare Hauser, Ueber Faulnissbacterin,' Leipzig, 1885. B. vulgaris is the commonest putrefactive bacillus. B. mirabilis differs from the above chiefly in the mode of growth on solid culture media.* Many putrefactive bacteria are capable of producing acids from sugar (vide Rahm, O., Centr. Bl., xiv. p. 22). B. termo. Not now considered a distinct species, but com- prises the two last and many other putrefactive bacteria* FlG. 41. -B.Jurfuris a. B. nrece. Slender rods, less than I /i in diam. Can grow without oxygen. Causes ammoniacal fermentation of urea. Wood has found in " puer " liquor Proteus vulgaris and B. subtilis, among others, and probably Bacillus urece (Cohn) and M. fulvus. Thiothrix or Beggiatoa, sulphur bacteria. Long threads, living in waters containing sulphuretted hydrogen, and assimi- lating sulphur. Beggiatoa alba. One of the most common. Threads dirty white, 2 or 3 p, in diam., containing granules of sulphur, common in sulphur springs as well as in soak pits and other waters con- taining sulphuretted hydrogen as a product of putrefaction. There are several other similar species. Their life-history has been specially studied by Winogradsky.f * For a short account of putrefactive organisms, see a paper by Wood, ' Journ. Soc. Chem. Tnd.,' 1906, p. 109. See also Lafar, p. 741, * Ueber Protein Faulniss.' f ' Morphologic und Physiologic der Schwefelbacterien,' Leipzig, 1888. 13 A C TERIOL O G V. SACCHAROMYCETES. 437 In addition to the true bacteria which act as ferments, there are certain other fungi which also have the power of producing fermentation ; especially those belonging to the class of Saccha- roniycetes or yeasts, which are the cause of ordinary alcoholic fermentation, although certain true bacteria have the same power. The yeasts, unlike the true bacteria, multiply by bud- ding or spores, and not by division. The different species of FIG. 43. YEAST. yeast vary in size, and are more or less round or elongated in form, but in general resemble the ordinary beer yeast closely in appearance, which may be conveniently studied in brewers' or German pressed yeast. Saccharomycetes my coder via is very frequent at the surface of tanning liquors, forming the thick and wrinkled scum which is so familiar. The cells are usually oval or elliptic, but sometimes cylindrical, 6-7 yu, X 2-3 //,, often forming thick and branched chains, and occasionally taking forms very similar to the hyphae 438 LEATHER INDUSTRIES LABORATORY BOOK. (see below) of the higher moulds. It was formerly supposed that it was an acetic acid ferment, but the true acetic acid fer- ment (see p. 434) is a bacterium frequently forming part of the scum produced by the mycoderma. It does not appear to produce alcohol in perceptible quantities, though it frequently causes a vinous odour, and, when freely supplied with oxygen, oxidises and destroys the acetic and other acids produced by bacterial fermentation. There are several other species of mycoderma which are either identical or very closely allied to the above. A few species of Sacckaroinyces produce lactic acid, e.g. Saccharo- myces acidi lactici (Grotenfelt). Andreasch has also isolated three species of yeasts from sour tan liquors which produce lactic acid. The branching stems or hyphae of the higher moulds, and especially of the mucors, appear to act as ferments when immersed in liquids, but the subject cannot be considered yet as being fully investigated. VanTieghem ('Ann.d. Sc.Nat.,' Ser.5, 1867, p. 210) showed that the mycelium, both of Penicillium and Aspergilhts, when growing in a solution of tannin, breaks up the tannin into gallic acid and FIG. 44. PenicilHum glaucnm glucose. Although their study does not properly belong to the subject of Bacteriology, it will be convenient here to mention one or two which are sure to be encountered by the microscopist. ASCOMYCETES (spores in tassels or heads unenclosed in sporangia). Penicillium glaucum, green mould, the most common of the mould fungi, is shown in Fig. 44. In its earlier stages it con- sists of a mass of hyphae or branched stems, composed of elon- gated cells, and forms a white mould. As it becomes developed it throws up stems, on the top of which the spore-bearing cells BACTERIOLOGY. 439 (conidia) appear, which, under the microscope, have the aspect of a tassel of bluish-green beads. This is most easily seen by examining dry, as the tassels immediately break up in the presence of moisture into single spores, each of which is capable of producing a fresh plant. The spores are about 3 * 5 //, and the mycelia from 0*7 to 4 yu, in diameter. Aspergillus spores grow on short conidia clustered on a club- shaped head. Aspergillus glaucus t pale green or bluish spores. FiG. 45. Aspergillus, showing spores, mode of conjugation, and eurotium form. FIG. 46. Fructification of Mucot. A. niger (Fig. 45), a dark chocolate brown mould, which grows well on acid fruits and liquids. This genus also produces large " resting spores " (^Eurotium) by sexual fructification, which were formerly considered to belong to a separate fungus and known as eurotium. PHYCOMYCETES (spores in membranous sporangia). Pilobolus. Hyphse 1-2 mm. high, glassy tufts ; occurs on excrement. Mucor mucedo, white mould ; hyphae colourless, simple or branched, 115 cm. long ; sporangia brown or black, thus dis- tinguished easily from those of P. glaticum which are bluish- green (Fig. 46). The moulds have much more destructive effect on tannins than the bacteria. Eitner* showed that the introduction of * Gerber,' 1889, p. 2. 