. . p * *- * + I * * * + ^ " ^ INKS THEIR COMPOSITION AND MANUFACTURE INCLUDING METHODS OF EXAMINATION AND A FULL LIST OF ENGLISH PATENTS BY C. AINSWORTH MITCHELL B.A.(Oxon.), #I.C. .AND " T. C. HEP WORTH v WITH 46 ILLUSTRATIONS, INCLUDING 4 PLATES LONDON CHARLES GRIFFIN & COMPANY, LTD, EXETER STREET, STRAND 1904 [All rights reserved] Printed by BALLANTYNE, HANSON London &> Edinburgh PREFACE SOME three years ago we were engaged in a scientific inquiry as to the composition of certain fluids used as writing ink. As this work led us beyond the limits anticipated, and to the making of many experiments not actually required at the time, and as there is need for a volume dealing adequately with the subject, we thought it advisable to embody the results in book form. We found, it is true, a few small books on ink and many allusions to ink-making in old volumes-and isolated papers in scientific journals ; but it seemed to us that the matter required more comprehensive treatment, and the pre- sent work may be regarded as an attempt to supply that want. As far as time permitted we have tested the various formulae quoted, but, as may be seen by reference to the patent list at the end of the book, there are so many cases of slight variations in composition that we have often contented ourselves with a record of the statements put forward. We have pleasure in tendering our best thanks to those who have assisted us in our work. To Mr. R. M. Prideaux, in particular, we are indebted for the excellent drawings of the various galls (pp. 37-46), the details of which could not have been nearly so w*ll shown by photography. Messrs. Newman and Co., of Soho Square, were good enough to afford us m uch information with regard to sepia vi PREFACE preparations, and to supply us with dried specimens, &c., tor analysis. The photographs of fossil cephalopoda were taken by us at the Geological Museum, Jermyn Street, by the courtesy of the Curator. To the authorities at Kew we are indebted for permis- sion to photograph in the Museum and in the Herbarium. The Badische Company kindly supplied us with specimens of aniline dye-stuffs and much valuable information regarding them. We have also to thank Messrs. Keller and Co., who have kindly allowed us to use certain blocks illustrative of printing-ink machinery, and have sent us samples of various permanent colours. Lastly, our thanks are due to Messrs. Madderton and Co., of Loughton, for specimens of permanent pre- parations made by them. C. A.M. T. C. H. GRAY'S INN, LONDON W.C. August 1904. CONTENTS HISTORICAL INTRODUCTION Ancient Egypt Old papyri Progress of writing Herculaneum frag- ments Carbon inks Iron gall inks The Lindisfarne Gospels Transition from carbon to gall inks Domestic ink making Scientific experiments Unoxidised gall inks Aniline inks German regula- tions Other inks .... . . Pages 1-14 SECTION I WRITING INKS CHAPTER 1 CARBON AND CARBONACEOUS INKS Sepia Source Manufacture Chemical composition Sepiaic acid British sepia Examination of commercial sepia Indian or Chinese Ink Lamp-black Composition Manufacture of lamp, black Old European methods Manufacture of Indian ink Qualities of Indian ink Examination of Indian ink Practical tests Carbon Writing Ink Ancient carbon inks Modern carbonaceous inks Pages 15-35 CHAPTER II TANNIN MATERIALS FOR INKS Galls : Origin Aleppo galls Chemical composition Chinese galls Chemical composition Japanese galls Acorn galls Oak-apple galls Other galls Tannins Classification of tannins Suitability of tannins for ink-making Chestnut bark and wood Chestnut extract Chestnut tannin Ink from chestnut wood Sumach viii CONTENTS Sumach tannin Ink from sumach Divi-divi Divi-divi tannin Ink from divi-divi Myrobalans The tannin of myrobalans Valonia The tannin of valonia Ink from valonia Oak- bark tannins Keactions of oak tannins Amount of tannins in oak-bark Ink from oak-bark Gallotannic acid Fermentation of gallotan- nic acid Properties Reactions distinguishing between gallotannic and gallic acids Pages 36-71 CHAPTER III NATURE OF INKS Constitution of ink-forming substances Influence of light and air Iron tannates Evidence of an intermediate blue iron oxide Tannates of iron Basic salts Methods of estimating tannates Procter's method Jackson's lead carbonate method Ruoss's ferric sulphate method Colorimetric methods Hinsdale's colorimetric method Mitchell's colorimetric method . : . Pasres 72-8(5 CHAPTER IV MANUFACTURE OF IRON GALL INKS The relative proportion of galls and ferrous sulphate Deductions from the composition of ink deposits Old type of iron gall ink Old formulas of iron gall. inks Unoxidised iron gall inks Gallic acid inks Japan inks ...... Pages 87-08 CHAPTER V LOGWOOD, VANADIUM, AND ANILINE BLACK INKS Logwood inks Logwood Logwood extract Hasrnatoxylin Hasmatein Iso-hcematein Addition of logwood to gall inks Log- wood inks without tannin Chrome logwood inks Ha?matein inks Use of logwood in patent inks Vanadium inks Black aniline inks Pages 99-111 CHAPTER VI COLOURED WRITING INKS Historical Coloured aniline inks Fugitiveness of aniline inks Patented coloured inks Pages 112-118 CONTENTS ix CHAPTER VII EXAMINATION OF WRITING INKS Fluidity of ink Penetration through paper Stickiness of writing Composition of commercial inks Schluttig and Neumann's stripe test Acidity, action on steel pens Stability on keeping Examina- tion of handwriting Old manuscripts Palimpsests Forged handwriting Bleaching agents Differentiation of writing done with different inks Photographic methods Mechanical erasure Chemical removal of writing Destruction of sizing Alterations and additions to writing Photographic distinction between different inks . Pages 119-131 SECTION II PRINTING INKS CHAPTER YIII EARLY METHODS OF MANUFACTURE Historical China Greec and Rome England Early printed books Early methods of manufacture Fertel's method of making ink Breton's method Savage's method of manufacture Modern methods of preparing ink Pages 132-140 CHAPTER IX MANUFACTURE OF VARNISH Boiled oils Burnt oil Varieties of lithographic varnish Andres' apparatus for boiling oil Apparatus with steam-jacket and air-blast Boiling with superheated steam Treatment with oxygen Linseed- oil substitutes Pages 141-150 CHAPTER X PREPARATION AND INCORPORATION OF THE PIGMENT Black for printing ink Modern apparatus Thenius' lamp-black furnace Furnace for producing black from pitch Other black pigments Carbon blacks Purification of lamp-black Composition of lamp- blacksMethods of examining lamp-blacks and gas-blacks Mixing the black and varnish Mixing the varnish and lamp- CONTENTS black Quack's mixing machine Werner and Pfleiderer's mixing machine Lehtnann's mixing machine Grinding Lehmann's grinding machines Machines by Neil, Jackson, Kingdon Litho- graphic printing ink Collotype ink . . . Pages 151-166 CHAPTER XI COLOURED PRINTING INKS Early methods Manufactured inks Painters' pigments Early gnor- ance as to proper pigments Half-tone process block Necessity for cleanliness Overlays Coarse-grain screens Theory of colour Diagrams of colour Peculiarities of pigments Permanency of pigments Yellow pigments Red pigments Blue pigments Green pigments Purple and orange pigments Brown pigments " Art " shades Three-colour printing Photographic falsification of colour Coloured screens Clerk-Maxwell's work Colour screens or niters Coloured light Pure pigments unknown General considera- tions Examination of trichromatic prints The half-tone dot Necessity for transparent inks Opacity of yellow pigments Supple- mentary key block Inks for cheques and bank-notes Patent inks for cheques ..,...., Pages 167-185 SECTION III INKS FOR MISCELLANEOUS PURPOSES CHAPTER XII COPYING INKS Various copying inks Patent copying inks Copying papers Copying ink pencils Manifold copying apparatus . . Pages 186-191 CHAPTER XIII MARKING INKS The ink plant of New Granada The Indian marking nut The Cashew nut Rhus toxicodendron Rhus venenata Rhus radicans Other vegetable juices Chemical marking inks Silver inks Gold marking inks Platinum marking inks Marking inks containing other metals Aniline marking inks Indigotin marking inks Alizarine marking ink Examination of marking ink Marking-ink pencils Pages 192-206 CONTENTS xi CHAPTER XIV SAFETY INKS AND PAPERS Various safety inks Resinous inks Traill's carbon gluten ink Soluble glass ink Other carbon inks Safety papers with special inks Patent permanent inks Patent safety papers Pages 207-213 CHAPTER XV SYMPATHETIC INKS History Various sympathetic inks Patent sympathetic inks Pages 214-218 CHAPTER XVI INKS FOR SPECIAL PURPOSES Ink powders and tablets Logwood ink powders Aniline ink powders Patent ink powders and dried inks Stencil inks Show-card ink Inks for rubber stamps Inks for writing on glass Hydro- fluoric inks Resin inks Foertsch's pencil for glass Inks for writing on metals Ink for writing on leather Ink for ivory surfaces Ink for writing on wood Fireproof inks Pages 219-22C. LIST OF ENGLISH PATENTS . . . . Pages 227-242 INDEX . . . . . Pages 243-251 LIST OF ILLUSTRATIONS Frontispiece Elizabethan housewife's recipe FIG. PAGE 1 Egyptian palette, brushes and pens 6 2 Egyptian slab and muller . . . . . . .6 3 Egyptian wax tablet 6 4 Common cuttlefish (Sepia officinalis) . ... 16 5 Fossil sepia 17 6 Fossil sepia 18 7 Dried sepia sacs 21 8 Chinese manufacture of lamp-black . 24 9 Chinese manufacture of ink ... 25 10 European lamp-black chamber 27 Brush drawing in Chinese ink by Japanese artist (Plate I.) To face 30 11 English double oak-apple gall . .... 37 12 English oak gall 37 13 English gall 37 14 Green Aleppo gall 38 15 White Aleppo gall 38 16 Section of white gall '38 17 Chinese gall 41 18 Japanese gall ... 44: 19 Aphis from Chinese gall 45 20 Aphis from Japanese gall 45 21 Oak-apple gall 46 22 Gall wasp (Cynips Kollari] 47 23 Sumach (Coriaria myrtifolia) ...... 55 24 Divi-divi pods 58 25 Myrobalans 59 26 Valonia 61 27 Trimble's apparatus for tannin determination ... 81 xiv LIST OF ILLUSTRATIONS FIG PAGE 28 Hehner's Nesslerising tubes . . . . . '. 8(5 29 Action of bleaching reagents on writing (Plate II.) To face 129 30 Free-fired pan for boiling oil 143 31 Andres' apparatus Ho 32 Steam-heated kettles U 33 Apparatus for hot-air treatment of oils ... 147 34 Lamp-black apparatus .152 35 Lamp-black apparatus .153 36 Quack's mixing machine . . V lfi<) 37 Lehmann's mixing machine 1(H 38 Lehmann's grinding mill . . . . . . Ifi'J 39 Colour diagram . . .171 40 Solar spectrum in triplicate (Plate III.) . . . To face 17S 41 Clerk-Maxwell's colour curves 180 42 Indian marking ink (Sf-nief&rpUQ anacareliuin) . . . 193 43 Cashew nut (Anaeardin m Occident tile] ..... 195 44 Rhiis toxicodendron . ' . 197 HISTOEICAL INTRODUCTION. CONTENTS. Ancient Egypt Old papyri Progress of writing Herculaneum fragments Carbon inks Iron gall inks The Lindisfarne Gospels Transition from carbon to gall inks Domestic ink-making Scientific experiments Unoxidised gall inks Aniline inks German regulations Other inks. Ancient Egypt. The earliest use of a liquid which can be described as "ink" is found in those documents on papyrus which have been among the archaeological treasures of Egypt. Although the history of Egypt has been traced back for a period of more than four thousand years, and papyrus was employed as a writing material there from very remote times, the oldest specimen of the material extant is a roll which dates from B.C. 2500.* This possibly refers to the oldest specimen which bears decipher- able characters, for Professor Flinders Petrie has found fragments of papyri which date from a thousand years earlier.f As Egypt is still the subject of exploration, and as perishable articles have been found of a still earlier period than that last mentioned, we may reasonably hope that ink- written records may some day come to light which will carry back the history of the country to a more remote time. Professor Flinders Petrie found in one tomb, dating* from 3500 B.C., baskets, a coil of palm rope, wooden mallets, and chisels left behind by the workmen, together with some pieces of papyrus which were almost white ; and he attributes the excellent condition of these things to the preservative nature of the clean dry sand in which they had been buried for so many centuries.} * British Museum- Guide, 1896, p. 312. t Journal of the Camera Club, Nov. 1897. I Ibid, MANUFACTURE Old Papyri. In Case A (Greek Papyri), British Museum, can be seen a number of specimens dating from the first century of the Christian era, and although in many cases * V the papyrus is merely in fragments, the ink is as black as it was the day that it was applied. The lettering in many of these papyri is extremely beautiful, and compares very favourably with much of the handwriting that some of us have to decipher to-day. And it would seem quite clear from an examination of many of these writings that the implement employed was a pen and not a brush. The papyrus in some instances is of a very light drab colour, and on this surface the old writing stands out with startling distinctness ; but when the material has assumed a dark brown or yellow tint, the writing is not so distinct, although the quality of the ink is quite as good. That papyrus was not a cheap material is- shown by a specimen here, labelled " Aristotle on the Constitution of Athens. The only extant MS. of the work, brought from Egypt in 1890. Written about A.D. 100, ia four rolls, in four different hands, on the back of the papyrus which had already been u?ed [in A.D. 78-79] for the acconipts of a farm-bailiff named Didymus, near Hermopolis." Another specimen of great interest lies close to the one first mentioned, namely, fragments of the Theogonia of Hesiod. It is written in a firm and large hand in very black ink, and the label tells us that its date is probably the fourth or fifth century, " contemporary with the early MSS. on vellum, and so marking the transition from the one material to the other." Progress of Writing. The various specimens shown in the King's Library at the British Museum, in Cases A E, are designed to illustrate the progress of writing from the second century B.C. to the fifteenth century of our era, and at the same time they afford testimony as to the kind of ink employed during the period covered. The basis of the black ink used on papyrus by the ancient scribes was undoubtedly carbon, a substance which had the advantage of being easily procurable, while at the same time it was indestructible except by fire. It was probably prepared in the form of vegetable or animal charcoal, and was mixed with gum, oil, or varnish. Possibly, for the finer writing, HISTORICAL INTRODUCTION 3 water, with gum or glue as a binding material, was the medium mostly employed, for it would flow more readily from the reed pen or quill used by the writer. It is certain that the art of writing has a remote antiquity, and that the power of recording thoughts in this way marks a distinct line of demarcation between civilised man and the savage. J t is a matter of interest to consider the many different materials which have been used for writing upon in early times besides papyrus. Soft wood cut into slices and planed and polished was used in various countries, the pen being a metal stylus, which simply scratched or indented the material. ~l7ater on, the wood tablet with a thin coatingof wax was employed, and the writing upon it in the case of ephemeral memoranda could be quickly effaced. Bark and palm leaves were also used for writings of a temporary character, and in some countries in later times both linen and silk have been so employed. The Chinese are credited with the invention of paper anterior to the Christian era, a statement which need not excite surprise when we remember that they anticipated Europe in the invention of printing by nearly a thousand years. We may assume that for many centuries before this the art of writing in China had been brought to some degree of perfection. Among the Roman antiquities found in Britain, which are now deposited at the British Museum, are many speci- mens of the stylus in ivory, bronze, &c., and some of these are armed with a sharp projection, with which guiding lines could be ruled across the waxen surface of the tablets, The reed pen was commonly used for writing on papyrus, and the steel pen was foreshadowed by a few specimens in bronze found in Italy, and one in England. This last is among the Romano-British antiquities in the British Museum. It consists of a tubular piece of bronze, about five inches in length, which has at one end a split nib, while the tube is gradually reduced in size towards the other extremity, where it ends in a solid piece, which was probably used for pressing down the wax in order to efface the writing. In the Mediaeval Room at the British Museum may be found many specimens of writing tablets, some dating 4 INKS AND THEIR MANUFACTURE before the seventh century, of great rarity and therefore great value. Some of these are what is known as "con- sular diptychs," so-called because these folded tablets were at one time sent as ceremonial presents by the Roman consuls on their appointment to official persons or to friends. Many of these tablets are of ivory and are beautifully carved, and the slabs or plaques are sometimes of such a size that the tusks procurable at the time they were made must have been of unusual dimensions, or the artificers had some means of bending the material, the secret of which is now lost. Herculaneum Fragments. In 1821 Sir Hiaiqrfiry Davy read a paper before the Royal Institution,* in which he described a number of experiments that he made with some fragments of papyri which had been found in the ruins of Herculaneum. Of some of these papyri he says : "The black ones, which easily unroll, probably remained in a moist state without any percolation of water ; and the dense ones, containing earthy matter, had probably been acted upon by warm water, which not only carried into the folds earthy matter suspended in it, but likewise dissolved the starch and gluten used in preparing the papyrus and glue of the ink, and distributed them through the sub- stance of the MSS." He made further experiments in the Museum at Naples. And of some of the MSS. he says : " These MSS. had been so penetrated by water that there were only a few folds which contained words, and the letters were generally erased, and the charcoal which had composed them was deposited in the folds of the MSS." He makes some general observations, of which the fol- lowing is worth noting : " I looked in vain amongst the MSS. and on the animal (sic) charcoal surrounding them for vestiges of letters in oxide of iron ; and it would seeni from these circumstances, as well as from the omission of any mention of such a sub- stance by Pliny, that the Romans, up to his period, never used the ink of galls and iron for writing : and it is very probable that the adoption of this ink, and the use of parchment, took place at the same time. For the ink, * Trans. Roy. Soc., 1821, ii. 191. HISTORICAL INTRODUCTION 5 composed of charcoal and solution of glue, can scarcely be made to adhere to skin r ; whereas the free acid of the chemical ink partly dissolves the gelatine of the MSS., and the whole substance adheres as a mordant ; and in some old parchments, the ink of which must have contained much free acid, the letters have, as it were, eaten through the skin, the effect being always most violent on the side of the parchment containing no animal oil.''