440 LEATHER INDUSTRIES LABORATORY BOOK. putrefaction bacteria into tan-liquors had no destructive effect on the tannins, and the same was proved by Kohnstein and Simand with regard to the lactic ferment. * On the other hand, it is well known that tannins are quickly destroyed by moulds ; Wood f mentions that the tannin in a sterilised sumach infusion remained practically constant for 23 days, while in one infected with Penicillium glaucum, 80 per cent, of the total was destroyed in the same time. Bacteriological Methods. It is generally useless to attempt bacteriological work of any elaborate kind without a separate and properly fitted laboratory, from the air of which germs can be rigorously excluded ; and therefore only a few simple methods of preliminary experiment which can be carried out reliably under ordinary conditions, will be here described. Cultures and inoculations may however be made in any ordinary laboratory, by providing a glass case with a sliding front like a balance case, but somewhat wider, the inside of which is wiped over with an antiseptic solution of carbolic acid and glycerine, and allowed to stand closed, until all germs have become deposited and adhere to the sides of the chamber ; if it be then carefully opened inoculations, etc. may be made by opening the tubes or Petri dishes inside the case, just sufficient aperture being left for the insertion of the operator's hands. His sleeves should be turned back, and hands and wrists washed with carbolic soap or an anti- septic solution. As every fermenting material contains, beside the organism producing the principal change, many others which are acci- dentally present, the first step in bacteriological work is to obtain what are called " pure cultures," containing one species of organism only. It is then possible to study the effects of this organism undisturbed by secondary fermentations, and to deter- mine what products are actually due to it. Since all ordinary materials and objects are swarming with bacteria, a necessary preliminary to pure culture is to obtain liquids and apparatus absolutely sterile or free from living germs, and for this purpose various methods are employed according to the nature of the material to be dealt with. Indiarubber stoppers, and other objects which will not stand a temperature of 1 50 C., maybe sterilised by washing with a * 'Gerber,' 1886, p. 253. t ' Journ. Chem. Ind.,' 1894, p. 221. BACTERIOLOGY. 441 solution of mercuric chloride of cri per cent., and afterwards rinsing with sterilised water. Chloroform and ether may be used for sterilisation of more delicate objects, and are easily driven off by heat, but are not very reliable, but useful as a means of checking the development of bacteria in liquids, without stopping the action of unorganised ferments or zymases (p. 429). Other chemical means of sterilisation may often be employed. N/4 NaOH solution will destroy most bacteria, and may be neutralised with an equal quantity of N/4 HC1, common salt only remaining, or the process may be reversed, mineral acids being also powerfully bactericidal. Hypochlorates, chlorine, or bromine, may also be used and removed or converted into inactive compounds by thio- sulphates or sulphites. Copper and mercury salts may be removed from liquids by SH 2 , and so on. Strong CaCl 2 solutions, or even prolonged salting, will kill many bacteria, and are easily removed by washing. These methods are often useful with hide or other tissues to which heat cannot be applied, when it is desired to sterilise them before submitting them to the action of other bacteria. The most important means of sterilisation, however, is the employment of heat alone. Platinum wires and other objects not easily damaged may be held for a moment in the flame of a Bunsen burner ; glass vessels and plates, cotton-wool and many other objects may be heated in an air-bath to I5OC., or sub- mitted to the action of steam at 110 C. (10 Ibs. per square inch pressure) for half an hour. As steam under pressure, however, involves the use of special appliances, it is often more convenient to employ the lower temperature of steam at atmospheric pressure by which bacteria are killed in the course of an hour, though their spore j are frequently much more resistant and need at least five or six hours for complete sterilisation. In many cases, as in the sterilisation of water and other liquids, it is preferable to heat for an hour to boiling point daily for three successive days, when the spores become developed and are killed by the later heatings. Sterilisation of liquids which are altered by boiling may be generally accomplished by heating daily for four or five hours to 60 C. for a week, though some few bacteria will thrive at this temperature. For the preliminary experiments which are likely to be undertaken in the tannery laboratory, an ordinary potato- steamer will be found to form a convenient steriliser for many purposes, and the air-bath of the laboratory heated to 1 50 C. 442 LEATHER INDUSTRIES LABORATORY BOOK. may also be useful. It is usual to plug the necks of flasks, test- tubes, and other vessels in which liquids are being sterilised, with cotton wool, which of course becomes sterilised at the same time. If from any cause the plug is withdrawn, as for instance to take samples of the liquid, it must be singed in the flame before it is replaced, in order to destroy any germs which have fallen upon it, and the vessel during the process must be held in an inverted or sloping position so as to lessen the danger of germs falling into the liquid ; and it need hardly be said that the platinum wire or pipette, which is employed for drawing samples, must itself be thoroughly sterilised before use. An- aerobic bacteria may be cultivated by placing the culture tube inside a wider tube 3 c.m. x 25 c.m., containing I grm. pyrogallol and 10 c.c. of a strong solution of caustic potash. The tube is then closed by an air-tight indiarubber stopper. They will also frequently develop in the lower part of solid media (gelatine, agar-agar) in test-tubes. Various nutritive materials are employed for the culture of bacteria. In Pasteur's researches, liquids were generally used ; but at present it is found convenient in many cases to employ solutions of gelatine or agar-agar, which at ordinary temperatures form a solid jelly. Many bacteria have the power of liquefying gelatine, probably by the production of a zymase ; and to the leather manufacturer it is not without importance to distinguish these, as there can be little doubt that they are active in pro- ducing solution of the hide fibre. Pasteur's original solution consisted of I grm. ammonium tartrate, 10 grm. crystallised sugar, and the ashes of I grm. of yeast, dissolved in 100 c.c. of water. Buchholz modified this nu- tritive solution by substituting 0*5 grm. of calcium phosphate for the yeast ashes. Cohn employed o- 10 grm. potassium phosphate, O' 10 grm. crystallised magnesium sulphate, 0*01 grm. tribasic cal- cium phosphate, 0*20 grm. ammonium tartrate and 20*00 c.c. of water. Many bacteria will live in these solutions, though they were originally designed for the study of yeasts and fungi. Bacteria usually thrive much better in meat broths than in these saline solutions. An ordinary mode of preparation is as follows : 250 grm. of beef muscle carefully fre*ed from fat is gently simmered for five hours in I liter of water, the fat as it rises being skimmed off. It is allowed to stand overnight in a BACTERIOLOGY. 443 cold place, solidified fat is removed, and the liquid made neutral to litmus by dilute caustic soda solution. It is then again boiled for ten minutes, filtered through paper previously wet with water, made up to one liter and measured into experimental flasks or test-tubes, in which it is sterilised. Many other nutritive solutions are employed, for the prepara- tion of which the bacteriological text-books may be consulted. In many cases it is advantageous to cultivate bacteria in the medium in which they normally exist ; thus, " drench " bacteria may be grown in a filtered and sterilised drench liquor, and the bacteria of bates, puers, etc., in their appropriate liquors simi- larly treated. The methods of cultivation in gelatine are, as a rule, more easy to handle successfully in ordinary laboratories than those in liquids, and are therefore of special interest for technical use. A suitable jelly for this purpose may be made by dissolving 6-15 per cent, (according to temperature) of the finest French gelatine in the nutritive broth which has been described ; or 1-2 per cent, of dry meat peptone (peptonum siccum), and if desired, I per cent, of pure glucose may be added to a simple solution of gelatine in water. These solutions may require to be neutral- ised with soda similarly to the meat extract, and must be purified by repeated filtrations. It is frequently necessary to clear with white of egg, in order to obtain perfectly bright solu- tions ; this is done by adding to each liter the white of one egg, which has been previously beaten to a froth with a little water. The temperature of the liquid must not exceed 4O-5o C. when the egg is added, but is then raised to boiling point, or even heated to 110 C. or more to coagulate the albumen, which carries down with it all suspended matter, and may be removed by filtration through flannel or paper. The coagulation of the egg should always be performed at a temperature at least as high as that which will be afterwards employed in sterilising, or a further precipitation may take place during that process. Where somewhat high temperatures are required in cultures, agar-agar is substituted for gelatine, as its jellies have a higher melting point, but this will not often be required in tannery work. The jelly after filtration is usually sterilised in test-tubes, 5 or 10 c.c. being measured into each, and the tube plugged with cotton-wool. The plugs should be of the pure cotton-wool used for medi- 444 LEATHER INDUSTRIES LABORATORY BOOK. cal purposes, tolerably light and compact, and covered during sterilising with tinfoil to lessen the destructive action of the steam. For some purposes (roll-cultures, p. 445) small cylindri- cal bottles of clear glass, and especially the thin blown bottles used for oil samples, are to be preferred to test-tubes. Flasks and bottles with flat sides have also been used. Several dozen tubes or bottles may be sterilised at once by placing in a wire rack in the steamer, and steaming for an hour daily for three successive days. Properly sterilised tubes should remain clear and unaltered for an unlimited length of time, but though the cotton plugs exclude all germs, in an ordinary laboratory moulds are apt to develop on them, and in the course of weeks to grow through and infect the sterilised media. This danger can be lessened by covering the plugs, after singeing, with a rubber cap or even with tinfoil, but these will also exclude air, which is indispensable to the development of many bacteria. Cultivation. As it is extremely difficult to avoid the admis- sion of germs from the air of an ordinary laboratory into nutrient liquids which are to be inoculated (see, however, p. 440), the employment of the nutrient gelatine will be first described. To inoculate this, a platinum wire (fused into a piece of glass tube) is usually employed, which may be either straight (needle), or bent into a loop (Oese}. This is ignited for a moment in the flame to sterilise it, then allowed to cool, and dipped in the liquid containing the bacteria. A test-tube of sterilised gelatine is held in the left hand upside down, and the plug of cotton-wool is drawn out between the little finger and second finger of the right hand, and the platinum needle (which during this operation has been retained in the hand) is thrust into the gelatine almost to the bottom of the test-tube, quickly withdrawn, and the plug of wadding replaced, frequently after first singeing it in a flame. In case of what is called "tube" or "stab " culture, the tube is now simply retained at a suitable temparature, which may vary according to the object of the experiment, from that of the laboratory up to anything below the melting point of the jelly. The bacteria multiply upon the sides of the prick, pro- ducing in many cases very distinctive appearances, some forms liquefying the gelatine and others spreading into it without liquefaction. Anaerobic bacteria only develop in the deeper parts of the tube, while aerobic are confined to the surface, while some species (" facultative " aerobic or anaerobic) can thrive BACTERIOLOGY. 445 under both conditions. Some species of bacteria also produce brilliant colouring, or fluorescence. Illustrations of these stab cultures may be found in most bacteriological text-books. This method, however, is only suitable for pure cultures, as it affords no means of separating the different bacteria. To accomplish this, Koch introduced the method of " plate culture," which consists in pouring the melted gelatine upon levelled plates of glass on which it is allowed to set, and which are then carefully protected under a bell glass from the possibilities of contamina- tion by aerial germs. These plates have been almost superseded by the " Petri dish," a flat glass box of about 4 inches diameter with a cover, as the latter are much easier to handle. In this case the bacteria, if not too numerous, develop at separate points upon the plate, forming colonies which may be counted, and which usually consist of one species of bacterium only. By inoculating fresh tubes from these colonies and