* The disintegration of the papyrus by the action of water, alluded to by Sir Humphry Davy,^ will be readily understood when we remember that this ancient writing material was made of thin strips cut from the reed and cemented together. The strips were laid side by side, and then other strips were laid across them at right angles, the whole being stuck together and placed under pressure so as to form a paper-like sheet. Papyrus was first used as a single sheet, or in lengthy documents as a long roll of different sheets joined together. Later on papyrus leaves were bound together as in a book. At a very early period papyrus was imported into Greece and Italy. It continued to be the chief writing material in Egypt until the tenth century, and was largely used in Europe after vellum had been introduced. Carbon Inks. We give illustrations (Figs. I, 2, and 3) of various writing implements dating back to about 1500 B.C., which are exhibited in the Egyptian department of the British Museum. The titles of these pictures sufficiently explain their nature. Chinese, or Indian ink, as it is commonly called in this country, was made at a very early period, according to Chinese historians as far back as between B.C. 2697 and 2597, the inventor being one Ticn-Trheu. Full particulars of the way it is manufactured are given in a subsequent chapter. Its base, like that of early Egyptian and other inks, is carbon. ' :< Tt-fftix. Hoy. iii. CentralM., 1888, xix. 587. 22 INKS AND THEIR MANUFACTURE British Sepia. Through the kindness of Messrs. New- man we have been enabled to examine several ink-sacs of cuttle-fish from Southampton, in the dried condition as received by them. The appearance of these is shown in the accompanying figure (Fig. 7). The general physical characteristics were very similar to those recorded by Prout (supra), but our specimens had the distinct fishy odour observed by Kemp in the case of the fresh liquid, and this became very marked on boiling the powdered substance with water. The powder contained 17.56 per cent, of moisture, and on ignition over a low Argand flame yielded 12.22 per cent, of ash, containing the following constituents : Silica, 0.28 ; calcium, 1.92 ; magnesium, 1.75 ; chlorine (including other halogens), 1.07; sulphuric acid, 1.84; total nitrogen, 8.42 per cent. When treated with boiling water the powder dissolved to a considerable extent, but repeated and tedious extrac- tion was necessary to remove the whole of the soluble matter. The black residue left on the filter amounted to 71.1 per cent, of the original substance, and contained 7.46 per cent, of nitrogen, calculated on the original powder. The brown solution when evaporated left a brown resin- like deposit, which on ignition gave 4.55 per cent, of ash, calculated on the original substance. On treating the insoluble residue with boiling 10 per cent, potassium hydroxide, 19.23 per cent, (calculated on the original substance) remained undissolved. Examination of commercial Sepia. Most, if not all, of the English manufacturers prepare sepia, exclusively from the cuttle-fish, but there is reason to believe that a large proportion of the so-called " sepia " of foreign origin is sepia in name only. A chemical means of distinguishing between genuine sepia and preparations consisting of lamp-black or other forms of carbon incorporated with glue, consists of treat- ing the powdered sample with boiling water untij tho- roughly disintegrated, filtering the liquid, and thoroughly washing the residue. In \}\Q case of sepia, this residue will contain a large CARBON AND CARBONACEOUS INKS 23 amount of nitrogen, and on ignition will leave a consider- able proportion of ash containing the constituents men- tioned in the previous section (p. 22). Lamp-black preparations, on the other hand, will leave a residue of practically pure carbon, containing only traces of nitrogen, and leaving but little ash on ignition. The whole of the glue, which would cause the finished prepara- tion to show a large proportion of nitrogen, will have been removed by the treatment with hot water and filtration. The main points to be considered in a manufactured sepia are the colouring power and permanency of the colour. Sepia was one of the pigments tried in the experiments of Dr. Russell and Sir William Abney (chap, vi.), and to quote the words of the latter : * " We are apt to look on sepia as one of the most permanent pigments ; as a matter of fact it is fugitive, and those who have examined sepia drawings made in the early part of the century will see there has been certainly a distinct fading in those drawing?." 2. INDIAN OR CHINESE INK. The extreme antiquity of the ink manufactured by the Chinese has already been mentioned in the Historical Intro- duction. According to ancient Chinese documents cited by Jametel,^ the earliest ink was a kind of vegetable varnish, and it was not till about the third century B.C. that the solid product prepared from lamp-black and glue was introduced. The province of Kiang-si enjoyed a monopoly of the manufacture, and the ink attained a high degree of perfection, its quality being maintained by special ink inspectors. This ink has also been prepared in Japan for many hun- dred years. The province of Om i produced a fine quality known as takesa, but the taikeiluku of Yainashiro was con- sidered the best. At the present day the best quality of Japarese ink is said to be manufactured in Nara or Matsuda. * Journ. Soc. Arts, 1889, xxxvii. 113. f L 1 Encre de Chine, d'npres des Documents Chinois, traduits par iVf. Jametel, Paris, 1 882. 24 INKS AND THEIR MANUFACTURE Lamp-black: Composition. Wben carbonaceous pro- ducts, such as oil, rosin, or tar, are burned with an insuffi- cient supply of air, the oxygen combines with the hydrogen forming water, whilst the carbon is to a large extent deposited in the amorphous form known as lamp-black. Fig. 8. Chinese manufacture of lamp-black. The amount of pure carbon in this soot is about 80 per cent., the remainder consisting of oily and resinous sub- stances with inorganic salts, notably ammonium sulphate. For ordinary commercial uses these impurities are not altogether disadvantageous ; but if a purer substance is required, the lamp-black is heated to redness in a closed CARBON AND CARBONACEOUS INKS 25 crucible to carbonise the organic substances, and then digested with hydrochloric acid and thoroughly washed with water to remove inorganic salts, the final product being nearly pure carbon. The purest form of lamp-black is obtained by passing a slow current of turpentine vapour Fig. 9. Chinese manufacture of ink. through tub3S heated to redness, and igaitiug the deposit in chlorine to remove the last traces of hyJrogan. Manufacture of Lx>ny-Ua?k: Chinee M^hrl. Th3 old3sb msthoi of which we hive any ra^ord is that which has b3en usel by the Chinese for csnturies.* Various * Jametel, loe. cit. p. n. 26 INKS AND THEIR MANUFACTURE substances have been used as the original source of their lamp-black, such as rice straw, pine wood, and haricot beans, but these have been for the most part discarded in favour of vegetable oils, and in particular that obtained from the seeds of Aleurites cordata, or tung-cil, which yields a brilliant black ink, deepening in tone with age. The oil is burned in small terra-cotta lamps, which are placed in terra-cotta chambers with a hole to admit air, and having a depression on the top in which water is placed. The smoke is collected in inverted terra-cotta cones with polished interiors, which are fixed above the flame. From time to time the cones are replaced by fresh ones, and the deposited soot removed by means of a feather, care being taken tc reject all oily particles. Figs. 8 and 9 are reproductions of two of the quaint illustrations copied by Jametel from ancient Chinese manuscripts. In some factories the terra-cotta condensing chamber is replaced by a hollow wooden ttmnel, having a hole bored in the wall to act as the ventilating shaft. A range of bricks inside supports the cones, of which about twenty are used at a time. The best season for the manufacture of lamp-black is at the end of autumn or beginning of winter. The terra- cotta condensers are placed in a room carefully protected from draughts, which would interfere with the regular deposition of the soot. The cones are examined hour by hour, since delay in changing them causes the lamp-black to assume a yellow tint. Julien* states on the authority of Chinese documents that the finest quality of ink is prepared from the lamp- black obtained from sesame oil, or from tung oil, whilst the soot of pine wood or deal is used for the commoner kinds. Strips of pine wood about 18 inches in length are burnt in a bamboj cabin, 100 feet in length, which is covered inside and out with paper, and divided into several com- partments by partitions, in each of which is an opening for the passage of the smoke. The deposit in the furthest compartment is the lightest and makes the best ink, whilst * Ann. cle Chun.. 1833, ^"- 38- CAEBON AND CARBONACEOUS INKS 27 that in the first and second compartments is very coarse r and is sold to printers, varnishers, and house painters. The quality of the lamp-black has a very great influence upon the character of the ink, and the Imperial ink is prepared from the very lightest and purest that can be obtained. Old European Methods. Lamp-black is manufactured on Fig. 10. European lamp-black chamber. a large scale from the resinous- imparities obtained as by- products in the manufacture of turpentine, and is also prepared from oil, tat, &c. The initial substance is burned in a furnace with an insufficient supply of a^r suitably regulated by apertures which can be opened or closed. The dense smoke is conducted through a flue into a cylindrical stone, brick, or cast-iron chamber, the sides of yhich are covered with sacking or sheep-skin. An iron cone is 28 INKS AND THEIR MANUFACTURE suspended within the chamber, which it fits so exactly that when lowered its edges scrape the suspended sacking and remove the deposited lamp-black (see Fig. 10). A small hole in the top of the cone allows the smoke to escape into the chimney of the cylinder, leaving most of its carbon behind. From time to time the suspended sacking is removed, scraped, and replaced. A more economical method is to conduct the products of combustion first through an iron tube, where oily substances are deposited, and then through a series of iron condensing chambers, where the carbon is deposited, the purest product being obtained from the final con- denser. This method of condensing is employed in the manufacture of the finest grades of lamp-black, the source of the smoke being fatty oils burned in lamps. An impure form of black of bad colour is prepared from certain kinds of coal, and is chiefly used for pitching ships. Other varieties of black are Spanish black from cork ; vine Hack from the twigs of the vine ; peach black from, peach kernels ; and German black, said to be obtained from a mixture of wine-lees, peach kernels, and bone shavings.* More modern methods of preparing black for printing inks are described in chap. x. Manufacture of Indian Ink. The fullest source of information on the Chinese methods 'of preparing the ink from the lamp-black is still C/ien-ki-Soiten's book as trans- lated into French by Jametcl. .From that we learn that the lamp-black is first sieved into glazed vases, and then dried in paper bags suspended in a dry chamber. The glue is prepared either from fish or from ox-hide, and is used in the proportion of four to five catties f to each pound of lamp-black. If too little glue be used, the ink is blacker, but not so permanent. The solution of the glue is poured through a sieve on to the lamp-black, and the paste thoroughly mixed and heated for fifteen minutes in a tightly closed vessel over boiling water. It is next pounded for four hours in a mortar (see Fig. 9), until the * Lewis, PkUotophico-technicum,) 1763, p. 377. f A catty = 800 grammes. CARBON AND CARBONACEOUS INKS 2$ mass becomes thoroughly pliable, after which it is mixed with musk and camphor and beaten into long sticks. These are then moulded into cakes weighing about 114 to 140 grammes, and the cakes dried by desiccation in well-burnt ash from rice straw, which is replaced daily by fresh ash. The desiccation takes from one to three or four days or longer, but if the process be continued too long the ink becomes pale and loses its brilliancy. The following proportions are given as the best for an ink that will become blacker with age : * Lamp-black from dryandra oil, 10 catties ; old ox-hide glue, 4J catties ; old fish glue, i catty ; extract of sou-mou and another Chinese aromatic plant, I catty. The addition of a small quantity of dried ox-tongue is said to give a violet tint to the ink, whilst finely powdered vegetable matter is added to produce a bluish tint. The glue must be white and transparent. It was formerly obtained from various substances, such as rhino- ceros' and stag's horn, but is now exclusively prepared from ox-hide or from fish. A decoction of the plant Hibiscus mutdbilis was formerly used, but according to Jametel has long been discarded. At the present day the only essential difference in the ink produced by different Chinese manufacturers is that different proportions and methods of incorporating the chief ingredients are employed. The methods of preparing Chinese ink, which are given in a history of China published by du Halde, a Jesuit mis- sionary ,f in 1735, agree in all essential details with the above account. Lamp-black from pine wood or from oil was mixed with glue or with gum tragacanth and aromatic essences, and the paste pounded, and stamped into tablets, which were finally dried for three to ten days in cold ashes. In Japan the lamp-black is obtained chiefly from sesame oil or from pine wood, and is mixed with ox-hide glue in a copper vessel surrounded by another vessel con- taining hot water. The plastic mass is beaten in wooden moulds into cakes, which, as in the Chinese method, are dried by contact with absorbent ash. * Jametel, loc. clt. p. 28. f Description cle V Empire He la Chine, Parin, 1735, ii. p. 245- 30 INKS AND THEIR MANUFACTURE Eislcr * describes a method of preparing Indian ink from animal and vegetable charcoal mixed with milk and thick gum water, and allowed to dry into cakes. A modern (European) method of preparing Indian ink consists of triturating the lamp-black with a dilute solution of potassium hydroxide, so as to form a cream. This is poured in a thin stream into slightly alkaline water, and the deposit collected, washed with water, dried, and incorporated with a decoction of the seaweed known as Irish moss or carrageen, to which a little musk has been added. In another process, a solution of gelatin is boiled under pressure for two hours, and then for one hour more, over an open fire until suitably concentrated, and then mixed with lamp-black that has been heated to redness in a closed crucible. The object of heating the gelatin is to convert it into the so-called gelatin-peptone, which does not solidify like ordinary gelatin. Indian ink thus pre- pared does not gelatinise in cold weather. Merim4e f a ^ so prepared Indian ink by mixing a strong decoction of galls with a solution of glue, washing the precipitate with water, and dissolving it in a fresh solution of glue, which is then concentrated to the required con- sistency and mixed with lamp-black. Dextrin js some- times used in place of gelatin or glue in the manufacture of cheaper qualities. .Lenher% asserts that he has prepared Indian ink of equal quality to the best kinds obtained from China by the following method : Petroleum or turpentine oil is burned in lamps to which the supply of air is limited. The smoke is conducted through a zinc tube, 100 feet in length, the inclination of which is slightly upwards. The soot deposited at the remote end of the tube is in a very fine state of subdivision, and requires but little purification to free it from the tarry matter which, if not removed, would cause the ink to have a brownish tint. For this purpose it is first boiled with nitric acid, then washed with water by decantation, then boiled with strong sodium hydroxide solution, and finally washed and dried. * Dintefass, 1770, p. 31. f loc. cit. p. 197. I Die Tinten Fabrikation, 1880, p. 180. CARBON AND CARBONACEOUS INKS 31 The purified product, consisting of nearly pure carbon, is mixed into a paste with a clear solution of guin, and heated and stirred until evaporated to the required con- sistency. It is now allowed to cool gradually, a little tincture of musk being added before it quite sets, and is finally kneaded on a flat plate, and pressed into metal moulds, from which the rods are ejected by tapping. An inferior kind of ink has been prepared by Lcnher from ordinary soot purified in a similar manner. This method of preparing ink from purified soot was published long before Lcnlicrs book appeared.