repeating the process, pure cultures are easily obtained. From the difficulty of excluding aerial germs the method is difficult in a general technical laboratory, but a modification of it (" roll culture ") suggested by Esmarch, may be successfully adopted. In this case the test-tube containing the gelatine, instead of being poured on a glass plate, is simply heated by immersion in water sufficiently warm to liquefy the solution, and is then revolved in an horizontal position so as to spread the gelatine uniformly over its sides, and at the same time is cooled by water from a tap, or by the application of ice. As in this operation the solution comes in contact with the plug of cotton- wool, which may contaminate it, and which it is apt to render impervious to air, it is advantageous to substitute for test-tubes the thin blown cylindrical bottles which have been mentioned on p. 444, in which the gelatine may be uniformly distributed by rolling in a shallow tray containing a small quantity of ice- water. The colonies will then develop upon the sides of the bottle, through which they may be examined with a low power of the microscope, or with a simple lens, to distinguish whether they liquefy the gelatine and to observe their other characteristics of growth and appearance. Portions of colonies may be with- drawn with a piece of platinum wire for the inoculation of fresh tubes. In some cases it is desirable, in order to obtain an approximate idea of the number of bacteria present in a water or fermenting 446 LEA THER IND US TRIES LA BORA TOR V BOOK. solution, to count the colonies which are formed ; this is most readily done by partially surrounding the bottle or test-tube by a piece of black paper in which an aperture exactly I centimeter square has been cut with a penknife. For plate cultures or Petri dishes, cover-glasses etched with a square of I cm. divided into squares of 2 mm. are still more convenient where the colonies are numerous. For roll cultures squares scratched on mica or celluloid may be used. . The number of colonies which can be counted with a lens on I cm. is determined in various parts of the tube, and the average multiplied by the whole area of the surface of the plate or tube measured in centimeters will give the number of colonies produced by the quantity of material introduced into the tube. The tube must be watched, as different species of bacteria develop with varying rapidity, and liquefying species soon cause the colonies to run together. It usually happens, especially with liquids occurring in the tannery, that the number of bacteria in a single drop, or even in the quantity of liquid taken up by a platinum wire, is so great that the colonies are too close together and run into one another. To avoid this a sterilised tube is inoculated, its contents melted and mixed by rolling, and from the mixture another test-tube is similarly inoculated, and from this again a third. In this way three successive dilutions are produced, one of which will pro- bably be of the right concentration to give satisfactory results ; and if a known quantity of liquid be taken, the dilution can be calculated, and from it the number of bacteria in the original liquid. For this purpose a drop may generally be reckoned as the 2Oth of a c.c., and if the test-tube contained 5 c.c. of solution, each dilution would be lOOth of the previous one. In infecting a tube with a drop of liquid, the gelatine should be liquified before withdrawing the cotton-wool, and the tube should be held in a sloping direction, to avoid as much as possible the entrance of atmospheric dust. Where larger dilution is required sterilised water may be advantageously substituted for the nutri- tive medium. The method of obtaining pure cultures with nutrient liquids is similar in principle to the above, but the dilution must be carried much further, in order that, out of a number of tubes infected, a certain proportion should not receive more than a single germ. The contents of the tubes after incubation are examined microscopically, and those taken which appear only to contain BACTERIOLOGY. 447 one species ; and, for the sake of additional safety, the processes of dilution and inoculation may be repeated in a further num- ber of tubes. As large dilution is required, it is better to use sterilised water for the purpose in the earlier stages. Microscopic Examination of. Bacteria. Solutions containing bacteria may be examined by placing a minute drop on a glass slip, covering with a thin cover, and employing a high power objective, as described in Sect. XXIV. For the mere recognition of the presence of bacteria a good dry 6th may be sufficient, but for detailed study a good oil immersion I2th is necessary. Abbe's illuminator, or some other condenser, is preferably used for lighting, generally with a narrow aperture of the diaphragm. If bacteria are to be studied in tissues such as the skin, sections must be made as described on p. 422, and the bacteria stained. It is frequently found that the recognition of bacteria, espe- cially in tissues, is extremely difficult, from their minuteness and absence of very characteristic form. This difficulty can be largely overcome by the use of stains which colour the bacteria, while leaving other parts of the tissue unaffected. The bacteria generally behave to staining agents very similarly to the nuclei of more highly organised cells. When the bacteria exist in liquids it is necessary, before staining, to attach them to a cover- glass to prevent their being washed away. The cover-glasses are made absolutely clean, by heating first in concentrated sulphuric acid, then rinsing in water, washing with alcoholic ammonia, and finally drying with a linen cloth perfectly free from grease. A small trace of the liquid containing the bacteria is then spread on the glass with a platinum wire, or distributed in a thin film by pressing a drop between two cover-glasses which are separated by sliding one after the other. The glasses are then dried in a desiccator over sulphuric acid or calcium chloride, and the bacteria are finally fixed on the glass by draw- ing the latter, held in the fingers, with its coated face upwards, three times, somewhat slowly, through the flame of a Bunsen burner or spirit lamp. Some little experience is necessary to learn the right amount of heating to employ, too much heat dis- organising