* In this country little, if any, Indian ink now appears to be manufactured in the form of cakes. During the war between China and Japan there was a great dearth of the ink, and although some of the largest dealers tried every possible source to obtain a supply, they were unable to do so. From this it would seem that the solid product is now only to be procured from the far East. This conclusion receives further confirmation from the fact that a large firm dealing in artists' materials now supplies much more of a liquid preparation of lamp-black than of the cakes of Indian ink. Qualities of Indian Ink. The ink is imported into England from China in the original boxes, each holding i Ib. According to the size of the sticks, 8, 20, or 40 may go to the pound, and are spoken of in the trade as "eights," "forties," &c. The sticks are of various forms, some being in squares, some in tablets, and some octagonal/ The best qualities of sticks are generally dis- tinguished by being gilt, and are stamped with very fine impressions, such as dragons, lions' heads, &c., which denote different qualities well recognised in the trade. They are obtained from Yutshing and Yenshing. The octagonal sticks are also of very fine quality. The sticks known as " Mandarin " are of fine quality, and are distingushed from ordinary sticks, which have also a lion on the top, by having a finer impression of the Chinese characters on their sides. The commonest kind are in the form of small sticks with white letters on the side. * Dingier* g polyt. Juurn. 1832, xliv. 237. 32 INKS AND THEIR MANUFACTURE Examination of Indian Ink. Among the Chinese the quality of ink is tested by rubbing the tablet on the palette. If only a faint sound is heard the ink is con- sidered to be of good quality (Si-mo), but if a loud noise is produced it is regarded as inferior (Tsou-mo). When rubbed with water, Indian ink should yield a uniform liquid, free from coarse particles or flakes. The best Chinese inks have a brilliant violet shade, whilst inks of the second quality are brilliant black, and inferior inks have a yellow tint. A good ink should not lose its inten- sity or brilliancy on keeping, and should colour paper a brilliant black. Inferior inks lack either blackness or brilliancy, or both. A practical test of the quality of an ink made from pine soot was recommended by Julien* This consisted in leaving a fragment in water, and noting the time before it rose to the surface. The better the quality the longer the ink was said to remain submerged. Practical Tests. We submitted several of the different grades of Chinese ink to practical tests, first of all re- ducing each to powder, and immersing o. I grm. in 10 c.c. of water. It soon became apparent that the better-class inks were far more readily soluble in cold water than were the cheaper kinds, some of the latter hardly colouring the fluid after some hours' soaking. The various samples were then put in a water-bath and raised to the boiling- point, but the cheaper grades were still more refractory than the others, and required to be rubbed down in a mortar before the particles of carbon were diffused in the liquid. After allowing the containing bottles to rest for an hour, it was found that the sediment of the best samples of ink was of a much finer character than that of the others. Our next experiment was to test the tinctorial value of the different samples by applying the solutions to What- man paper, first of all with a full brush covering a long strip of paper while it was pinned on a sloping drawing- board. Each strip of paper was treated with a different sample of ink, and when the first coat was dry, a second * Ann. de Chim., 1833, liii. 314. CARBON AND CARBONACEOUS INKS 33 was applied, not covering the whole of the strip, but leaving a small portion at the end with the first coating untouched. A third, fourth, and fifth coat followed, each falling short of the preceding one, until at the end of the strip a strong black represented the sum of all. A glance at the results at once showed the advantage of employing the better class of ink, for the cheaper kinds were by comparison lacking in covering power, and there were present particles of carbon which gave rise to streaks under the brush. The best inks worked far more smoothly than the inferior kinds, and opacity was reached with fewer washes. And from what has been already stated it will be evident that in the better class of material the labour of rubbing down the pigment from the solid stick is reduced to a minimum. From our examination of the sediment formed in the inks under examination, it would seem that in the better grades lamp-black of much finer quality is employed than is used in the manufacture of the cheaper kinds. These sticks of Chinese ink are exceedingly brittle, and those rendered unsaleable by breakage are commonly ground up in water to form the liquid ink so much employed by draughtsmen and artists in " black and white.'" Chemical Composition of Commercial Indian Inks. The specimens of the four grades of ink submitted to the practical tests described above gave the following results on analysis : Indian Ink. Water. Carbon residue. Nitrogen in residue. Nitrogen in Original Ash. Ink. Per cent. Per cent. Per cent. Per cent. Per cent. I. Octagonal stick . 8.16 53-9 0.0 7-74 4.08 II. Lion stick, tine letters III. Lion stick, coarse 7.20 52-53 4.87 3.69 letters IV. Small stick, coarse 9-93 49.64 7.26 4-96 letters 9.40 57-04 6.84 4.OI 34 INKS AND THEIR MANUFACTURE The fact that the residue left, on extracting the soluble substances with hot water, is free from nitrogen, affords simple means of distinguishing between Indian ink and pure sepia (see p. 22). 3. CARBON WRITING INKS. Ancient Carbon Inks. The characters on Egyptian papyri and in Latin and Greek MSS. are frequently much darker and more distinct than those written centuries after with modern iron-gall ink. The latter can be readily destroyed by various chemical agents, such as acids and bleaching agents, and their permanency is also largely dependent on the relative proportion of iron and tannin in the ink, and on the manner in which they have been kept. Astle, who was Keeper of Records in the Tower of London, and thus had exceptional opportunities of study- ing MSS. of all ages, found that the black ink used by the Anglo-Saxons in documents of the seventh, eighth, ninth, and tenth centuries had preserved its original intensity much better than that used at later periods, especially in the sixteenth and seventeenth centuries, which was fre- quently very faint. It was rare to find faded writing in documents before the tenth century. Astlc * came to the conclusion that this was due to the earlier inks containing carbon ; but Blagden, on testing the writing with potas- sium ferrocyanide, found that iron was present in every instance.f It is impossible to determine the exact period when carbon inks were replaced by iron-gall inks, though it was probably early in the present era (cf. Historical Introduc- tion). The ink of the Greeks and Latins, like the modern Oriental inks, was a mixture of finely divided carbon with a solution of gum or glue, sufficiently dilute to flow from a reed. In reality they were only modifications of the Chinese inks described above, and in some cases were even dried before use. Thus Vitruvius + states that atramen- * Origin of Writing, 1803, p. 209. f Trans. Roy. Soc,, 1787, Ixxvii. [ii.] 451. j Lib. vii. 10. CAEBON AND CARBONACEOUS INKS 35 turn was prepared from the soot of pitch pine collected on the walls of a marble chamber, mixed with gum (glutinum) and dried ; and Dioscorides * gives the proportions of soot to gum as three to one. Evidently the brilliancy of the black deposit and the more fluid character of iron-gall inks led to their gradually superseding carbon inks for writing purposes, of which no mention is made in mediaeval literature. It is true that Wecker in 1582 gave a formula for an atramentum 'perpetuum, but this was really a printing ink consisting of linseed oil and lamp-black, and there is no reference to an aqueous carbonaceous ink in his book or in that of Oanneparius ( 1 660). Modern Carbonaceous Inks. Lewis ^ in 1764 made various suggestions for rendering ink more permanent, some of which are described more fully in chap. xiv. His principal plan was to add finely divided lamp-black or ivory black to a good iron-gall ink, but such ink could be bleached by chemical means to destroy the gall ink and then washed with water to remove the carbon. Other chemists have made use of an essential oil, or of a varnish or saponified resinous substance, or a solution of gluten, to retain the carbon in suspension. Of various old formulae on these lines mention may be made of the following : Westrunitfs Ink.\ Galls, 3 parts ; Brazil wood, I part ; water, 46 parts. Boil until reduced to 32 parts. Strain and add ferrous sulphate, ij part ; gum arabic, ij part; indigo, i-J part; and lamp-black, f part. Close's Ink. Powdered copal (25 parts), in lavender oil (200 parts), mixed with lamp-black (2\ parts), and indigo (I part). If too thick the ink was thinned with turpentine. Sheldrake's Ink.\\ A mixture of asphalt dissolved in turpentine with amber varnish and lamp-black. The whole question of the best means of rendering writing safe from attempts to remove it is discussed in chap. xiv. * Opera, lib. v. cap. 183. t loc. cit. Nicholson's Journ. of Nat. Philosophy 1802, iv. 479. Nicholson's Diet, of Chem., 1820. || Ibid. CHAPTER II. TANNIN MATERIALS FOE INKS. CONTENTS. Galls : Origin Aleppo galls Chemical composi- tionChinese galls Chemical composition Japanese galls Acorn galls Oak-apple gaily Other galls Tannins Classi- fication of tannins Suitability of tannins for ink-making Chestnut bark and wood Chestnut extract Chestnut tannin Ink from chestnut wood Sumach Sumach tannin Ink from sumach Divi-divi Divi-divi tannin Ink from divi-divi Myrobalans The tannin of myrobalans "Valonia The tannin of valonia Ink from valonia Oak- bark tannins Reactions of oak tannins Amount of tannins in oak bark Ink from oak bark Q-allotannic acid Fer- mentation of gallotannic acid Properties Gallic acid Properties^-Reactions distinguishing between gallotannic and gallic acids. GALLS. Origin. Curious vegetable excrescences, known as galls, are frequently formed upon the branches, shoots, &nd leaves of trees, and especially upon the oak. They are produced by the female of certain species of insects, of which the best known are the hymenopterous gall-wasps (Cynipidce), which puncture the young tissues and deposit their eggs. Under this stimulus the plant juices accumu- late at the point of puncture, and a gall is gradually formed, which serves as the home of the larva. It is pos- sible that some virus injected simultaneously with the egg plays a part in the development of the gall, but the main essential appears to be the presence of the living larva. Should the egg of the insect perish from any cause no gall is formed, or if the larva dies the gall ceases to grow. Galls vary greatly both in size and shape, some, e.g., the Californian " flea seed," being very minute, whilst others, like the large galls on the roots of certain oaks, are several inches in diameter. Some galls are round and smooth like TANNIN MATERIALS FOR INKS 37 the English oak-apples ; others, like the Aleppo galls, are crowned with protuberances ; whilst others again assume fantastic forms, as in the case of the " artichoke gall " found on certain French oaks, the curious English galls shown in Figs. II, 12, and 13, and the Chinese and Japanese galls (Figs. 17 and 1 8). The forms and colours of the different kinds of galls are remarkably con- stant, and afford a means of distinguishing between the Fig. ii. English double oak-apple gall. Fig. 12. English oak gall. Fig. 13. English gall. insects, often of very similar appearance, that produce them. In the majority of cases galls contain only one larva, and are described as " monothalarnous," whilst others 38 INKS AND THEIR MANUFACTURE afford shelter and food to a colony of larvae, and the term " polythalarnous " is applied to them. Aleppo Galls. The ordinary nut-galls of commerce are commonly known as Aleppo, Turkey, or Levant galls. They are produced by the female of a gall wasp, Cynirjs gallon tinctorice, upon the branches of a small oak, Quercus infectoria, which is abundant on the Syrian coast, and on the east of the River Jordan. The insect pierces its way out of the gall after five to six months, and the unin- habited galls are then known as white galls, from their pale colour. These contain considerably less tannin than galls which still enclose the larva, and have therefore a Fig. 14. Green Fig. 15. White Fig. 16. Section Aleppo gall. Aleppo gall. of white gall. smaller commercial value. The best galls are selected ahead and harvested before the insect escapes, and from their colour are known as blue or green galls. The value of a given sample of galls depends to a large extent upon the proportion of white galls it contains. Hence, fraudulent attempts are sometimes made to arti- ficially close the holes left by the insect, and so make the galls to appear to still contain the larva.* A section of the nut would readily detect this fraud (see Fig. 16). Aleppo galls vary somewhat in size, but usually average from 8 to 15 mm. in diameter. They are globular or pear-shaped, and are crowned with numerous tubercles (Fig. 13). The colour ranges from greenish black to pale * Allen, Commercial Organ. Anal. TANNIN MATERIALS FOR INKS 39 yellowish green, whilst the interior is pale brown or yellowish green. The appearance of white galls is shown in Fig. 1 5 , and in section in Fig. 16, the latter showing the small canal through which the insect made its way to the surface. When a thin section of an inhabited gall is examined under the microscope it is seen to consist of an external layer of small cells, forming a sort of bark : beneath these are cellular layers of parenchyma, some of the cells con- taining tannin and chlorophyll ; then come radial cells surrounding the central cavity, in which lies the larva in the midst of an alimentary mass. Smyrna galls appear to be a commercial variety of Aleppo galls, being somewhat larger and darker in colour, and often containing a larger proportion of white galls. Chemical Composition. By treating 500 grains of the best Aleppo galls with distilled water, Davy* obtained an infusion of specific gravity 1.068, containing 185 grains of solid matter, consisting of 70.27 per cent, of tannin ; 16.75 P er cent, of impure gallic acid ; 6.48 per cent, of gum and other extractives; and 6.50 per cent, of salts of calcium and other metals. From the results of tannin determinations made by later chemists there appears to be little doubt that Davy had not extracted the whole of the soluble constituents of the galls, for on this basis the insoluble woody fibre amounts to 63 per cent, of the total substance. In 1845 Guibourt^ made a very exhaustive examination of Aleppo galls ; and his results, still quoted as final in text-books, are as follows: Tannin, 65.00; gallic acid, 2.00 ; ellagic and luteogallic acids, 2.00 ; chlorophyll, 0.70; brown alcoholic extract, 2.50; gum, 2.50; starch, 2.00 ; woody fibre, 10.50 ; sugar, proteid, potassium and calcium salts, 1.30; and water, 11.50 per cent. A later, though less complete, analysis is that of Watson Smith, who found Aleppo galls to have the following composition: Tannin, 61.65; gallic acid, 1.60; woody * Trans. Roy. Soc., 1803, xciii - 2 33- t Archiv. der Pharm., 1846, li. 190; Hist. Xat. des Drogues, 1849, ii. p. 286. 40 INKS AND THEIR MANUFACTURE fibre, 15.68 ; water, 12.32 ; and colouring matter and loss, 8.75 per cent. Biichne-r* obtained the following amounts of extractive matter by treating the powdered galls with different solvents : Per cent. Substances extracted by ether .... 77.00 ,, by ether and alcohol . 80.40 ,, ,, by cold water . . . 86.50 A specimen of commercial nut-galls examined by us contained 44 per cent, of tannin, determined as gallo- tannic acid by the method described in chap. iii. Chinese Galls. The curious variety of galls exported from China are not formed by a gall- wasp like most of the commercial galls, but are produced by a small aphis (Aphis Chinensis) upon the leaf, stalks, and shoots of Ehus semialata, a tree growing abundantly in sandy places in Northern India, China, and Japan. The aphis is about ^V of an inch in length by about T V in breadth at the base of the abdomen, which gradually widens out from the thorax (see Fig. 19, p. 45). The gall is at first dark green, and gradually changes to yellow before the larva escapes through the walls bursting open, the Chinese peasants collecting them shortly before the change takes place. The aphides are killed by exposing the galls in osier baskets to the action of steam. The gall is naturally covered with a light powder termed " salt powder" by the Chinese, and used by them for flavouring soup and as a medicine. f As imported into Europe the galls are pale grey in colour, and have a hornlike appearance, and a curious odour resembling that of freshly tanned leather (Hepwortli and Mitchell). They vary greatly both in size and in form, but a characteristic shape is shown in Fig. 17. They have a horn-like texture, and when broken open present a hollow interior contain- ing a little chalk-like dust with darker particles, which when examined under the microscope are seen to be dried aphides. * Rep.f. Pharm., 1851 [3], vii. 313. f Pereira, Pharm. Journ., 1844, iii. 384. TANNIN MATERIALS FOE INKS 41 According to von Rebling* an average-sized gall contains more than 3000 aphides, and by treating the debris with warm water these swell up to about -^ of an inch in size. DLL Halde\ gives a description of these galls, which, he states, are termed ou-poey-tse by the Chinese and are used by them in the preparation of various medicinal compounds. He also states that their formation is due to a small insect. Chinese galls were first imported into Europe in the eighteenth century under the name of " Oreilles des Indes," but they did not become a regu- lar article of commerce until about 1850. They are now largely used in Germany and America as the raw material for the manufacture of tannic acid, and they form one of the princi- pal and cheapest tannin materials for the manufacture of ink. In fact, according to Dieteridi,^ gall inks are now prepared from them almost exclusively in Germany. Biichner compared their com- mercial value with that of ordinary gall-nuts. In 1851 good average blue Aleppo galls cost iocs, to 1058. per cwt. , whilst Chinese galls fetched 6$s. to 68s. per cwt. Thus, taking into account the amount of readily soluble tannin in the latter, they were I J to I J times cheaper than Aleppo galls. Chemical Composition. Specimens of Chinese galls were examined in 1817 by Brande,\\ who found them to yield 75 per cent, of soluble matter to cold water, the residue consisting of woody fibre with 4 per cent, of resinous matter soluble in alcohol. The residue from the aqueous * Archiv. f. Chem., 1855, cxxxi. 280. t Description de V Empire, de la Chine, 1735, p. 496. | Pharm. Manual, 1897, p. 680. Rep. f. Pharm., 1851 [3], vii. 329. || Trans. Roy. 8oc., 1817, evil. 39. Fig. 17. Chinese gall. 42 INKS AND THEIK MANUFACTURE extract was found to consist mainly of tannic acid with a little gallic acid. From the absence of extractives (gums, &c.), Brande concluded that these galls would not be suitable for tanning purposes, and, in fact, he found that leather prepared with them was very brittle when dried. On the other hand, he found this property rendered them par- ticularly suitable for the manufacture of ink, and the ink prepared from them proved to be less liable to become mouldy than that from ordinary galls. In 1849, Stein* described a variety of Chinese galls as possessing an odour of tobacco, and containing the following constituents: Ash, 2. CO; tannic acid, 69.14; other tannins, 4.0 ; green saponifiable fat, 0.97 ; starch, 8.20; woody fibre, 4.9; and " inert " matter, 12.96 per cent. The tannin was completely extracted by boiling the powdered galls three times with eight times their weight of water. It was regarded by Stein as identical with the tannin of ordinary galls. The ash contained potassium, calcium, magnesium, iron, chlorine, and phosphoric acid. Bley's-\ results are similar to those of Stein, viz., gallo- tannic acid, 69.0; resin and fat, 3.0; gallic acid, ex- tractives and protcids, 4.0; starch, 7.35; woody fibre, 8.65 ; and water, 8.0 per cent. Buclmer's^ analysis in 1851 gave the following results : Tannic acid, 76.97; fat and resin, 2.38; extractives soluble in water and some salts, 0.89 ; gums and salts, 5.94; and starch, woody fibre and mineral matter, 13.82 per cent., calculated upon the substance dried at 100 C. When extracted with ether these galls yielded 79.35 per cent, of soluble matter, of which 76.97 per cent, (on the original substance) dissolved in water. Buchner was unable to confirm Stein's conclusion as to the presence of other tannins in addition to gallotannic acid. He also came to the conclusion that the tannic acid was identical with that of oak-bark, and that gallic acid was only pre- * Dingler's polyt. Journ., 1849, cxiv. 433. t Archil-, d. Pharm., 1850, cxi. 297. j loc. clt, p. 323. TANNIN MATERIALS FOR INKS 43 sent in the galls in very small proportion. The mineral matter was found to consist principally of magnesium phosphate. Tannic Acid. The proportion of tannic acid found by titein, Bley, and Buchner is substantially the same when calculated upon the dried substance, viz., Stein, 79.43 per cent.; Blcy, 75 per cent.; and JJnchner, 76.97. Viedt * gives the proportion of tannic acid in Chinese galls as about 72 per cent., whilst Iskikama t found 77.4 per cent. Samples recently analysed by the authors J have given the following results: Moisture, 10.70; ash, 1.43; and substances soluble in water, 78 per cent. The tannin determined by Procter's method was 68 per cent. Viedt (loc. cit.) asserts that Chinese galls do not contain the necessary ferment for the conversion of the gallo- tannic acid into gallic acid, and that therefore they cannot be used for the manufacture of ink unless a small propor- tion of Aleppo galls or of yeast be added to the infusion. We are unable to confirm Viedt 1 s statement, which is also altogether at variance with the results obtained by van Tieghem, who has clearly demonstrated that the con- version of tannic acid into gallic acid is brought about not by a pre-existing ferment, but by the action of certain mould fungi, We have prepared ink by adding ferrous sulphate to a decoction of Chinese galls without any addition of either yeast or other galls, and found that it behaved just like ordinary gall ink, giving a writing which rapidly became black on exposure to the air. Moreover, insoluble deposits formed on exposing the ink to the iiir, and these deposits contained 6.86 to 7.56 per cent, of iron, results very near to those obtained with ink from gallotannic acid or ordinary Aleppo galls. Japanese Galls. These galls are closely allied to the Chinese galls, and are frequently stated to be identical with them. They are produced by Aphis Chinensis, or an * Dingier" 1 * polyt. Journ., 1875, ccxvi. 453. f Chem. Xews, 1880, xlii. 274. j Unpublished. Coiiqrtes Rendits, 1867, Ixv. 1091. 44 INKS AND THEIR MANUFACTURE allied aphis upon tbe shoots of Rims japonica (Siebold) or Rlius javanica (Murray). (See Fig. 20, p. 45.) They must, however, be regarded as at least a distinct variety, and in fact they are so recognised in commerce, though for ink manufacture the two varieties are used indis- criminately. According to Procter * the Japanese galls are smaller and paler, and are usually more esteemed. Ishikama t states that considerable quantities of Chinese galls were formerly imported into Japan, but that in 1880 only the native product was used. The Japanese galls (Kibushi) are plucked from the trees between July and September, and are placed in boiling water in wooden tabs for thirty minutes, and -p, ^*f\. then dried in the sun for ^1 three to four days. They are ^HB }f' \ stored in warehouses in 4BBM 11) Kiyoto, often for several years, before being used. C| The reactions given by the j^Jp tannin they contain are iden- tical with those of ordinary gall-nut tannic acid. Fig. 1 8. Japanese gall. The amount of tannin de- termined by the permangan- ate process in seven samples of different ages up to eight years ranged from 58.82 to 67.7 per cent. The old galls were very brittle, and gave much darker decoctions than the fresh galls, but did not contain less tannin. The commercial Japanese galls that we have had the opportunity of examining J undoubtedly differed both in size and shape from the Chinese product, were also softer, and had very much thinner walls. A typical Japanese gall is shown in Fig. 18. These galls contained 10.46 per cent, of moisture, 1.96 per cent, of mineral matter, and yielded 50 per cent, of tannin when boiled for three hours with successive portions of water. Mr. R. M. Prideaux, who has kindly made a micro- scopical examination of the debris in some of these Chinese * Text-book of Tanning, p. 28. f Ckem. News, 1880, xlii. 275. j Unpublished results. TANNIN MATERIALS FOR INKS 45 and Japanese galls, informs us that the two aphides are not demonstrably of different species. Those from the Chinese galls were uniformly smaller than those from the Japanese galls, and lacked the rudimentary wings of the latter ; but it would be necessary to follow out the entire life history of both in the growing galls before being able to deteruiine with any certaint}^ the specific value of the differences observed in the dead debris. (See Figs. 19 and 20.) Acorn Galls (Knopperri). These galls, also known as Piedmontesc galls, are produced by the female of Cynips Fig. 19. Aphis from Fig. 20. Aphis from Chinese gall, x 18. Japanese gall, x 18. quercus-calicis on different oaks (Q. pedunculate, Q. sessi- flora, &c.), in the forests of Austria and Hungary, espe- cially in Dalmatia, Slavonia, and Croatia. It is a large gall, 35 to 50 mm. in length, by 35 to 40 mm. in breadth, resembling Aleppo galls in having a crown of five or six points at the top. The interior is spongy, and has a spheroidal chamber containing the larva in the centre. This gall is the same as the pomme de Chene of Rtaumur* The galls are collected from August to October, after they have fallen from the trees, and are sold either whole or in the form of powder, or as an extract. They contain less than 45 per cent, of tannin, which, according to Lowe^ is the same as that of other galls, giving analytical results corresponding with the formula, C 14 H 10 9 . * Diet, des Sciences Medicales, art. '' Galles." f Zeit. anal, Chem., 1875, xiv - 4 6 - 46 INKS AND THEIR MANUFACTURE Eitncr * made an examination of the Knoppern collected in 1884 in different districts of Austria, and found them to contain about 12 per cent, of moisture, whilst the pro- portion of tannin ranged from 23.94 to 35.02 per cent. Knoppern galls are sometimes used in the manufacture of ink, though according to Viedt f their use is not com- mon, probably owing to their comparatively low propor- tion of tannin. A similar gall is also produced on the Quercus infectoria of Asia Minor, but is spherical, and has the tubercles round the centre instead of at the top. Oak- Apple Galls. The common galls known in England as oak apples are produced by a species A ^^^^^ of Cynips on the branches of the oak, f ^fl| ^|k Quercus robur, and appear to be closely J M ,\ allied to, if not identical with, the gal Is Ijl I formed on that oak throughout Central Mr^ & I Europe. They are perfectly spherical (see Fig. 21), and of a light greyish-green J or reddish colour. British galls contain very much less Fig. 2i.-0ak-apple tannin than Ale PP g alls > and generally gall. less than Knoppern. Braithwaite J obtained only an insignificant amount from Devonshire galls, but did not state what method of determination he employed. In 1856 Vinen made an examination of the galls pro- duced by Cynips quercus pctioli, after the escape of the insect. 100 parts of the galls digested with ether and water gave 26.74 parts of extract, containing 17 parts of tannic and gallic acids. According to Vinen these galls were at that time used in Devonshire for the manufacture of ink. In 1847-48 the oaks in East Devonshire became in- fected with Cynips Kollari (Fig. 22), and the galls also appeared suddenly in 1860 in great quantities in the woods * Dingler's polyt. Journ., 1885, cclv. 485. t Ibid. 1875, ccxvi. 453. J Pharm. Jtntrn. Trans., 1855, xv. 544. Hid. 1856, xvi. 137. TANNIN MATERIALS FOR INKS 47 to the North of London. According to D' Urban* these galls contained a considerable amount of tannin and made excellent ink. As there was considerable doubt as to the commercial value of British galls, and conflicting statements had been published as to the amount of tannin contained, Judd made a series of experiments on galls at different seasons, the tannin being precipitated in each case by means of Fig. 22. Gall wasp (CynijJis Kollari}. x 5^. alum and gelatin. He found that old galls hanging on the trees in December contained on the average 15.97 P er cent, of tannin, whilst mature imperforated galls gathered in August contained on the average 17.65 per cent., and half-developed and shrivelled galls 13.44 per cent. An ink of average quality was prepared from the old perforated galls. An analysis of a specimen of Cheshire galls made by Watson Smith in 1869 gave the following results : Tannin, * Pharm. Joum. Trans., 1863, xxii. 520. 48 INKS AND THEIK MANUFACTURE 26.71; gallic acid, trace; woody fibre, 47.88; moisture, 20.61 ; and colouring matter and loss, 4.80 per cent. Specimens of old oak-apple galls collected by us during the winter in Surrey contained only 1 1 per cent, of tannin as determined by Procters hide-powder method, but when examined by a colorimetric method the amount of gallic and tannic acids in terms of gallotannic acid was 30.7 per cent. These galls yielded a good ink, and there seems to be no reason why English galls should not be used in admixture with the richer foreign varieties by ink manufacturers. The French galls sometimes met with in commerce are slightly larger than ordinary oak-apples, which they closely resemble in general appearance. They are formed upon the shoots of Quercv.s ilex in Mediterranean districts. Probably some of the varieties of Punjab galls are obtained from this species of oak. Other Varieties of Galls. There are numerous other kinds of galls, some of which are of considerable impor- tance as tanning materials, but they do not appear to have been used in the manufacture of ink, though probably some of them would be suitable for the purpose. The small-crowned Aleppo galls, which are occasionally found mixed with ordinary Aleppo galls, are also pro- duced upon Quercus infectoria, but by a different insect ((7. polycera). They are about the size of a pea, and have a circlet of small projections at the top. Pistachio galls are produced by Anopleura lentisci on plants of the pistada order, and are exported from Bok- hara together with pistachios. They are red galls, about the size of a cherry, and have a characteristic taste. Mecca or Bassorah galls are produced upon an oak by Cynips insana. According to an analysis by Bley* they have the following composition: Tannic acid, 26.00; gallic, 1.60; fatty oil, O.6o ; resin, 3.40; extractives and salts, 2.00 ; starch, 8.40 ; woody fibre, 46.00 ; and moisture, 12 per cent. Tamarix galls, also known as red galls, are formed on Tamarix orientalis and other plants of the same order. * Archil", der Pharm., 1853, Ixxv. [2], 138. TANNIN MATEEIALS FOR INKS 49 They are of a bright red colour, and are about I cm. long by 0.5 cm. broad. They are extensively employed in dyeing and tanning, and in India they are used medicinally by the natives. Similar galls are produced on T. articulata in Morocco. The galls formed on the American " live oak," Q. virens, contain 40 per cent, of tannic acid, and are very similar to Aleppo galls. A soft, spongy, and very astringent gall is formed on the Californian oak, Q. lobata.* Terebinth galls are due to the action of aphides on certain species of Terebinthacecc growing in the countries bordering on the Mediterranean. They are red in colour, long and flat, and have horn-shaped projections. Within them is a large cavity in which fragments of the aphides can usually be discerned. They contain a considerable amount of tannin and a resinous juice that readily exudes. These galls are sometimes termed apples or galls of Sodom. Watt\ states that the galls produced upon Pistacia terebinthus in India are regarded by the natives as value- less, though the leaves are used for dyeing and tanning. They are sold in Bombay as pistachio galls. TANNINS. The substances to which the general name " tannin " has been applied are compounds possessing certain com- mon chemical and physical characteristics. They are widely distributed throughout the vegetable kingdom, and it is not improbable that many of them are individual substances, just as are the different fatty acids that occur in vegetable oils. When separated in a state of purity or approximate purity, tannins are odourless white or brown substances, with a very astringent taste. They are insoluble in chloroform and carbon bisulphide, but dissolve in water, alcohol, and ether. They yield blue or green insoluble compounds with iron salts, and most of them are precipitated by potassium chromate. They also usually combine with antimony, with lead, arid many other metals to form insoluble salts. With * Trimble, Tlie Tannins, vol. i. p. 63. f Diet, of Economic Products. 50 INKS AND THEIR MANUFACTURE lime water they yield precipitates of varying colour, and with gelatin they form an insoluble compound (leather). Tannins are soluble in concentrated sulphuric acid, the solution on heating becoming first red (rufigallic acid), and then black (metagallic acid). They are oxidised by nitric acid and l>y potassium permanganate, the latter reaction } forming the basis of a quantitative method of determination Classification of Tannins. Tannins are frequently described as " iron-blueing " or " iron-greening," ac- lull t tu- / insr cording to the colour of the precipitate they form with iron salts. This difference is evidently one of constitu tion, for, as Stenhouse* first showed, one group of tannin can be converted into gallic acid and yield pyrogallol, whilst the other group does not give these reactions. Thus, when heated to 160 C., different products of decomposition are formed, the " iron-blueing " tannins, of which gallotannic acid may be taken as the type, yielding metagallic acid and pyrogallol, whilst the " iron-greening " tannins produce metagallic acid and catechol. Thorpe's method of preparing pyrogallol by heating gallotannic acid in glycerin has been used as a qualitative test of the nature of the tannin : I grm. of the tannin is slowly heated to 160 C. in 5 c.c. of glycerin, and the temperature then raised to 200-210 C. for 20 minutes. The liquid is then diluted with 10 c.c. of water and ex- tracted with an equal volume of ether (or extracted with ether without previous dilution, Trimble), and the residue from the ethereal extract dissolved in water and tested for pyrogallol or catechol by the following tests : f Reagent. ryrogallol. i per ceiit. solution. Catechol. I per cent, solution. Ferric chloride . ; Ferric acetate . - . Lime water Pinewood moistened with hydrochloric acid Ked, turning brown. Dark purple. Purple, then brown. No change. Green colour. 1 5) Clear red. Violet colour. Melting-point . 131 c. mC. * Mem. Chem. Soc., 1842, i. 133. f Trimble, The Tannins, i. p. 26. TANNIN MATERIALS FOR INKS 51 When a tannin is heated with dilute hydrochloric acid (2 per cent.) in a sealed tube at 100 C. an insoluble precipitate of ellagic acid (crimson colour with nitric acid) may be formed. When boiled with nlkalies the ''iron-greening" tannins yield protocatechuic acid and phloroglucinol, or acetic acid, whilst the "iron-blueing" tannins are converted into gallic and ellagic acids. The elementary composition of the different tannins has been suggested by Trimble * as a possible means of clas^i- fication. Thus, the gall tannins, or "iron-blueing " group, contain about 52 per cent, of carbon and about 3-5 per cent, of hydrogen, whilst the "iron-greening" tannins have 60 per cent, of carbon and 5 per cent, of hydrogen, e.g. : Group I. Carbon. Hydrogen. Per cent. Per cent. Gallotannic acid . 52.10 3-52 Chestnut wood tannin . 52.11 4.40 Chestnut bark tannin . 52.42 4.67 Chestnut tannin (Xa*s) . , 52.07 3-97 Sumach tannin (Lotve) . 52.42 . 