the bacteria and preventing their staining, while too little fails to fix them firmly on the glass. If the liquid con- tains gelatine or agar-agar, it is not so readily fixed by heating, from the solubility of these substances in water ; the addition of a small quantity of white of egg renders it less liable to wash off. 448 LEATHER INDUSTRIES LABORATORY BOOK. The difficulty may be also overcome by exposing the prepara- tion for a short time to the vapours of formaldehyde or parafonn, which have the property of rendering gelatine insoluble. Preparations are frequently made from plate cultures (p. 445) by pressing a clean cover-glass og the colony which it is desired to transfer, which then adheres to the glass, sometimes showing very characteristic forms. It is subsequently treated as above to fix it. The basic aniline colours are the most important stains for bacteria ; they are employed dissolved in water, either alone or occasionally with the addition of aniline and other bases. As the aqueous solutions do not usually keep well, it is frequently con- venient to make saturated alcoholic solutions by shaking the powdered colour with alcohol in sufficient quantity to leave a part undissolved. These concentrated solutions are mixed with water in proportion of i-io or more, as may be found necessary to give the required colour. The most generally useful colours are magenta or fuchsine, methyl violet 5 B, gentian violet, methy- lene blue, and Bismarck brown. Staining methods are, however, now so numerous that the reader must be referred to the special text-books. A most useful book of reference is Squire's ' Methods and Formulae ' (p. 427). The sections or the cover-glasses on which the bacteria are fixed are placed in one of these solutions in a watch glass, and remain in it ten minutes or more, until the bacteria are sufficiently- coloured. The solutions are frequently employed warm, where the bacteria do not readily take the stain. Instead of water a saturated solution of aniline, made by shaking three or four grams of aniline oil with 100 cubic centimetres of water, and filtering, is frequently employed to dilute the alcoholic colour solutions. The preparation is then washed with water, which usually removes the colour from the tissue, leaving the bacteria stained, which may now be examined without further treatment, or may be dried and mounted with a drop of Canada balsam dissolved in xylene, which is preferable to chloroform, as less liable to decolorise the bacteria. For the same reason, if a clearing liquid is required, oil of cedar is preferable for the purpose to oil of cloves. In the foregoing Section the writer is greatly indebted to the suggestions of his friend, Mr. J. T. Wood, who" has been good enough both to read and to criticise the manuscript before print- ing, and to assist in examining the proofs, INDEX. ABBE, 295 Abies pectinata, 143 Absorption of non-tannins by hide- powder, 207 tables, 208, 210 Acacia arabica, 136, 147 catechu, 138, 147, 148, 154 Acacias, 142 Acetic acid, 25 Acetyl value, 318, 320 Acid dyes, 375 in hides and skins, 102 oleic, 321 palmitic, 321 stearic, 321 value, 301 determination of, 304 Acidity of chrome liquors, 265 Acids in alum and chrome, determina- tion of, 253 in leather, 366 Air-oven, 193 Albrecht, see Will Alcoholic potash solution, 305 Alkaline salts, titration of, 26 sulphides, 87 Allen, 78, 104, 153, 326, 344 Allihn, 267 Alizarine, 18, 382 yellow, 121 Alkali in soap, 278 Alkalies, determination of, 48 Alkaline carbonates, 35 chromates, 29 Alum, ammonia, 254 and chrome, determination of acids m, 253 potash, 254 Alum, tannages, 249 Alumina, 365 determination of, 250 in leather, determination of chro- mium and, 263 Aluminium sulphate, 254 American Leather Chemists' Associa- tion, 203 official method of tannin analysts, 216 shake method, 203 sumach, 409 Amines, 87 Ammonia, 26, 87 alum, 254 determination, Schloessing, 62 by distillation, determination of, 63 Ammonium salts and ammonia, 71 Ampelopsis hederacea, 117 Analysis of arsenic sulphide, 57 of egg-yolk, 351 of leather, 355 of used liquors and spent tans, 21 5 , 241 Andreasch, 154, 155, 407 Andreasch's reactions with alcoholic solutions of tannin, tables, 164-167 Animal waxes, 346 " Antigalline," 131 Appelius, see Paessler and Schall, 265 Arbousier, 410 Arbutus unedO) 410 Archbutt, 39 Arsenic, determination of, 58 sulphide, orpiment or yellow, 57 Ascomycetes, 438 Ash in egg-yolk, 353 2 G 450 LEATHER INDUSTRIES LABORATORY BOOK. Ash in oils and fats, 288 in tanning materials, 201 Aspergillus. 439 Available" lime, determination of, 54 Azo-colours, 378 BACILLUS, 433, 436 Bacteria, classification, 430 in limes, 90 Bacteriological methods, 440 Bacteriology and mycology, 427 Balance, chemical, 2 Baldracco, 312, 342 Balland and Maljean, 368, 369, 370 Ballantyne, see Thomson Barium hydrate, 21 Barkometer, 8 Barrett, 181 Bartel, 267 Earth and Schroder, 128 Baryta in leather, 364 Basic chrome solutions, 256 dyes, 375 reagent for, 375 lead acetate, 269 Bates, 92 Baum, 176 Baumd degrees, 8 Bayer, 254 Becker, see Popp Bees-wax, 346 Beilstein and Grosset, 253 Bell, 304 B act. furfur is, a and )8, 94 Benedikt and Ulzer, 318, 321 Bennett, see Procter Berthollet and Werner, 106 Bicarbonates, hydrates, carbonates, and, 28 Blackings for boots, 386 for leather, 385 Blank and Finkenbeiner, 82 Blockey, see Procter Boegh, 235 and Thorsen, 334 Borates, 74 % Borax bead, 363 Boric acid, 74 detection of, 74 determination of, 75 in egg-yolk, 352 (Boracic) acid and borates, 26 Boron, 74 Bottcher, 66 Bottinger, 142, 145, 146 Brasilin, 141 Bromine water as reagent for tannins, 149 Brown and Martin, 43 Buchholz, 442 Burettes, 3 CALCIUM hydrate, 21 Calculation of Lowenthal's determina- tion, 229 of results, 1 3 of tannin analysis, 200 Camera lucida, 405 Campbell, 434 Candle filter, 187 Cape sumach, 142 Carbolic acid, analysis of, 78 neutral oils in, 78 water in, 78 Carbonates, alkaline, 35 and bicarbonates, hydrates, 28 Carbonic acid, free, 39 Carbonic acid, determination of free, 39 Carini, 99, 100 Carnauba wax, 347 Caroubier, 410 Carpene, 171. Catechins, 137 Catechol, 105, 116, 117 distinction from pyrogallol and phloroglucol, 118 ortho-carboxylic acid, 129 tannin reactions, tables, 158-160 tannins, 108, 137 Catechutannic acid, 137 Catellagic acid, 123 Cathcart microtome, 403 Caustic alkalies^ 26 Centigrade thermometer, 6 Ceratonia siliqua, 410 INDEX. 