3-56 Group II. Carbon. Hydrogen. Per cent. Per cent. Oak-bark tannin (av. of 9) . 59-79 5.08 Kino tannin (Bergltolz) 59-65 , 4-87 Oak-bark tannin (Etti) .... 59-25 4.99 Catechu tannin (Lowe] .... 61.93 4.80 Tormentilla tannin (lie/nbold) 60.75 4.65 The tannins in Group I. give a white precipitate, chang- ing to blue with lime water, whilst in the case of the tannins in Group II., the colour of the precipitate is light * The Tannins, ii. p. 132. 52 INKS AND THEIR MANUFACTURE pink, changing to red or brown. Bromine water precipi- tates the tannins in the second group, but not those of the first group. Suitability of Tannins for Ink-making. Only the " iron-blueing " tannins are suitable substances for the manufacture of black ink, as has been shown by SMuttig and Neumann* who found that mixtures of extracts of pine, cafcechu, quebracho, kino, and hemlock with solutions of iron salts gave bright green colorations on paper, but after six months' exposure only rust-like stains were left. Good black inks can be prepared from algarobilla, divi- divi, myrobalans, valonia, and sumach, all of which contain " iron-blueing " tannins. Oak-bark tannin, although an "iron-greening" tannin, also contains a substance giving a blue precipitate with iron salts, and can therefore be used in the manufacture of ink (vide infra). The most important of the tannins suitable for ink are described individually in the following pages. CHESTNUT BARK AND WOOD. The Spanish, or Sweet Chestnut (Castanea vesca), is a large tree, frequently 80 feet or more in height, which grows abundantly in the countries surrounding the Medi- terranean, and in sheltered districts as far north as Scot- land. In America it is common in many of the States as far west as Indiana. The fruit is the well-known chestnut, which is largely imported into this country. Chestnut Extract. An aqueous extract of chestnut wood or bark is prepared extensively in Pennsylvania and Virginia, the decoctions being subsequently evaporated to a solid mass. According to Trimble^ it is impossible to manufacture a good extract without the use of a vacuum pan. It is said to be frequently adulterated with molasses or glucose, and is itself employed to adulterate oak bark extract. * Die Eisengallustlnten, p. 38. f The Tannim; ii. p. 129. ' TANNIN MATERIALS FOR INKS 53 Chestnut Tannin. Sheldon* who appears to have been the first to call attention to the value of chestnut wood as a tanning and dyeing material, asserted that it contained twice as much tannin as oak bark. Trimble (loc. ciL) found air-dried chips to contain 7.8$ per cent, of tannin, which is slightly higher than the amount found by Sheldon; whilst Simand\ found 8.5 per cent, in chestnut wood, and 23.52 per cent, in chestnut- wood extract of 31 Be., the determinations being made by Lowenthal's permanganate method. Nass^. was the first to prepare a tannin from chestnut wood, and to determine its composition and properties. The aqueous extract of the wood was fractionally precipi- tated with sodium chloride, and the final fractions dialysed and then extracted with acetic ether. In this way he obtained a white preparation which was soluble in water, alcohol, ether, and glycerin, and gave the following reactions when tested in a one per cent, solution. Reagents. Chestnut tannin. Gallotannic acid. Ferrous salt No change. No change. Ferric ammonium sul- phate Blue-black precipi- Blue-black precipi- tate. tate. Tartar emetic + ammo- nium chloride . Bromine water . Slight precipitate. No precipitate. Slight precipitate. No precipitate. Lime water . Light precipitate, be- White precipitate, be- coming light blue. coming light blue. Sulphuric acid (1:9) . No deposit on boil- No deposit on boil- ing. ing. When heated to 200 C. it was converted into pyrogallic and metagallic acids, and gave an acetyl derivative closely resembling that of gallotannic acid. Its elementary composition was also found to be very similar to that of gallotannic acid, as is shown by the fol- lowing results obtained by Nass and by Trimble : * A trier. Journ. Science, 1819, i. 313. f Dingler's polyt. Journ. 1885, cclv. 487. t Zeit. anal. C/iem., 1886, xxv. 134 ; also Trimble, The Tannins, ii. p. 124. loc. cit., p. 127. 54 INKS AND THEIR MANUFACTURE Chestnut tannin (AVm). Chestnut wood tannin (Trimble). Chestnut bark tannin (Trimble). Gallotaunic acid. Per cent. Per cent. Per cent. Per cent. Carbon . 52.20 52.42 52.11 52.17 Hydrogen 3-97 4.67 4.40 3.10 In Trimbles opinion this similarity in composition and reactions renders it highly probable that chestnut tannin is identical with the gallotannic acid from galls. Ink from Chestnut Wood. Sheldon (loc. cit.) in 1819 found that chestnut wood contained ^ as much substance giving a black coloration with iron (i.e., tannin), as was present in logwood (hgernatoxylin). He stated that it was probably unequalled as a material for ink, since it gave ii rich blue-black colour with iron, whilst galls or sumach used in the same proportion had a redder shade. The ink formed by chestnut decoction was blue, but on paper it dried, yielding an intense black. The permanency of the ink was tested by exposing the writing to the sun and air, and was found highly satisfactory. Schluttig and Neumann* however, in their comparative tests on the stability of inks prepared from different tannin materials, found that chestnut iron-ink, originally blue-black, was fainter than the ink from most of the other "iron-blueing" tannins (p. 50). In 1825 Giroud took out a patent (Eng. Pat. No. 5285) for a substitute for galls, to which he gave the name of " damajavag." This was prepared by soaking I cwt. of the wood of the chestnut tree, or shells of the nut, with water for twelve hours, and then boiling it with 180 to 200 quarts of water and evaporating the decoction to a paste, which was to be used in the manufacture of ink, or in tanning. An ink prepared by us from chestnut extract had a good blue-black colour. On standing exposed to the air for a month it yielded a deposit containing 7.37 per cent, of iron. * loc. cit. p. 38. TANNIN MATERIALS FOR INKS 55 SUMACH. Sumach or Sumac is the name given to the leaves of various plants belonging to the natural order Elms. Of these the Sicilian sumach, Rims coriaria, grows wild Fig. 23. Sumach (Coriaria myrtifolia) . in Spain, Portugal, and other Mediterranean districts, and is also widely cultivated in these countries. The most esteemed variety of Sicilian sumach, known as Alcamo, 56 INKS AND THEIR MANUFACTURE occurs in commerce as a light green powder with an aromatic odour. A second and inferior variety, which is chiefly used in dyeing, has a more yellow shade and con- tains less tannin. The best French sumach is very similar to that grown in Sicily. Another French variety, known as redou, is obtained from Coriaria myrtifolia (Fig. 23). In preparing sumach for the market, the branches are dried in the sun, and the leaves removed and ground to powder in mills. The leaves of the Venetian sumach, Ehus cotinus, a shrub cultivated in Italy and the south of France, contains a yellow dyestuff, and a tannin which gives an olive-green compound with iron salts, and is therefore unsuitable for ink-making. In America two species of Rkus, R. copallina and R. glabra, both of which contain much less tannin than Sicilian sumach, are extensively used as tanning materials. Sumach Tannin. The proportion of tannin in sumach varies considerably, but the usual limits are from about 1 3 to 20 per cent. Stenhouse * was the first to show the similarity in com- position a::d properties between the tannin of sumach and gallotannic acid, both yielding gallic acid and pyrogallol. The percentage composition of his sumach tannin was: Carbon, 49.73 to 50.12; hydrogen, 3.64 to 376; and oxygen, 46.24 to 46.5 I. Lowe + obtained a purer product by extracting Sicilian sumach with alcohol, treating the residue from the extract with water, extracting the tannin by means of acetic acid, and purifying it by repeatedly dissolving it in water, and precipitating it with sodium chloride. Gallic acid (which was not identified) would be left in solution in the sodium chloride treatment. Lowe confirmed Stenhouse's statement of the formation of gallic acid from the tannin. Crystals of the latter were obtained by heating the tannin solution either alone or with 2 per cent, of sulphuric acid for several hours in a sealed tube placed in a brine bath. * Mem. Chem. Soc., 1842, i. 135. f Zeit. anal. Chem., 1873, xii. 128. TANNIN MATERIALS FOB INKS 57 The properties of the Sicilian sumach tannin were found to be identical with gallotannic acid, and analysis showed them to have the same composition, corresponding with the formula C 14 H 10 9 . Lowe was doubtful whether the tannin of other species of sumach could also be regarded as identical with gallo- tannic acid. Thus a tannin prepared from Tyrol sumach contained 52.3 per cent, of carbon and 3.8 per cent, of hydrogen, corresponding with the formula C 16 H^0 10 . Moreover, this tannin differed from that of Sicilian sumach in not yielding gallic acid when heated in a sealed tube with sulphuric acid. Sumach contains a small quantity of a yellow dyestuff, quercetrin. Ink from Sumach. As the tannin of sumach is identical, or at least allied to that of galls, it was to be anticipated that it would yield an ink of a very similar character, only modified slightly by the colouring matter of the leaves. In fact, Ribeaucourt * found that the ink made from it had a greenish shade. Lewis A who made experiments in 1763 with sumach as an ink material, came to the conclusion that it was inferior to galls as a source of tannin. Schluttig and Neumann^ however, have shown that sumach iron-ink is but little inferior in durability to ink prepared from Chinese galls, and superior to "Knop- pern " ink. According to Viedt,^ ink is occasionally pre- pared from sumach on a manufacturing scale. DIVI-DIVI. Dim-dim is the name given in commerce to the dried pods of the South American shrub, Ccesalpinia coriaria (Fig. 24), which was not known in Europe until the latter . halt' of the eighteenth century. It grows in low-lying marshy lands, attaining a height of twenty to thirty feet. The pods are of a dark brown colour, and about one and a half to three inches in length. They have a very astrin- * Ann. de Cltlm., 1792, xv. 156. f loo. tit. p. 382. J loc. tit. p. 38. Dingier' s polyt . Journ., 1875, ccxvi. 453. 58 INKS AND THEIR MANUFACTURE gent taste, due to the tannin, which is for the most part concentrated in the rind immediately beneath the epi- dermis. The recorded amount of tannin ranges from 30 to 52 per cent. A commercial sample examined by us contained 36 per cent. Divi-divi Tannin. Stenhousc * separated a tannin from divi-divi, which he found to have the following com- position : Carbon, 50.12 ; hydrogen, 3.72: and oxygen, 4.62. This tannin yielded gallic acid and pyrogallol, and Fig. 24. Divi-divi pods. formed deep blue insoluble compounds with ferric salts, and was thus very similar in composition and properties to gallotannic acid. In a more extended research, Lowe | found that the tannin of divi-divi behaved with most reagents like gallo- tannic acid, from which it was distinguished, however, by yielding a deposit of ellagic acid when heated in aqueous solution in a sealed tube. He therefore described this tannin as ellagitannic acid, * Mem. Cliem. Soc., 1842, i. 141. f Zeit. anal. Client., 1875, xiv. 35. TANNIN MATERIALS FOR INKS 59 and ascribed to it the formula C 14 H 10 10 , which maylDe regarded as gallic acid, C 14 H 12 10 , minus 2 atoms of hydrogen, or gallotannic acid, (J u tT 10 O 9 , plus I atom of oxygen. Lowe, also found the same tannic acid in myrobalans. A specimen of divi-divi examined by us contained 34 per cent, of tannin determined colorimetrically, and expressed in terms of gnllotannic acid. Ink from Divi-divi. Stenhouse (loc. cit.) states that calico printers had attempted to use divi-divi as a sub- stitute for galls, but had not found it satisfactory, owing to the large proportion of other extractive matters (gums). Fig. 25. Myrobalans. In the case of ink this would not be so objectionable, and in fact Viedt * asserts that divi-divi is sometimes used in Germany as a source of ink-tannin. An ink was prepared by us from an extract of divi-divi (5 grms.), treated with I grm. of ferrous sulphate. The deposits yielded by this ink contained from 6.77 to 7.77 per cent, of iron. MYROBALANS. The dried fruit of different species of Terminalia grow- ing in India and the East Indies is sold as a tanning and dyeing material under the name of myrobalans (Fig. 25). The ripe fruit weighs between 5 to 10 grms., and has a very astringent taste, due to the tannin in the husk. * Dingier 's poly 't. Journ., 1875, ccxvi. 453. 60 INKS AND THEIR MANUFACTURE The Indian species, T. chebula, which yields the " black " or " chebulic " myrobalans of commerce, is exten- sively used in conjunction with iron salts as a black dye, and is also employed in the manufacture of ink. The earliest mention of the possible use of myrobalans as a substitute for galls is that made by A. Johnson * in a communication to the Society of Arts in 1 80 1, in which he stated that the natives in India used them to give a black colour to leather, mixing the powder with iron filings and water. A committee of the Society appointed to report on the subject found that the pulp and outer husk of the fruit gave a rich black colour with ferrous sulphate. Ink prepared by us from myrobalaus was of a good blue black colour, and yielded insoluble deposits containing about 6 per cent, of iron. The Tannin of Myrobalans. Loii'e\ found about I per cent, of gallic acid in myrobalans, and extracted a tannin which contained 49.42 per cent, of carbon and 3.16 per cent, of hydrogen, corresponding with the formula Ci 4 H 10 10 . When heated in a sealed tube at io8-uo C., a solution of this tannin yielded a deposit of ellagic acid ; and from this fact and the elementary analysis, Lowe con- cluded that it was ellagitannic acid, identical with that of divi-divi. Z'6lffel\ confirmed Lowe's statement of the occurrence of I per cent, of gallic acid, but found that the tannin was a mixture of ellagitannic acid, and a glucoside of gallotannic acid, the former being in the greater proportion. A specimen of myrobalans examined by us was found to contain 39 per cent, of tannin determined colorimetric- ally, and expressed in terms of gallotannic acid. VALONIA. Valonia is the commercial name for the acorn cups of certain species of oaks growing in Asia Minor and different * Trans. Soc. Arts, 1801, xix. 343. t Zeit. anal. Chem., 1875,^1^. 35. J Arch, der Pharm., 1891, cCxxix. 155. TANNIN MATERIALS FOR INKS the most important are 61 parts of Greece, of which cegilops and Q. macrolepis. The best sorts are gathered before the fruit is quite ripe in April, those beaten from the trees in September and October being poorer in tannin. The cups (containing the acorns) are first partially dried on the ground and then conveyed by mules to Smyrna, where they are stored in warehouses until slight fermenta- tion sets in and causes the acorns to fall from the cups. Fig 26. Valonia. If exposed to rain after gathering, the acorn cups turn black and lose a considerable amount of tannin by fermentation. As met with in commerce valonia consists of semi- circular prickly backed cups, about 50 mm. in diameter (see Fig. 26). The amount of tannin they contain varies greatly with the district, species of oak and time of collection, but usually ranges from about 20 to 45 per cent. 62 INKS AND THEIR MANUFACTURE The following percentages have been recorded inter aha: 32.4 (Handtke); 38 (Gallow); 22.6-39.2 (Rothe) ; and 31.6 to 35.64 (Simand). Eitner* examined eighteen samples of different origin of the harvest of 1886, and found them to yield from 42.4 to 51.9 per cent, of total extract, 1.6 to 3 per cent, of soluble ash, and 21.28 to 30.2 per cent, of tannins. The best valonia is that obtained from Smyrna, the Greek and Albanian products being held in much smaller esteem. In 1852 the prices per cvvt. were as follows : Smyrna, 143. to 155. ; Morea, ics. to I2s. ; and Camata, 145. to 165. (Tomlinsori). The prices given by Procter in 1885 were considerably higher, viz., Smyrna, I2s. 6d. to 2Os. 6d. ; Morea, IDS. 6d. to iSs. 6d. ; and Camata, 155. to igs. per cwt. The Tannin of Valonia. This appears to be mainly ellagitannic acid, judging by the results of Bottinger* who extracted the tannins from valonia, divi-divi, and algarobilia, and prepared the acetyl derivative of each. The amount of acetyl in the valonia tannin (44.1 per cent.) was nearly the same as that of divi-divi (43.19 per cent.) and algarobilia (43.9 per cent.), and hence Bottinger concluded that the preparations were identical in com- position. In the colorimetric estimation of the amount of tannin in terms of gallotannic acid we found the solution yielded a much bluer tinge than the standard solution of gallo- tannic acid, and it was necessary to add a slight trace of an aniline colour to the latter, in order to match the tint. In duplicate determinations we found (i) 59.1 per cent, and (2) 57.5 per cent. The amount found by Procter's hide powder method was 20 per cent., so that the tannin in valonia appears to have a greater tinctorial value than gallotannic acid. The filtrate from the hide powder con- tained iron-colouring substances (gallic acid), correspond- ing to 2.5 per cent, of gallotannic acid, Ink from Valonia. Vaionia yields a very rich bluish black ink, and appears to us to be a very suitable raw * Dcr Gerlet; 1887, xiii. 18. -f Bcr d. d. cJiem. &e$., 1884, xvii. 1503. TANNIN MATERIALS FOR INKS 63 material for the manufacture, especially if used in admixture with Chinese galls. The deposits yielded by the ink on exposure to the air are very similar to those given by gall or divi-divi inks. Thus the deposits examined by us contained from 11.5 to 12.8 per cent, of iron oxide. OAK-BARK TANNINS. Owing to the fact that infusions of oak bark give a blue coloration with iron salts the tannins present were formerly regarded as identical with gallotannic acid. This error was first pointed out by Stenhouse* who showed that an infusion of oak bark differed from a solution of gallotannic acid in not yielding gallic acid or pyrogallpl. In 1867 Ghraboivshi found that instead of gallic acid an amorphous red compound, " oak red," was produced, which Etti\ obtained by boiling an oak tannin with dilute sulphuric acid, and concluded to be an anhydride with the formula 2C 17 H 6 9 -H 2 = C 34 H 26 5 . An extended series of researches on the oak tannins were the'n made independently by JStti, by Lowe, and by Bottinger, but the most conflicting results were obtained. Thus Etti prepared an oak tannin which did not dissolve in water and gave a green coloration with ferric salts, whilst Lowe obtained soluble tannins with the formulae C 28 H ?8 14 and C 28 H 30 15 , which gave blue precipitates with iron solutions. Trimblel has given an excellent summary of these different results, though without succeeding in reconciling them. He himself has made numerous preparations, and has found the average composition of nine of these to be as follows : Carbon, 59.79; hydrogen, 5.08; and oxygen, 35.13 per cent. results which correspond best with the formula of Etti's tannin, C 20 H 20 9 . In Trimble's opinion there is no question but that oak tannins give green colorations with iron salts, and he attributes the blue colorations given by oak-bark infusions * Mem. Chem. Soc., 1842, i. 140. t Monatsh.f. Chem., 1880, i. 262. j The Tannins, ii. p. 90* 64 INKS AND THEIR MANUFACTURE to the presence of an associated colouring-matter. In the case of the chestnut oak he separated this " iron-blueing " compound by first precipitating the oak tannins with neutral lead acetate and then treating the filtrate with basic lead acetate. Reactions of Oak Tannins. On heating oak-bark tannins to 190 C. catechol is formed as the main decom- position product, whilst on fusion with caustic alkali protocatechuic acid is obtained. The colour reactions vary greatly with the species of oak whence the tannins were derived, which is evidence that they are not identical. Trimble* gives the following table of the reactions of the tannins separated from two species of oak bark com- pared with those given by gallotannic acid : Keagent. English oak, Q. robur. Indian oak, Q. sessiliflora. Gallotannic acid. Copper sulphate. Copper sulphate + ammonia. Precipitate. Red brown preci- pitate. No precipitate. Brown precipi- tate. Stannous chloride and HC1. Violet colour. Violet colour. Slight green colour. Sodium sulphite. Bromine water. Ferric chloride. Ferric chloride + ammonia. Pink colour. Yellow precipitate. Blue green colour and green preci- pitate. Purple bi*own pre- cipitate. Yellow colour. Yellow precipi- tate. Green colour aud precipitate. Purple brown precipitate. Slight pink colour. No precipitate. Blue colour and precipitate. Purple precipi- tate. Ferric ammonium sulphate. Blue green colour and green preci- pitate. Green colour and precipitate. Blue colour and precipitate. Lime water. Precipitate turning pink. Precipitate turn- ing pink. Precipitate turn- ing blue. Amount of Tannins in Oak Bark. Procter gives the proportion of tannins in European oak bark as IO to 12 per cent, whilst Trimble* found the bark of different * loc. cit. ii. p. 48. TANNIN MATERIALS FOR INKS 65 species of American oaks to contain from 4.04 to 14.21 per cent., whilst an English oak bark gave 12.37 P er cent, calculated on the dry substance. The chief species of oak from which the commercial bark is derived are Q. pedunculata, sessifl>ora, and rubescens, the first of which usually contains more tannin than the others. Eitner* has shown that the amount of tannin varies with the season. Thus in the case of the bark from Q. pedunculate he obtained the following results : April, 14.80; May, 10.71 ; June, 12.33; July, 9.8 ; and August, 1 1.23 per cent. Weiss f analysed commercial oak barks of different origin with the following results : Tannin. Ash. Origin. Per cent. Per cent. Hungarian (3) 10.36-13.47 5.68-7.31 . German (3) 11.87-16.18 6.27-8.52 French (3) . 13.82- 16.22 6.14-7.77 Danish (3) . 13.86- l6.22 6.66-7.81 Swedish 12.02- 14.59 5-55-7-05 Average . 13-5 6.82 Ink from Oak Bark. Stenhpuse (loc. cit.) found that a good blue-black ink could be prepared from an infusion of oak bark, in which respect it differed from the infusions of kino, larch, and alder barks, which only gave green colorations with iron salts. According to Prechtl,% oak bark was used fifty years ago in conjunction with other substances in the manufacture of ink. Thus he gives the following formula for the pre- paration of ink from oak-bark galls and Knoppern : Galls, 9 Ibs., logwood i Jibs., rasped oak bark, 81bs., Knoppern 61bs., gum 2lbs., ammonium chloride Jib., infused in 40 quarts of water and 24 quarts of vinegar, and the infusion mixed with ferrous sulphate. * Der Gerber, iv. 85. t Ibid. 1885. ii. 181. I TechnoL Encijclop.. 1852, xviii. 460. 66 INKS AND THEIR MANUFACTURE Since the pure oak tannins are "iron-greening" sub- stances, whilst the blue-black colour given by oak-bark infusions with iron salts is only due to the presence of an associated substance (vide supra), the use of oak bark for ink is not economical. Moreover, ink prepared exclusively from the infusion has been shown by Scliluttig and Neu- mann* to be somewhat less stable on exposure to light and air than the inks from galls, divi-divi, or other sub- stances containing " iron-blueing " tannins. GALLOTANXIC ACID. The tannin which is best known is that contained in galls, and to this the name of gallotannic acid has been given to distinguish it from quercitannic acid and other tannins. It is present in Aleppo, Chinese, and Japanese galls and in Knoppern, and has also been identified in sumach, myrobalans, and algarobilla. Pelouze f prepared gallotannic acid by extracting pow- dered gall with ether containing water, and showed that on exposure to the air in an aqueous solution it gradually yielded an insoluble deposit consisting mainly of gallic acid. Strecker^ came to the conclusion that gallotannic acid was a glncoside, which was decomposed on fermentation in accordance with the equation C, 4 H 28 0.> 2 - C 6 H 12 6 + 2C 14 H 10 9 - H,0 Glncoside. Glucose. Taimic acid. Subsequently it was shown by Scliiff^ thai perfectly pure gallotannic acid was free from glucose, and was an anhydride containing two gallic acid groups, i.e., digallic anhydride. In his opinion the glucose in Strecker's pre- paration was originally present in the galls, and had been extracted simultaneously with the tannic acid. Trimble || concludes that although gallotannic acid can be so purified as to be eventually only digallic anhydride, * Die Eisengallnxtintcn, p. 38. j Ann. Chem. P/tanit., 1833, liv. 337. j Ibid. 1854, xc. 238. Ibid. 1873, clxii. 43. i| The Tannins, i. p. 29. TANNIN MATERIALS FOR INKS 67 it is rarely if ever met with in that state in the '* pure " article of commerce, which contains variable amounts of glucose in a loose state of combination. In his opinion the commercial article must be regarded either as a gluco- side of digallic acid, or as a mixture of the glucoside and of the pure anhydride. Sdiiff (loc. cit.) prepared a pentacetyl derivative of gallo- tannic acid, which melted at 137 0., and had a composi- tion agreeing with the formula C 14 H 5 (C 2 H 3 0) 5 O a . This compound was insoluble in water and cold alcohol, and gave no coloration with iron salts. The following constitutional formula represents the for- mation of this acetyl derivative, and also the formation of gallic acid by the hydration of the gallotannic acid : OH, ,OH OH/ ! \COv OH V - , ' / Fermentation of G-allotannic Acid. Chevreul showed that by keeping a solution of gallotannic acid in a sealed tube so as to exclude atmospheric oxygen, it could be kept unchanged for an indefinite period. It has frequently been asserted (e.g., by Viedt, p. 43), that the conversion of gallotannic acid is due to the action of an oxidising enzyme. It was shown, however, by van Tieghem * that this was not the case, but that the spon- taneous change was due to the action of two mould fungi, Penicillinm glaucum and Aspergillus niger in the presence of air. By inoculating solutions of gallotannic acid with the spores of these fungi, he was able to effect a complete conversion of that substance into gallic acid in a few days. The fermentation only took place within the liquid, for when there was only a surface growth a very small amount of gallic acid had been produced after several days' vigorous fermentation. From the results of Saccs t experiments it * Complex Itc/idt/*, 1867, Ixv. 1091. t Hrid. 1871, Ixxii. 766. 68 INKS AND THEIE MANUFACTURE would seem that ordinary yeast also possesses the hyclro- lysing property of these two ferments. Properties of Gallotannic Acid. Gallotannic acid is a yellowish- white, glistening, amorphous powder, which is readily soluble in water, alcohol, and ether. When heated with dilute acids (Stenlwuse) or fermented [vide supra) it takes up water and is converted into gallic acid, whilst when boiled with alkaline solutions it yields gallic and ellagic acids. When heated alone to 160 C. it is decomposed and yields a sublimate of pyrogallol (see p. 50). Gallotannic acid gives dark violet or blue precipitates with iron salts. It is precipitated quantitatively by lead salts, with which it yields white compounds, and it forms white unstable gelatinous precipitates with antimony. It combines with gelatin to form an insoluble compound (leather). GALLIC ACID. Gallic acid (C 7 H 6 O,. + H.,0), which was discovered by Sckeele, is present naturally in small proportion in various vegetable substances, such as tea, galls, and myrobalans (about i per cent.). It is obtained from gallotannic acid by fermentation with certain mould fungi (p. 43), or by the hydrolysing action of dilute acids : C 14 H 10 9 + H,0 = 2C 7 H 6 B . Its constitutional formula shows that it may be regarded as benzoic acid, in which three atoms of hydrogen are replaced by hydroxyl groups : C 6 H,(OH), . COOH + H 2 0. Properties. Gallic acid crystallises in white silken needles, which melt above 200 C. It is much less soluble in water than is gallotannic acid, I part requiring 130 parts at 12.5 0. to bring it into solution. It is more soluble in absolute alcohol, 100 parts of which at 15 C. dissolve 27.95 parts, whilst 100 parts of ether at the same temperature only dissolve 2.5 parts. When heated alone to about 215 C. it is decomposed TANNIN MATERIALS FOR INKS 69 with the formation of pyrogallol (C H 3 . (OH) 3 ) and water. When heated with sulphuric acid at 100 C. it gives off red vapours of rufigallic acid, whilst under the influence of arsenic acid at a high temperature it yields ellagic acid. It combines with alkalies to form salts, which, in alka- line solution, absorb oxygen from the air and turn brown. Ferric salts are reduced by gallic acid with the formation of blue-black compounds containing iron in the ferrous condition (Chevreul). Ferrous sulphate free from ferric salts gives no colora- tion with gallic acid, but ferric sulphate gives a blue colour and eventually a precipitate ( Wackeiibrodcr). Unlike tannic acids, gallic acid yields no insoluble compound with gelatin. Various formula? for the preparation of ink from gallic acid are given on pp. 96-98. Reactions distinguishing between G-allotannic and Gallic Acids. It has been generally accepted that tannic acid gives black precipitates with ferric salts, and no coloration with ferrous salts,* but Ruoss^ has recently shown that these statements are incorrect. He has found that tannic acid gives a black precipitate with ferric acetate, and a black precipitate or coloration with ferrous acetate. Moreover, he has also proved that on adding a solution of a ferric salt drop by drop to a solution of tannic acid only a dark coloration (and no precipitate) is obtained, the iron tannate being readily soluble in an excess of tannic acid. Since gallic acid behaves in the same way, a dark coloration with ferric salts is inconclusive. Rnoss has therefore devised the following two new re- agents, which he has found to be both characteristic and very sensitive : Ruosss Reagent /. (i) solution of 20 grms. of ferric sulphate per litre ; (2) solution of 28 grms. of crystalline sodium carbonate per litre ; (3) acetic acid (sp. gr. 1.04), containing 5 grms. of sodium tartrate per litre. The tannin solution is diluted to such an extent that on * /-'.'A? Schluttig and Neumann, lot', fit. p. 18. j /e'^. anal, ('hem., 1902, xli. 725. 70 INKS AND THEIR MANUFACTURE adding the ferric sulphate solution drop by drop it still remains slightly transparent when the maximum colour has been reached. About 10 c.c. of such a solution are treated with the iron solution (i), which is added drop by drop until the colour ceases to become darker. The same number of drops of solution (2) are then added, and twice that quantity of solution (3). When the liquid is shaken and allowed to stand a black precipitate is obtained in the case of tannic acid, whilst gallic acid yields no such precipitate. The reaction is capable of detecting o.ooi per cent, of tannic acid. Ruoss's Reagent II. (i) a solution of 10 grms. of ferric sulphate + 15 grms. of sodium acetate + 1.7 grms. of sodium tartrate per litre; (2) a solution of 1.25 grms. of gelatin in 125 c.c. of hot water, made up to a litre with glacial acetic acid (sp. gr. 1.064). Ten c.c. of the tannin solution are treated with solution (i), added drop by drop until the colour ceases to darken, and then with the same quantity of solution (2). After being shaken and left for some time a flocculent blue- black precipitate indicates tannic acid. Rvoss's Oxidation Reaction* One drop of the ferric sulphate solution (20 grms. per litre) is added to 10 c.c. of the tannin solution, diluted as required in the test with Reagent I. A permanent dark coloration is obtained with tannic acid, whilst gallic acid gives a black coloration, immediately changing to yellow. If ferric acetate be used instead of ferric sulphate the dark coloration is permanent with gallic acid as well as with tannic acid. Gricssmayer' 1 s Reaction^ for tannic acid consists of adding one drop of a solution of tannin to a very dilute solution of iodine. The liquid becomes colourless, and on now adding a drop of a dilute solution of ammonia a blood-red colour is produced. Ruoss has pointed out that the reaction is also given by gallic acid, and is therefore inconclusive. * Ice. cit. p. 732. f Classen, HanttftHcJi dermal. Anal., p. 163. TANNIN MATERIALS FOR INKS 71 Hydrogen Peroxide as a Reagent. It has recently been found by Mitchell and Hepworth that on adding hydrogen peroxide to a solution containing tannin and ferrous sulphate, there is an immediate black precipitate, the tannin being precipitated quantitatively, or nearly so, as a basic tannate ; whereas gallic acid treated in the same way yields a dark-brown solution, but only a slight trace of any insoluble compound. The precipitates yielded on ignition from 30 to 34.5 per cent, of ferric oxide, and thus approximated in composition to one of Ruoss's basic tannates (p. 79). We attempted to base a quantitative method of sepa- rating tannin on this reaction, but were unable to obtain concordant results. CHAPTER III. NATUKE OF INKS. CONTEXTS. Constitution of ink-forming- substances Influence of light and air Iron tannates Evidence of an inter- mediate blue iron oxide Tannates of iron Basic salts Methods of estimating tannates Procter's method Jackson's lead carbonate method Ruoss's ferric sulphate method Colorimetric methods Hinsdale's colorimetric method Mitchell's colorimetric method.. Constitution of Ink-forming Substances. The property possessed by gallic and tannic acids of forming blue com- pounds with ferric salts has been attributed by Schiff'* to the presence of free phenoloid hydroxyl groups, to which is also due the analogous colorations obtained with other compounds of the aromatic series. Thus, when a coloration is obtained with ferric chloride the presence of a free hydroxyl group may be inferred, and vice versa. F.OY instance, a violet coloration is given by phenol, salicylic acid, phenyl-sul phonic acid, &c. ; a Hue coloration by gallotannic acid, gallic acid, pyrogallol, arbutin, and many derivatives of tannic acid ; a green colour by ma&y tannins, resculetin, and parresculetin ; a red or reddish- ciolct colour by phloridziu, ty rosin, &c. ; whilst no colora- tion is obtained with picric acid, dinitro-hydroquinone, acetylgallic acid, &c. Schiff also came to the conclusion that the intensity of the colour stood in relation to the number of free hy- droxyl groups, the substances giving violet colorations containing only one free hydroxyl, whilst deep Hue-black colorations were produced by compounds containing several free hydroxyl groups. Thus, phenyl-sulphonic * Ann. ('hem. Pharm.. 1^71, clix. 164. NATUEE OF INKS 73 /OH acid, C H H,<^ , gives a violet colour, whilst gallic \SO,.OH acid, C 6 H 2 (OII) 3 COOH, gives a blue-black colour. Scliiff's work was extended in a special direction .by Kostaneckiy* who investigated the relation between the constitution of certain organic dyestuffs and their tinctorial properties. He found that phenoloid colouring matters combine with oxide mordants when they possess two hy- droxyl groups in the ortho position. la a subsequent communication, Kostanecld f gave the name of " tinctogen group " to that atomic grouping which enables dyestuffs to combine with oxide mordants. The further question of the formation of permanent "inks" upon vegetable fibres was thoroughly studied by Sc.ldv.Uirj and Ncumann.l In ord?r to determine whether any phenol compound giving an intense coloration with iron salts was suitable for ink, they made a series of tests in which each substance was dissolved in water (with a little alcohol if required), and then treated with the same proportion of a solution of ferrous sulphate. The liquids were allowed to run down white paper stretched at an angle of 45, so as to form stripes 3 to 6 mm. in breadth as in their " Stripe test " (p. 121), which were then allowed to dry. In the case of phenol, resorcin, hydroquinone, phloro- glucinol, orcin, triacet} 7 lgallic acid, trimethyl-pyrogallol, and some other compounds, nothing but a faint yellow stain due to iron oxide was obtained. On the other hand, dark violet colorations of varying intensity were given by gallic and tannic acids, pyrogallol- carboxylic acid, methyl and ethyl esters of gallic acid, potassium pyrogallol-sulphonate, and haematoxylin. From these and similar experiments Schluttig and Neu- mann established the fact that in order to yield colours forming a permanent ink on paper, the compound must contain three hydroxyl groups in juxtaposition. For * Bfti: (1. d. chem. GM.. 1887, xx. 3146. t I'titf. 1888, 3113, foot-note. J Die Eixengalluistutten, p. 16. 