45* Ceresin, 3 r 3 - wax, 346 Chlorine, determination of, 39 Cholesterol, 312, 313, 316 Chromates, 255 alkaline, 29 Chrome, determination of acids in alum and, 253 from liquors, recovery of, 265 hide-powder, Kopecky's, 205 - liquor, glucose, 257 liquors, acidity of, 266 solutions, basic, 256 Chromed hide-powder, Paessler's, 198 Chromic acid, 255 determination of, 257 chromate, 256 oxide, 255 Chromium, 254 and alumina in leather, determina- tion of, 263 compounds used in tanning, 254 in solutions of chrome salts, deter- mination of, 262 Chromous oxide, 254 Churn for tanning tests, 240 Clamp for Koch's apparatus, 184 Classification of tanning materials, tables, 157 Claus, 133 " Coal-tar " bates, 84 bate, 92 Coast cod oil, 331 Cod oil, 330 Cohn, 94, 97, 442 Colorimetric determination of traces of iron, 40 Colour determination, 236 Commaille, 172 Commercial materials, testing of, 22 Common salt, analysis of, 69 - in egg-yolk, 353 Complete gravimetric water analysis, 43 Congo red, 18 Connective tissue, 417 Constitution of dyes, 376 Continental microscope, 395 Control of hides and skins, 98 Coriaria myrtifolia, 409 Corium, 417 Correction for absorption of tannin by filter paper, 213 Corypha cerifera, 348 Cotton seed stearine, 327 Creolins, analysis of, 78 Cresotinic acid, 84 Cupric sulphate and ammonia as reagent for tannin, 1 50 Currier's inks, 385 Crystallised oxalic acid, 24 Cylinders, 3 DALE, sec Gladstone Dalican, 294 Darton, 171 Davy, 1 68 Deal shaving reagent for phloroglucol. 152 Decinormal solutions, 22 Decomposition of tannins by dilute acids, 118 Degras, 337 former, 314 specific gravity of, 339 - water in, 338 Dekker, 134 Detannisation, 214 by gelatine, 229 Detection of boric acid, 74 Determination of acids in alum and chrome, 253 of alkalies, 48 of ammonia by distillation, 63 by Schloessing's method, 62 of arsenic, 58 of ash in oils and fats, 288 of " available " lime, 54 of boric acid, 75 of chlorine, 39 - of chromium and alumina in leather, 263 in solution of chrome salts, 262 of formaldehyde, 8 1 of glucose electrolytically, 274 of free carbonic acid (Archbutt), 39 of magnesia, 36 of neutral oils in carbolic acid, 78 2 G 2 452 LEATHER INDUSTRIES LABORATORY BOOK. Determination of sulphur, 58 as sulphide, 55 of sulphuric acid, 46 of " total dry matter " in tanning materials, 192 of traces of iron, colorimetric, 40 of water in carbolic acid, 78 Dieterich, 154 Digallic acid, 130 Diphloroglucolcarboxylic acid, 134 Dipyrogallolcarboxylic acid, 134 Dissolved organic matter in lime liquors, 88 Distillation, 9 Distilled grease stearine, 329 oleine, 332 Dog-dung, 92 Dolphin oil, 310 Drench, chemical examination of, 94 Drenches, 93 Drying apparatus, 193 precipitates, 13 Dyes and dyestuffs, 374 constitution of, 376 dye-tests, systematic, 388 EGG-YOLK analysis, 35 1 ash in, 353 boric acid in, 352 common salt in, 353 fatty matter in, 352 sampling, 351 water in, 351 Eitner, 57, 88, 89, 144, 203, 439 and Meerkatz, 173 Elastic fibre, yellow, 419 Ellagic acid, 121, 123 Ellagitannic acid, 136 English de'gras, 329 microscope, 394 Epidermis, 412, 417 Ernst, 121 Erodin, 92 Estimation of sulphurous acid and sulphites, 27 see Determination Ether value, 303 Etti, 116, 132, 145 Enzymes, 428 Evaporator and dryer, Reed's com- bined, 193 Evaporation, 9 Extraction of tanning materials, 1 80 Extracts, insoluble matter in, 181 FAHRENHEIT thermometer, 6 Fahrion, 341 Fat-cells in connective tissue, 419 extraction, 285 glands, 418 Fats in leather, 357 melting point of, 291 solidifying point of, 291 Fatty acids, 314 matter in egg-yolk, 352 in soaps, 278 Fehling's solution, 267 Fermentation, viscous, 246 Ferric alum as reagent for tannin, 1 50 Filter paper correction factors, 189 for tannin absorption, 188 for tanning solutions, 186 Filtration, 12 of tanning materials, 213 Finkenbeiner, see Blank Fisetin, 141 Fish fat, 327 liver oils, 3^4 oils, 335 Flavellagic acid, 125 Flavones, 141 Flemming, 424 Fluorescence, tanning materials, 154 " Foots," 327 Formaldehyde, 79, 156 detection of, 79 determination of, 81 Formanek, 385 Formic acid, 25 Fractionation, 10 Freda, 131 Free acids in liquors, 242 Procter's method of deter- mining, 243 Koch's method of deter- mining, 244 Free mineral acids in leather, 367 Fresenius, 251 and Griinhut, 83 INDEX. 453 GADD, 301 Gallic acid, 119, 121 Gallipoli, 326 Gall nuts, 110 Gallotannic acid, 121, 130 Gambier, 142 sampler, 178 Gantter, 172 Gautier, 138, 139 Gawalowski, 311 Gelatine culture, 443 Gerland, 170 Gladding, 278 Gladstone and Dale, 298, 344 Glucose and dextrine in leather, 360 chrome liquor, 257 copper tables, 272, 273 detection and estimation of, 266 determination electrolytically, 274 filter-tube, 271 in German leathers, 360 volumetric determination of, 267 - Von Schroeder's gravimetric deter- mination of, 269 Glynsky, 10 Goppelsroeder, 384 Graebe, 122 Graefe, 293 Grain, 418 Gravimetric determination of glucose, Von Schroeder's, 269 Green, 375, 382 Griessmeyer, 122 Grinding of tanning materials, 179 Grosset, see Beilstein Griinhut, see Fresenius Giinther, 133 Guthrie, A., 54 HACKFORD, see Trotman Haematin, 382 Hsematoxylin, 141 Hair bulbs, 414 development, 412 roots, 418 structure, 415 Hallopeau, 89 Hammer, 224 see Miintz, 173 Hanus, 301, 306, 308 Hardness, permanent, 32 temporary, 31 Harrison, see Lamb Hart, 96 Heal, 260, 261 Hehner, 32, 34, 254, 301 value, 314 Hellon, 274, 41 1 Hen-dung, 92 Hide, specific gravity of, 99 Hides, lime in, 101 and skins, acid in, 102 control of, 98 Hlasiwetz, 133 Holmes, see Procter Hoppenstedt, 245 Horse-fat, 323 Hiibl, 306 Hughes, 153 Hunt, 131, 140, 225,253 Hydrates, carbonates, and bicarbo- nates, 28 Hydrochloric acid, 24 normal, 20 specific gravity, 24 Hydrometers, 8 IGNITION, 11 of precipitates, 13 Indicators, 16 Indigo, 382 - pure B.A. S. F., 227 Inoculation, 440 International Association of Leather Trades Chemists, official method, 212 commission on tanning analysis, 211 Insoluble matter in extracts, 181 Iodine solution, 306 value, 209, 303 determination of, 306 table of, 309 lodometric determination of iron, 251 Iron, colorimetric determination of traces of, 40 determination, permanganate method of, 251 iodometric determination of, 251 Isocholesterol, 313 Iso-vanillic acid, 129 454 LEATHER INDUSTRIES LABORATORY BOOK. JAPAN wax, 346, 348 Jean, 155, 172, 341, 353, 367 Jean's method of tannin determina- tion, 230 Johansen, 