74 INKS AND THEIR MANUFACTURE /OH instance, hydroquinone C G H 4 \ does not yield an ink, X OH whilst haematoxylin and gallic acid, each of which con- tains three adjacent hydroxyls, C 6 H 2 (OH) ;{ OH (i) Lr. n W JoH(2) Ha?aiatoxyliii, Gallic acid, give permanent colorations. The colour produced by the other substances were as resistant to the action of water, and in some cases (e.g., esters of gallic acid and haematoxylin) more resistant than the ordinary inks of gallic and tannic acids. Influence of Light and Air. On exposing stains given by the different compounds that formed inks for six weeks to the action of a current of air and bright sunlight, the following results were observed : Completely bleached. Colours of paroxybenzoic acid and ortho-carboxylic acid. Yclloivish-grey. Pyrogallol-sulphonic acid, tribrompyro- gallol, dibrom-gallic acid, and tannic acid. Dark, grey.- Monobrom-gallic acid, pyrogallol-carboxylic acid, and gallic acid. Dark, brown. Pyrogallol . Greenish, or Bluish-black. Pyrocatechin, protocatechuic acid, methyl arid ethyl esters of gallic acid, and haema- toxylin. From these results it appears that the inks of tannic and gallic acids are not the most permanent, but are far ex- ceeded in this respect by logwood (hsematoxylin) and other inks. The behaviour of tannic acid ink was remarkakle, for it was the faintest of the colorations in its group ; whereas if the accepted formula, in which there are five hydroxyl groups, of which three at least are adjacent (see p. 67), be correct, it should have been one of the darkest. ^ The stability of the inks was found to stand in propor- tion to their darkness on exposure. NATURE OF INKS 75 Schluttig and Neumann consider that these experiments show conclusively that a determination of gallic or tannic acid in an ink (as prescribed by a German statute, p. 14), without reference to the presence of other compounds of the same character, is of no value as a test of the perma- nency of that ink. It has recently been shown by one of us (Mitchell)* that this law of atomic grouping, established by Scliluttig and Neumann, in the case of iron inks, also applies to inks con- taining ammonium vanadate in place of iron. IRON TANNATES. Compounds of Iron and Tannic Acid. Numerous metallic compounds of tannic acid have been prepared, but the iron salts are of primary importance in the manufacture of ink. Although nickel, cobalt and manganese are so closely allied to iron, it is remarkable that none of them forms an C 14 H 8 9 ' (C 14 H 9 9 ) 5 Schluttig and Neumann.^ on repeating Wittsteirts experiments, obtained a series of five spontaneous pre- cipitates which they removed from the ink from time to time, the final one being collected after an exposure of five weeks. These five precipitates, dried at 1 00 C., were found to contain from 6.27 to 6.6 1 per cent, of iron, the average in the five being 6.35 per cent. Precipitates obtained in the same manner and allowed to dry spontaneously in the air contained on the average 4.8 per cent, of iron. A complete analysis of one of these air-dried deposits gave the following results: Carbon, 35.77; hydrogen, 5.19; iron, 4.80; and oxygen, 54.24 per cent. When these precipitates were dried at 100 C., the dark violet colour changed to black, and they were * Jahresber. der Chem. (von Berzelius), 1848, xxviii. 221. f Ann. Client. Pharm., 1875, clxxv. 176. I Die Eisengallvstlnten, Dresden, 1890, p. 44. 90 INKS AND THEIR MANUFACTURE Amount of Iron in Spontaneous Deposits from Inks. s Time o -~ *- - No. Tamiin, &c. O 3 Si! 3 before col- lecting the 'E~ g ^ "2 $ if * 55 * deposit. * Grms. C.c. /0 /0 I Gallotannic acid (86 %) (2 grms.) 3 100 i month 822 5-75 2 >5 3 days. 849 5-94 x Gallotaunic acid (86 %) (3 grins.) 3 IOO i week. 7.85 549 2 " " " 8.63 6.04 I Gallic acid (3 grms.) (very slight 2 250 2 mouths 25 I7-.S deposit) I Chinese galls, decoction from 12 3 200 10 days 98 6.8 grms. 2 >> 11 v 7 11 10.3 7-2 3 V 11 11 ,. 3 weeks 10.8 7-55 4 11 " " '-' 2 10.81 7-56 I Aleppo blue galls, decoction from 5 i 2OO 4 days 8.62 6.04 grms. 2 11 11 11 11 ,, ,, TO 11.03 7-72 3 11 " " 3 weeks 10.6 7-4 I English oak-apple galls, from 5 i 200 2 weeks 14.8 10.3. grms. 2 ., i week 129 9.0 3 " " " i 13-1 9.2 I Japanese galls (5 grms.) i 3CO 6 days 11.31 7-9i 2 11 11 11 3 weeks ii 02 7.71 3 6 days n.6 8.12 I Divi-divi (5 grms.) i 200 2 weeks 9.70 6.77 2 Ink decanted from first sediment " " 2 ii. ii 7-77 I Yalonia (5 grms.) i 200 i week 11.53 8.! 2 11 11 . " '' 2 weeks 12 78 8.9 I Myrobalans (5 grms.) i 2OO 2 weeks 8.21 5-74 I Chestnut extract (5 grms.) 5 200 i month !0.53 7-37 MANUFACTURE OF IRON GALL INK 91 then found to contain 6.34 per cent, of iron, or the same proportion as the deposits dried directly at that tempera- ture. The ratio of iron to tannin was thus as I : 14.27 (or I part of ferrous sulphate to 2.88 parts of tannin) a result which agreed fairly well with Dieterictis empirical propor- tion of i part of iron to 1 5 parts of tannin. As there was considerable discrepancy between the composition of the deposits obtained by Wittstein and by Schluttig and Neumann we thought it advisable to once more repeat the work, using not only tannin, but also extracts of different kinds of galls. Our results, which, as will be seen, were obtained under very varying conditions, are summarised in the preceding table. Our results with pure gallotannic acid are thus more in agreement with those of Wittstein than with those of SchhUtig and Neumann. It was hardly to be expected that galls should not contain other substances besides gallotannic acid forming insoluble compounds with iron, and this probably accounts for the higher percentage of iron found in the deposits from English oak-apple gall ink, &c. We found that the precipitates attacked the paper if dried on the filter at 100 C., and we therefore in most of our experiments washed the deposits into a platinum basin, in which they were subsequently dried and ignited. The ratio between the iron and gallotannic acid in our dried deposits was as I : 16, which corresponds with a ratio of i : 3.22 between the ferrous sulphate and gallo- tannic acid. Hence each part of ferrous sulphate requires 3 parts of pure gallotannic acid. Since, however, the proportion of tannic acid varies in each kind of material employed, the proportion of tannin material must naturally vary correspondingly. The following table, giving the approximate proportions of different materials, is based upon the average amount of tannin they contain, and on the results of the preceding experiments. It is assumed that practically the whole of the tannin is extracted in each case. 92 INKS AND THEIR MANUFACTURE Proportion of Tannin Materials required ly 1 part of Ferrous Sulphate. Tannin material. Containing pure taunic acid. I Parts by weight required. Per cent. (circa). Commercial gallotannic acid 86 3-8 Aleppo galls ... 62 5-o Chinese galls 75 4-3 Japanese galls 62 5-o Acorn galls (Knoppern) 3 ii English oak-apple galls 26 12.5 Chestnut wood . . .. . .> 9 36 extract . . 20 16 Sumach . ... . 22 14.6 Valonia* .*.''. ... 30 ii Divi-divi . . . *. ' < 40 8 Myrobalans ... ... 30 ii OLD TYPE OF IRON GALL INK. When solutions of gallotannic acid and ferrous sulphate are mixed the liquid at first remains colourless, and it is only when oxidation takes place that a violet-black solution and eventually a violet-black deposit is formed. In the older type of iron gall inks it was therefore necessary to expose the liquid to the air for some time to obtain an ink which would give writing of sufficient im- mediate blackness, although even the writing with the colourless solution of gallotannic acid and ferrous sulphate gradually becomes black when dried. In other words, a "provisional colour" was formed by partial oxidation of the ink, and the insoluble deposit was kept in suspension by the addition of a sufficient quantity of gum arabic. Ink thus oxidised yielded an immediate black writing, but had the drawback that that portion of the ink in which the oxidation was complete did not penetrate into the fibres of the paper, but was attached to the surface by means of the gum and could be washed off. * See also colorimetric results on p. 86. MANUFACTURE OF IRON GALL INK 93 In a good iron gall ink of the old type it was therefore essential to have only so much of the ink oxidised as to give an immediate black colour, leaving the remainder in an unoxidised state to penetrate into the paper, and form the black insoluble oxidised tannate within the fibres. On boiling an iron gall ink the oxidation process is accelerated, and there is also some decomposition of the tannate, so that a complete ink should not be boiled. Provisional Colouring Matters. The paleness of the writing with unoxidised ink has also been obviated in many inks by the addition of logwood extract (p. 102), or more recently of various aniline colours (p. 13). Such colouring matters as Prussian or Turnbull's blue, ultra- marine, or the various blue compounds of .copper, are quite unsuitable for the purpose, since they are either too insol- uble or react with the tannin and injure the colour of iron tannate. The most suitable and most widely employed substance as a provisional colour is indigo, the presence of which is a characteristic feature of the so-called " aliza- rine " inks (vide infra). Old Formulae of Iron Gall Inks. The earliest method of preparing iron gall ink that we have discovered is that of the Elizabethan domestic ink, the formula of which is shown in the frontispiece. Elizabethan Ink. Rain water (or claret wine or red vinegar), i quart ; galls, 5 oz. ; ferrous sulphate, 4 oz. ; gum, 3 oz. After five days' soaking, the extract from the galls was heated just to the boiling-point with the ferrous sulphate (see also the rhyme of de Beau Chesne, p. 12).! Canneparius* (1660). Galls, 3 oz., macerated in 30 oz. of white wine for aix days, and the extract mixed with i oz. of ferrous sulphate and 2 oz. of gum arabic, and left for four days. Lewis 1[ (1760). Galls, 3 oz. ; rasped logwood, i oz. ; water, 2 to 3 parts ; gum, varied at discretion, but about J oz. per pint. The ink to be shaken daily for ten to twelve days. " Celebrated Black Dresden Ink " ( 1 770). J - Galls, 2 Ibs. ; ferrous sulphate, J Ib. ; gum, 6 oz. ; alum, 2 oz. ; verdi- * De Atram-entis, p. 270. f loc. clt. p. 377. 94 INKS AND THEIR MANUFACTURE gris, I oz. ; and salt, I oz. ; in 2 quarts of vinegar and 2 quarts of rain water. Decanted after two days and shaken daily for eight days. Eisler* (1770). Galls, 4 oz. ; ferrous sulphate, 2 oz. ; gum, I oz. ; in a quart of rain water. Ribeaucourt f (1792). Galls, 2 oz. ; ferrous sulphate, I oz. ; copper sulphate, J oz. ; gum, I oz. ; and logwood, I oz. ; in 24 oz. of water. Reid I (1827). Galls (i Ib.) extracted twice with 3 pints of boiling water, and the extract (2 quarts) mixed with 3^ oz. of ferrous sulphate and the same quantity of gum. UNOXIDISED IRON GALL INKS. The use of indigo as a means of improving the colour of ink was mentioned by JSisler in 1770 (loc. cit.\ and was used in this country by Stephens^ in 1836. In 1856 Leonhardi,\\ of Dresden, patented in Hanover an ink consisting of an extract of 42 parts of Aleppo galls and 3 parts of madder in 120 parts of water, mixed with ii parts of indigo solution, 5-!- parts of ferrous sulphate, and 2 parts of metallic iron, dissolved in crude acetic acid. Subsequently the madder was omitted as superfluous, but the inks still retained the name of "alizarine" ink, although quite free from alizarine. The more suitable name of " isatin " inks never met with popular acceptance. In " alizarine " inks the process of oxidation is pre- vented as far as possible, thus keeping the liquid free, to a large extent, from insoluble deposit, and giving it much greater power of penetration into the paper. The pre- sence of the indigo makes the writing immediately blue, and it subsequently changes to black as the oxidation of the iron tannate proceeds within the fibres of the paper, the oxidation process being completed in eight days at the most. The addition of indigo also increases the permanency of the ink, so that the writing offers much more resistance * Eisler, Dintefass, p. 7. f Ann. de Chim., 1792, xv. 113. J Philos. Mag., 1827, ii. 114. Mechanics' Mag., 1836, xxv. 229. || Dingier' s polyt. Journ., 1856, cxlii. 141. MANUFACTUKE OF IRON GALL INK 95 to the action of bleaching agents than ordinary iron gall inks. Owing to the absence of gum the inks flow more readily from the pen, and are less liable to clog ; but, on the other hand, the presence of free acid in considerable pro- portion causes the pen to be corroded. Thus we found that an ordinary steel pen left in a typical commercial "alizarine" ink from which air was excluded had lost 5 per cent, in weight after six weeks, whilst the ink itself had become semi-solid. Indigo blue is soluble in concentrated sulphuric acid, and the solution can be diluted to a great extent without yielding a deposit. Viedt* gives the following method of preparing " aliza- rine " ink : a 5 to 6 per cent, solution of sulphindigotic acid is treated with sufficient iron to form the necessary amount of ferrous sulphate for the tannin present. The excess of free acid is then nearly neutralised with chalk or marble, leaving only a slight amount to retard atmo- spheric oxidation of the ink. The clear solution is then decanted from the insoluble calcium sulphate and mixed with a 5 to 6 per cent, decoction of galls, yielding a green solution through the mixture of the yellow gall extract and blue indigo solution. Inks containing neutral indigo carmine, i.e., the sodium or potassium salt of sulphindigotic acid, yield deposits much more readily than inks containing free sulphuric acid, though the latter also form sediments in time. Indigo carmine is prepared by dissolving indigo in sulphuric acid, adding alkali, and collecting and washing the precipitate. Prollius' " Alizarine " Ink, which was recommended by Bley\ as superior to any then sold, was prepared from (i) i\ Ibs. of galls, with sufficient water to yield 5 Ibs. of decoction; (2) 4 oz. of indigo powder mixed with ij Ibs. of fuming sulphuric acid, and allowed to stand for 24 hours ; then diluted with 5 Ibs. of water, treated with 8 oz. of powdered chalk, and 8 oz. of iron filings, filtered and added to (i). * Dingier'*? polyt. Jo-urn., 1875, ccxvi. 533. f Ibid. 1857, cxlv. 77. 96 INKS AND THEIR MANUFACTURE With the object of reducing the corrosive action of the sulphuric acid in the ferrous sulphate upon steel pens several manufacturers have proposed to ignite ferrous sulphate until a white powder was left. It is difficult to see what advantage such a process can have. Desormaix's gall ink* and Heinle 's non-corrosive wi&t were prepared in a similar manner. GALLIC ACID INKS. Edcl,\ in the course of his investigation on gall inks, pointed out that after the conversion of gallotannic acid into gallic acid more than twice as much ink was produced. Thus 448 parts of galls required 144 parts of ferrous sul- phate, but after the conversion of the gallotannic acid into gallic acid 336 parts of the iron salt were necessary to obtain an ink of the same intensity. To effect this conversion in practice, he exposed a decoc- tion of lib. of galls to the air for ten days with continual daily shaking, and then added to each quart of the liquid 3^ pints of water, 9 oz. of ferrous sulphate, and 9 oz. of gum. Dieterich has also recommended inks prepared from gallic acid solutions obtained by the oxidation of gall extracts or tannin solutions. Oxidised Gall Extracts. 200 parts of powdered Chinese galls are moistened with water and kept in a warm place (20 25 C.) until quite mouldy, the water being renewed daily, so that the galls feel moist but not wet. After eight to ten days the fermentation is complete, and the galls are ex- tracted with successive portions of hot water, and filtered after the addition of some talc, the total amount of extract and washings amounting to IOOO parts. Oxidised Tannin Solutions. 100 parts of tannin, 100 parts of water, and 20 parts of hydrochloric acid (sp. gr. 1.16) are heated for ten hours on the water bath at 80 90 C., and then gradually diluted with 900 parts of dis- tilled water. * Nicholson's Diet, of Chein., 1820, p. 507. f Prechtl's Technol. Encyclop., 1852, p. 460. | P/nlosojrft. Mag., 1827, ii. in. Pharm. Manual, 1897, p. 680. MANUFACTURE OF IRON GALL INKS 97 For writing inks Dieterich finds that either ferrous or ferric salts can be used with such oxidised solutions, but for copying inks only ferrous salts can be employed (see chap. xii.). He gives the following directions for preparing inks on these lines : I. Gall Ink. 1000 parts of the oxidised gall decoction are mixed with 100 of ferric chloride solution containing 10 per cent, of iron, the ink left for two weeks in closed flasks and then decanted. II. Oxidised Tannin Office Ink. 100 parts of tannin, 100 of water, 200 of ferric chloride solution (10 percent. of iron), and 10 per cent, of crude hydrochloric acid (sp. gr. 1. 