145 Jolles, 89 Jones, A. S., 182 Wood, J. T., 92 KAMMFETT, 323 Karnowski and Tambor, 139 Kathreiner, 62, 171, 187, 225, 228, 424 Kauschke > 376 Kjeldahl method, 64 Koch, 182, 200, 269 Koch's apparatus, clamp for, 184 method of determining free acids in liquors, 243 tanning extraction apparatus, 1 83 Koerner, 142, 143 Kohnstein, see Sim and, 244, 440 and Simand's method of determin- ing tota free acid in liquors, 244 Koninck, see Nihoul Kopecky, 199, 204, 205, 265 Kopecky's chromed hide powder, 205 Kostanecki, 140, 141 and Tambor, 139 Kottstorfer, 304 Kubel-Tiemann, 42 LACMOID, 18 Lactic acid, 25 anhydride, 26 Lactones, 317 Lamb, 371 and Harrison, 407 Lambert, A., 106 Lanoline, 327 Leather, acids in, 366 analysis, 355 ash analysis, 363 determination of chromium and alumina in, 263 determination of fats in, 357 of free mineral acids in, 367 of moisture in, 357 of water-soluble matter in, 358 Leather, glucose and dextrine in, 360 mineral ash, 356 preparation for analysis, 355 total ash in, 362 uncombined tannin in, 359 Lebbin, 80 Le Bel Henniger, 10 Lecithin, 353 Lemme, 83 Lepetit, Dollfus and Gansser, 127 Lewkovvitsch, 311, 313, 314, 318, 325, 326, 328, 335, 347 Lime in hides, 101 liquors, dissolved organic matter in, 88 water reagent for tannins, 153 Limes, bacteria in, 90 and liming, 85 solubility of, 53 Lipochromes, 312, 313, 316 Liquid media, 442 Liquors, microscopic examination of, 246 valuation of, 247 volatile acids in, 246 Litmus, 1 8 Lorenz, 298 Lovibond's tintometer, 236 Lowe, 121, 130, 131, 145 Lowenthal, 139, 140, 149, 172, 242 LowenthaPs determination, calcula- tions of, 229 method of tannin analysis, 223 Liebermann-Storch, 316, 347 McCANDLlSH, see Procter Mackay, 333 Maclurin, 142 M'Culloch, 260 Magenta, 379 Magnesia, 365 determination of, 36 Magnetic iron in sumac, 201 Maljean, see Balland Manganese, 365 Marking vessels, '4 Martin, see Brown Dennis and Co., 84 INDEX. 455 Maschke, 315 Maumene', see Thomson and Ballan- tyne, 299 Mann, 305 Measuring, 3 instruments, 3 microscopic objects, 405 Meerkatz, 173, 197 Melting point of fats, 291 of waxes, 345 Metellagic acid, 124 Methyl orange, 16 Metric system, 4 Meunier, 368, 370, see also Vignon Micrococcus, 432 Micro-photography, 406 Microscopic examination of bacteria, 447 of liquors, 246 focusing, 398 objectives, 392 objects, mounting, 399 structure of skin, 411 Miller, 106 Mineral acids in oils and fats, 289 oils, 344 - waxes, 345 Mittenzwey, 171 Mixed and doubtful tannins, reactions of, tables, 161 Moeslinger drying oven, 193 Mohrs float, 3 - liter, 5 Moellon, 338 Monier, 172 Montan wax, 293 Moringatannic acid, 142 Morus tinctoria, 141 Mucor, fructification of, 439 M tiller, 46, 169 Munson, see Tolman Miintz, 173 Muscular fibre, 402 Myrica nagi, 141 Myricetin, 141, 142 NATURAL colouring matters, 140 sugars present in tanning materials, 276 Natural tannins, 142 Nauclea gambir, 138, 148, 154 Neatbfoot oil, 324, 325 Neubauer, 224 Neutral oils in carbolic acid, 78 New " International " method of tan- ning analysis, 21 1 Nierenstein, 80, 107, in, 116, 117,. 120, 130, 156 Nihoul, 50, 61, 65 and Koninck, 371 Nitro-colours, 377 Nitrogen determination by Kjeldahl,. 64 Nitroso-colours, 378 Nitrous acid as reagent for tannin, 151 Non-tannins, 2' 2 by hide-powder, absorption of, 207 tables, 208, 210 Normal hydrochloric acid, 20 potassium hydrate, 2 1 sodium carbonate, 2 1 solution, 19 hydrate, 2 1 sulphuric acid, 22 OAK twig section, 404 Objectives, microscopic, 392 Oil immersion, 396 Oils and fats, 281 ash in, 288 mineral acid in, 289 specific gravity of, 289 water in, 288 oxidation of, 299 Oleic acid, 321 Olive oil, 326 Organic impurities, 41 Orpiment or yellow arsenic sulphide, 57 Oser, 227 Osmic acid, 424 Osseine, 200 Osyris compressa, 142 Oxalic acid, crystallised, 24 Ox-hair root, 415 Oxidation of oils, 299 Oxidised fatty acids, 314 Ozokerite, 346 45 6 LEATHER INDUSTRIES LABORATORY BOOK. PAESSLER'S chromed hide-powder, 198 and Appelius, 50 and Sluyter, 368, 370 and Spanjer, 243 Palmer, 426, see also Richards Palmitic acid, 321 Panniculus adiposus, 418 Papilla, 418 Paraffin, 313 - wax, 345 Para-rosaniline, 379 Parker, 148, 149, 170, 177 and Payne, 175, 187, 242 and Procter, 76, 84, 182 and Russell, 99 method of tannin determination, 232 Pars'papillaris, 418 Pasteur, 94, 442 Pasteur's solution, 442 Patten, see Hart, 96 Payne, E. M., 83, 96, 101. 175, 187 and Parker, 148, 149, 170 and Pullman, 79, 90 Payne- Pullman, 86 Pelouze, 1 10 Penicillium glaucum, 438 Perkin, A. G., 122, 123, 124, 125, 126, 133, 136, 140 and F. M. Perkin, 114 and Steven, 114 and Yoshitake, 137, 138, 139 Permanent ether value, 317 hardness, 32 saponification value, 317 Permanganate method of iron deter- mination, 25 i Pleifer, and Wartha, 33, 34 Phenolphthalein, 17, 381 Phlobaphenes, 126 Phloroglucol, 115, 116 distinction from catechol and pyro- gallol, 118 Phloroglucolcarboxylic acid, 134 Phycomycetes, 439 Phytosterol, 313 Pigeon dung, 92 Pipettes, 3 use of, 4 Pistacia lentiscus, 1 54, 409, 508 Plate culture, 445 Polariscope, 402 Popp and Becker, 93 Porous alumina, 200 Posternak, 96 Potash alum, 254 Potassium bichromate, 255, 257 hydrate, normal, 2 1 Precipitates, drying, 13 ignition of, 13 washing, 12 Precipitation, 12 Preparation of hide-powder, 198 of infusion of tanning materials, 212 Priestman, A., 177 Priestman, H., 408 Procter, 76, 84, 182, 207,224, 225,226, 243, 298 and Bennett, 234 and Blockey, 207 and Holmes, 299, 335 and McCandlish, 60, 265 and Searle, 368, 369 Procter's method of determining free acids in liquors, 243 tannin extraction apparatus, 185 Protocatechuic acid, 129 Puers, 92 Pullman, see Payne Pure cultures, 446 Pycnometer, 6 Pyrogallol, 113 distinction from phloroglucol and catechol, 118 tannin reactions, tables, 162 tannins, 108 Pyrogallolcarboxylic acid, 134 QUANTITY of tanning materials used for analysis, 191 Quebracho, 142 Quercetin, 141, 142 Ouercitannic acid, 144 Quercitron, 142 ' Quercus prinus, 154 pubescens, 145 INDEX. 457 Quercus robter, 112, 148 tinctoria, 142 Quinol, 105, 118 RAMSPACHER, 173 Raschig, 46, 47 Rawson, 227 Reagent for basic dyes, 375 Realgar, 57 Real's tanning material extractor, 190 Reaumur's thermometer, 6 Recovered wool-fat, 327 Recovery of chrome from chrome liquors, 265 Red arsenic, 57 Reds, solubility of, 181 Reed's combined evaporator and dryer, 193 Refractive constants, table, 297 index, 294 Refractometer, 295 Reichert-Meissel value, 314 Resin acids, 313 oils, 343 Resins, determination of, 316 Resorcinol, 105, 117 Resorcylic acids, 129 Rhus coriaria, 407, 408 cotinus, 141 metopinm, 409 Richards and Palmer, 170 Richter, 117 " Roll culture," 445 Romijn, 81 Rosin, 316 - oils, 343 Rosolic acid, 380 Rufigallic acid, 120 Ruhsam, 269,1318, 321 Rusma, 57 Russell, G. H., 99 SACCHAROMYCETES, 437 Sadlon, 127 Sampling, egg-yolk, 351 of tanning materials, 177 Sand in sumach, 201 Saponihcation value, 302 determination of, 304 Schall, see Appelius Scheuch, 54 Scheurer, see Vaubel Schiff, in, 122, 123, 130, 131, 132,. 134, 137 Schift's reaction, 80 Schinia, 409 Schloessing's apparatus, 62 Schmidt, 318 Schmitz-Dumont, 267, 286, 328, 343,. 344 Schroder, see Earth Seal oil, 337 Searle, see Procter Sebaceous gland, 416 Section of calf skin, 412 cutting, 402 microscopic, 42 1 Sequoia gigantea, 143 " Shaking machine," 202 Shark liver oil, 310 Shukoff, 292, 294, 345 Siedentopf, 431 Simand, 224, 225, 244, 245, 267, 315,. 440 and Weiss, 173, 174 Sluyter, see Paessler Soaks and soaking, 85 Soaps, alkali in, 278 analysis of, 277 fatty matter in, 278 water in, 277 Sodium carbonate, normal, 21 solution, normal, 19 chloride solutions, specific gravity of, 71 hydrate, normal, 21 sulphide, 55 sulphite reagent for tannin, 152 thiosulphate solution, 306 Sod oil, 337 Softening water, 37 Solidifying point of oils and fats, 29 1 " Soluble non-tanning matter," filter bell, 195 Solubility of limes, 53 of reds, 181 458 LEATHER INDUSTRIES LABORATORY BOOK. Soxhlet tube, 285 Soxhlet's tanning material extractor, 190 Spanjer, see Paessler Specific gravity, 6 of de'gras, 339 of hide, 99 of hydrochloric acid, 24 of oils and fats, 289 of sodium chloride solutions, 71 of sulphuric acid, 23 temperature reaction, 299 Spectroscope, 385 Spent tans, analysis of used liquors and, 215, 241 tanning materials, 248 Spermaceti, 310 Sperm oil, 310 Spiegel, 75 Spren^el tube, 7, 290 Standard solutions, 14, 18 Stannous chloride and hydrochloric acid as reagent for tannin, 152 Statement of results in water analysis, 48 .Stearic acid, 321 Stenhouse, 107 Sterilisation, 441 Steven, see Perkin Stiasny, 57, 87, 156, 368, 371, 372 " Stinco," 409 Stoddart, 168 Stone, see Scheuch, 54 Strength of tanning solution for ana- lysis, 191, 211 .Sudoriferous glands, 414 Sugars present in tanning materials, natural, 276 Sulphide of arsenic, analysis of, 57 of sodium, 55 Sulphides, alkaline, 87 Sulphites, 73 Sulphur, determination of, 58 Sulphuric acid, 22 determination of, 46 normal, 22 reagent for tannin, 153 specific gravity, 23 Sulphurous acid, 73 and sulphites, estimation of, 27 Sumach and its adulterants, effects of heating, table, 408 magnetic iron in, 201 sand in, 201 Sutton, 253, 254 Sweat glands, 418 , TABLE, analysis of waters, 52 constants of fats, 350 of fatty acids, 350 of liquid oils. 349 effects of heating sumach and its adulterants, 408 refractive constants, 297 Tables, qualitative recognition of tan- ning material^, 157 et seq. Tallow, 321 Tamarisk, 410 Tamarix africana, 155,410 Tambor, see Karnowski, and also Kostanecki Tannage, valuation of, 247 Tannin absorption, filter paper cor- rection for, 1 88 analysis, American Leather Chem- ists' official method, 216 calculation of results of, 200 general regulations, 2 1 1 Lb'wenthal method of, 223 determination, Jean's method, 230 - Parker -Payne method, 232 extraction apparatus, Koch's, 183 - Procter's, 185 Real's, 190 Soxhlet, 190 - Von Schroeder's, 190 solution for analysis, strength of, 191, 2ii solutions, filter paper for, ib6 test, 239 Tanning materials, ash in, 201 classification tables, 157 determination of total dry matter, 192 extraction of, 180 grinding of, 179 natural sugars present in, 276 sampling of, 177 INDEX. 459 Tanning materials spent, 248 total soluble matter in, 194 used for analysis, quantity of, 191 Tannins, decomposition by dilute acids of, 118 by heat, 112 preparation of, 1 10 purification of, 1 10 Temporary hardness, 31 Terreil, 171 Testing of commercial materials, 22 Thermometer, Centigrade, 6 Fahrenheit, 6 Reaumur, 6 scales, 6 Thomson and Ballantyne, 299, 336, 337 Thorsen, see Boegh Titanium, 364 " Titer test," 294 Titration of alkaline salts, 26 Tolman, and Munson, 298 Total ash in leather, 362 - free acids in liquors, Kohnstein and Simand's method of, 244 solids, 212 soluble matter in tanning materials, 194 solubles, 212 Trillat, 80 Trimble, 108, in, 113, 144, 145, 146 Trimble's oak bark reactions, tables, 163 Triphenylmethane, 379 Trotman, see Wood and Hackford, 169 " Tube " culture, 444 Turin conference, 185, 186, 187 Turnbull, 201 Twaddell degrees, 8 Twitchell, 347 TwitchelPs method, 316 ULZER, see Benedikt " Uncombined tannin " in leather, 359 "Unorganised ferments," 428 Unsaponifiable matter, determination of, 309 Use of the microscope, 391 Used liquors and spent tans, analysis of, 215, 241 VACUUM drying oven, 194 Valuation of liquors, 247 of tannage, 247 Vanillic acid, 129 Vaubel and Scheurer, 171 Vegetable waxes, 346 Vienna Research Station, 204 Vignon and Meunier, 353 Vine leaves, 410 Virginia-creeper, 117 Viscous fermentation, 246 Volatile acids in liquors, 246 Volumetric analysis, 14 determination of glucose, 267 Von Schroeder, 181, 198, 225, 226, 267, 277, 360 gravimetric determination of glucose, 269 tanning material extractor, 190 Vournasos, 132 WARTHA, see Pfeifer Washing precipitates, 12 Water analysis, complete gravimetric, 43 statement of results, 48 in degras, 338 - in egg-yolk, 351 - in oils and fats, 288 in soaps, 277 soluble matter in leather, 358 softening, 37 Waters, table, analysis of, 52 Waxes, melting point of, 345 Weighing, I Weingartner, 375 Weiss, 197, 207, 210 ; see also Simand Werner, see Berthollet Whale oil, 336 Wijs, 301, 306, 308 Will and Albrecht, 134 Winter-Ely th, 274 Wislicenus, 175, 200 Wood, 93, 94, 96, 301, 429, 432, 435, 440, 448 460 LEATHER INDUSTRIES LABORATORY ROOK. Wood and Trotman, 90 Wool-fat, 310 Wuensch, 368, 37 1 YEAST, 437 ' Yellow arsenic sulphide, orpiment or, 57 Yocum, 175, 202 " Yorkshire grease,' 327 Yoshitake, see Perkin Z6LFEL, 143 Zwenger, 122 Zymases, 428 LONDON = PRINTED BV WILLIAM CLOWES AND SONS, LIMl'TED, CHEAT WINDMILL STREET, W., AND DUKE STREET, STAMFORD STREET, S.E. THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $1.OO ON THE SEVENTH DAY OVERDUE. 25 FEB LD 21-100m-7,'39(402s) 03867 333813 UNIVERSITY OF CALIFORNIA LIBRARY