1 6) are mixed and heated for ten hours at 8o 90 C. The liquid is then diluted with 700 parts of hot water, left for an hour at the same temperature (with renewal of the evaporated water), cooled, kept in a closed flask for two weeks, filtered, and diluted to 1000 parts. With inks thus prepared the writing is at first hardly perceptible, so that a provisional colour is necessary, as in the case of the following formulas : Blue-black Iron Gall Ink. Three parts of phenol blue in 400 parts of water are mixed with 600 parts of oxidised gall ink (I.) and I part of phenol, and left for a week in a loosely covered flask, after which the clear ink is decanted. Violet-black Ink. Prepared in the same way, except that 1.5 parts of phenol blue 3 F., and 2.0 parts of Ponceau red R.R. are used, instead of the 3 parts of phenol blue as the provisional colour. Bed-black Ink. Six parts of Ponceau red R. used. Green-black Ink. Six parts of aniline green used. Black Ink. The provisional colour consists of 10.5 parts of aniline green D, 9 parts of Ponceau red R, and I part of phenol blue 3 F. " Alizarine J| Ink. Four parts of indigotin, and 2.4 parts of aniline green as colouring matter. Blue-green Ink. 1.5 parts of phenol blue, and 2.5 parts of aniline green as colouring matter. For the formulae of gallic acid copying inks on these lines see chap. xii. State of Massachusetts Official Ink. This is a mixed tannic and gallic acid ink, containing the following con- 98 INKS AND THEIR MANUFACTURE stituents : Dry gallotannic acid, 23.4 ; gallic acid crystals, 7.7 ; ferrous sulphate, 30; gum arabic, 25 ; dilute hydro- chloric acid, 25 ; and phenol, i, in looo parts of water. JAPAN INKS. When an iron gall ink has been oxidised so as to have become converted for the most part into the insoluble black iron tannate (p. 13), it no longer possesses the penetrating properties of the freshly prepared ink, and requires the addition of a considerable amount of gum to keep the in- soluble powder in suspension. Such ink gives an immediate black writing, which dries on the surface of the paper with a varnish-like gloss, whence this sort of ink was termed Japan ink by Ribeaucourt. Since the oxidation has taken place within the ink instead of partially within the fibres, as in the old type of partially oxidised gall inks, Japan inks are more easily re- moved than other gall inks. They have also the drawback of stickiness and of the excess of gum clogging the pen, whilst they also readily yield large deposits. Ribeaucourtfs Japan Ink contained the ingredients given in the formula on p. 94. In Neiuton's English patent (No. 836; 1865), complete iron gall ink was oxidised by being percolated through narrow openings in the bottom of a vessel; whilst Carter (Eng. Pat., No. 1982; 1873) obtained the same results by subjecting the ink to the action of a current of air. Both of these processes produce "Japan" inks of the very opposite type to " alizarine " inks. CHAPTER V. LOGWOOD, VANADIUM AND ANILINE BLACK INKS. CONTENTS. Logwood inks Logwood Logwood extract Haematoxylin Hsematein Iso-hfematein Addition of log- wood to gall inks Logwood inks without tannin Chrome logwood inks Hasmatein inks Use of logwood in patent inks Vanadium inks Black aniline inks. LOGWOOD INKS. Logwood. This well-known dyeing material consists of chips of the wood of Hcematoxylon Campechianum, a large tree (40 to 50 feet high) belonging to the Ocesalpiniacece. It forms large woods on the Atlantic side of Central America, in Mexico, and in the West Indies. It was first discovered by the Spaniards in the Bay of Campeachy, in Mexico, and exported by them into Europe.. When introduced into England in the reign of Queen Elizabeth, it was soon employed to adulterate other dyes,, and its use was prohibited as " affording a false and deceitful colour" injurious to the Queen's subjects, "and discreditable beyond seas to our merchants and dyers." This Act was not repealed until 1661. The constant hostilities between the Spanish and English led to the tree being acclimatised in the W'est Indies in 171 5, though subsequently a treaty was concluded giving the English the right of cutting and exporting the wood from Campeachy. The best Indian logwood is not so valuable as the Mexican product, whilst Honduras wood is intermediate in value. In preparing it for the market, the wood is first divided into logs about 3 feet in length, which are then cut into- 100 INKS AND THEIR MANUFACTURE chips by means of a revolving drum provided with steel catting knives. Logwood Extract. Formerly the chips were moistened and exposed to a fermentation process, but manufacturers now endeavour to exclude all oxidising influences. Three methods are in use in the preparation of the log- wood extracts of commerce.* The finely divided chips are frequently digested with hot water under a pressure of I to 2 atmospheres, this process yielding a large extract, which, however, contains a considerable proportion of resins, fats, and other impurities ; (2) the French method of boiling the chips with water at the ordinary pressure, which yields a smaller though purer extract; and (3) a diffusion process, in which an apparatus similar to that used in the sugar industry is employed. The yield by this process is smaller than in (i) or (2), but the shades of colour are finer. If the wood has not undergone any fermentation, the extracts contain chiefly haematoxylin and but little haema- tein. They are sold either as liquids, with a density of about 10 Be., or as solid gum-like masses. The tinctorial value of an extract is usually determined by practical dyeing tests, the amount of colour fixed on wool mordanted with potassium bichromate and tartaric acid on treatment with a definite quantity of the dried sample, being compared with that given by a sample of standard quality. Logwood extract may be adulterated with molasses or tannin materials, though according to Rupe (loc. cit.) adul- teration is not so frequent now as formerly. Hsematoxylin [C 16 H 14 6 ]. The colouring matter of logwood does not occur ready formed in the cells, but in the form of a compound, hcematoxylin, which becomes purple on oxidation. Heematoxylin was first discovered by Ckevreul f in 1810, and termed hcematin by him, a name subsequently changed to hcemato&ylin by Erdmann, t in order to prevent con- fusion with the colouring matter of blood. * Rupe, Die Chem. der natur. Farbstoffe, p. 107. f Ann. Ckim. Phys., 1812, Ixxxii. 53, 126. j Ann. diem. Pharin., 1842, xliv. 292. LOGWOOD, &c., BL J AdK IKKS 101 It was obtained by Hesse * in the form of colourless crystals containing 3 molecules of water [which they lost at I2OC.], by extraction with ether containing some water and in the presence of alkali bisulphate. Properties. Hasmatoxylin has a sweet taste and is slightly soluble in water, but dissolves readily in alcohol and ether. The crystals turn red on exposure to the light (in the absence of air) without changing in composition. Solutions of silver and gold salts are rapidly reduced by them, as is also the case with Fehliug's solution. Stannous chloride gives a rose-coloured precipitate, ferrous ammo- nium sulphate a slight violet black precipitate, and lead acetate a white precipitate changing to blue. On dry distillation hsematoxylin yields pyrogallol and resorcinol (or a derivative). From a consideration of the results obtained on acetyla- tion Reim"\ concluded that hgematoxylin had the constitu- tional formula C 6 H 2 (OH) 3 C 6 H 4 Haematein (C 10 H 12 6 ). The colouring principle of log- wood, which is first formed by the oxidation of the pre- existing hgematoxylin, was discovered by 0. Erdmann (loc. <.)-C w H M 6 + = C 16 H 12 6 + H 2 . It forms small anhydrous yellowish-green crystals with a metallic lustre. These are only slightly soluble in water, alcohol or ether, though the solution in water (0.06 per cent, at 20 C.) has an intense colour. It is readily soluble in alkalies yielding violet or purplish brown compounds. The ammonium compound, C 16 H 12 6 .2NH 3 is precipitated by most metallic salts. Thus it yields a violet blue precipitate with copper sulphate, a dark violet with alum, and a black one with iron ammonium sulphate, whilst it reduces a solution of silver nitrate (Hesse). I * Ann. Ckem. Pharm., 1859, cix. 332. f Ber. fl. d. diem. GfS., iv. 329. loc. clt 102 ' INKS AND THEIR MANUFACTURE Hgematein is reduced to hasmatoxylin by treatment with sulphur dioxide or hydrogen. Iso-hsematein. When haematein is treated with concentrated sulphuric acid, it dissolves, forming a brown solution, from which, on standing, a crystalline P TT O ^ compound, 16 n rj 5 h S0 4 , is deposited (Hummel and Perkiri). * If hydrochloric acid be heated with hsematein in a sealed tube the colour of the solution changes to dirty yellow, and on evaporating the liquid a crystalline de- posit of iso-haematein hydrochloride is left. By treating this with silver hydroxide, and concentrating the solu- tion in vacuo, iso-hcematein is left as an amorphous mass with a greenish metallic lustre. Jso-hasmatein has the same composition as hasmatein, which ifc also resembles in its general reactions, though many of the metallic compounds have a more reddish shade of purple. Addition of Logwood to Gall Inks. The advantage of adding a small proportion of logwood decoction or extracb to iron gall inks has long been known. In 1763 Lewis t made a series of experiments to determine whether such an addition had any injurious effect upon the stability of the ink, and found that the colour of the ink was materially improved without lessening its permanency. He recommended an ink consisting of 3 parts of galls, I of ferrous sulphate, and I of logwood in 40 to 60 parts of water, with gum arabic in the proportion of J part to 20 of ink. Eisler's Logwood Gall Ink \ was prepared from the fol- lowing ingredients : Logwood, 8 oz. ; ferrous sulphate, S oz. ; vinegar, i quart ; rain water, J quart ; galls, 4 oz, ; gum arabic, 4 oz. ; alum, 2 oz. ; and indigo, I oz. The ink was left for fourteen days in the sun before using. Ribeaucourt in 1792 agreed with Lewis's statement as to the advantage of logwood in iron gall ink, and recom- * Ber. d. d. chem. Ges., xv. 2337. f loc. clt. p. 377. | Dintefasx, 1770, p. 8. Ann. de C/iiui., 1792, xv. 113. LOGWOOD, &c., BLACK INKS 103 mended that its proportion should be half that of the galls. These earlier results have been fully confirmed in our time by the experiments of ScTiluttig and Neumann^* who found that ink prepared from hasmatoxylin and iron was even more permanent than those prepared from gallo- tannic or gallic acids. Reid's Gallic Acid Logwood Ink. Eeid-\ found that when gallic acid was employed in place of gallotannic acid logwood might be added in the proportion of I J parts to i part of the former. He prepared an ink 011 these lines by exposing a decoc- tion of galls (i lb.), to the air for ten days with occasional daily agitation, so as to convert the gallotanuic acid into gallic acid, and then adding a decoction of logwood ( i J Ibs.), 1 8 oz. of ferrous sulphate, and 18 oz. of gum. This ink is decomposed by alkalies and alkali carbon- ates, the iron being precipitated. Logwood Inks without Tannin. Since haematoxylin, the active agent in logwood, contains three adjacent hydroxyl groups it follows the general rule established by Schluttiy and Neumann of yielding an ink with iron salts. The iron logwood inks have a greenish shade, which gradually changes to black as the writing dries, Alum logwood inks have a deep violet-black colour, and chromium logwood inks a violet colour changing to black, Chromic acid added to logwood gives a deep black precipitate, whilst potassium chromate yields a black ink, and if added in excess a black precipitate. In fact, as Viedt J has shown, precipitates are gradually formed by oxidation in logwood inks containing alum or iron or copper salts, though more slowly than in iron tannin ink. The addition of logwood in excess does not prevent this, but the deposition is retarded for a long time by completely excluding the air. Reiniges Iron Logwood Ink. This may be taken as a * Die Eisengallustuiti'n, p. 33. f Ph'dos. Jfag., 1827, ii. 115. j DlnyleSs polyt. Journ., 1875, ccxvi. 456. Ibid. 1857, cxliii. 240. 104 INKS AND THEIR MANUFACTURE typical ink of this class. It is prepared by dissolving 2 grms. of logwood extract, and 3 grms. of ferrous sulphate, in 100 c.c. of water, then adding 10 grms. of crystalline sodium carbonate, and finally, 2 grms. of oxalic acid. After complete settlement the ink is decanted from the sediment, and a suitable proportion of gum added. This ink gives a good black writing, but we have found that the characters become somewhat brown after two or three months. The writing gives a red coloration with hydrochloric and other acids, due to the logwood, and a bluish-green with acidified potassium ferrocyanide solution indicating the iron. It is gradually bleached by bromine water. Reid* who investigated the character of the inks formed on adding ferrous sulphate to logwood, found that the greenish-blue compound first formed was gradually oxidised to a brownish-black compound. Copious deposits were given, hdwever, by such logwood iron inks, and hence Reid concluded that logwood should not be employed alone, or should not exceed a third of the amount of the galls in mixed inks. Bb'ttgers Alum Copper Logwood lnk.-\ This is pre- pared by boiling i part of alum, 2 of copper sulphate, and 4 of logwood extract with 48 parts of water, and filtering the solution. The filtrate is a red-violet ink, which writes pale violet, but rapidly darkens and soon becomes jet black. On treatment with bromine water the writing is changed to red, and then to faint brownish yellow. The great drawback of the ink is its instability, and it must be kept in tightly corked bottles, which should contain as little air as possible. Ink prepared by us and kept in a corked bottle containing air had yielded a dense deposit in six weeks, and gave only faint writing. Violet-Black Rouen Ink (Encre Ueu rouennaise).+ This consists of a decoction of 75 parts of logwood in 600 parts of water, to which is added 3 J parts of alum, 3 parts * Philos. Mag., 1827, ii. 114. f Dingier''* polyt. Juttm., 1857, cxliii. 240. I Ibid, 1859, cliii. 77. LOGWOOD, &c., BLACK INKS 105 of gum arable, and I J parts of sugar candy. The ink is allowed to stand for two or three days and then strained. Inks containing only logwood and alum write with a reddish-violet colour, which gradually changes to a dark, though not absolutely black, shade. Viedt's Copper Logwood Ink. * In Viedfs opinion copper sulphate should always be used in preference to ferrous sulphate in logwood ink, since it gives a blacker writing. He recommends the following formula: Logwood extract, 20 kilos, in 200 kilos, of water, mixed with a solution of IO kilos, of ammonium alum in 20 kilos, of boiling water, and the mixture treated with 0.2 of sul- phuric acid and 1.5 of copper sulphate in 20 litres of water. In order to obtain a darker liquid the ink is exposed to the air for some days before being bottled, thus producing a " provisional colouring," similar to that given by indigo in " alizarine " inks. Inks' of this type are sold under different names, e.g., Chemnitz violet-black ink. To obviate the paleness of the writing given by such inks as this Stark has prepared a writing and copying ink which gives immediate black characters by adding a little chromate to a copper logwood ink, the chrome ink in this case representing the "provisional colouring" matter (cf. P- 13)- Viedt (loc. cit.) states that he has never known an ink of this kind to gelatinise. A great objection to all logwood copper inks is that they cannot be used with steel pens, which gradually withdraws the copper from them. The addition of free sulphuric acid, as in Viedt's copper logwood ink (supra), retards the formation of a deposit, but causes the ink to corrode steel pens. Chrome Logwood Iiiks. Runge^ discovered that by adding a very small proportion of potassium chromate to a decoction of logwood a deep black fluid was obtained, which could be used at once as a writing ink, though by * Dingier' 1 s polyt. Journ., 1875, ccxvii. 76. f Gnmdrlss dcr Chem., 1847, ii. 205 ; Dingier'' 8 polyt. Journ., 184$, cix. 225. 106 INKS AND THEIR MANUFACTURE increasing the amount of chromate a black precipitate was produced. Runge's Chrome Ink. This was originally prepared by boiling 10 Ibs. of logwood with water until 80 Ibs. of decoction were obtained, and adding potassium chromate in the proportion of I part in 1000. If solid extract of logwood be used, 1 5 parts are dis- solved in 1000 parts of water, and I part of potassium chromate added. Runge claimed for this ink the advantage of yielding permanent black writing, and of not acting upon steel pens. Gopel* considered that Runge's formula had too large a proportion of logwood, as shown by the red-brown edges of drops of the ink on white blotting paper, and advocated the following proportions : logwood extract 24, potassium chromate 2, and water IOOO parts. In the proportion recommended by Karmarsch (i of chromate to 8 of logwood extract), the ink is too grey, pointing to an excess of the chromate. Although Eunge had been able to use his ink con- tinuously for two years, it has been found by others that after a time a coagulation due to some unknown cause may occur. Stein f made numerous experiments to find a remedy for this, and eventually found that the addition of 4 grains of mercuric chloride to a bottle liquefied the coagulated ink and prevented it from becoming thick again. Viedt. however, asserts that Stein's remedy is useless, and that a better remedy is the addition of sodium car- bonate, as in Bottgers writing and copying ink (infra). He found that such ink, when kept in a well-closed ink- stand, remained fluid for two years, and hence concluded that the best means of preserving the ink was to com- pletely exclude the air. Bottger's Modification of Runge's Ink. Fifteen parts of logwood extract are dissolved in 900 parts of boiling * Dingier'* polyt. Jouni., 1859, cli. 80. f Ibid. 1850, cxv. 77. j Ibid. 1875, ccvii. 76. Ibid. 1859, cli. 431 ; 1869, csci. 175. LOGWOOD, &c., BLACK INKS 107 water, and 4 parts of crystalline sodium carbonate dissolved in the clear decanted solution ; a solution of I part of potassium chromate in 100 parts of water is finally added. We have prepared ink by each of the above methods. Runges original formula gives an ink which yields very black characters, but requires the pen to be frequently filled, or the writing appears faint. The ink kept in a test-tube closed with cotton wool was perfectly liquid after three months, though it then yielded writing with a browner tinge. A simultaneous experiment in which a drop of formalin had been added to the ink gave analogous results. Ink prepared at the same time by Bottgers modification, and kept in a large well-closed flask (containing air), had a slight mould on the surface and gave dirty brown writing. Hence the addition of phenol or other preserva- tive is essential for this ink. According to Viedt, Plasters " Chrome Ink Powder' 1 and Poncelet's " Ink loithout Acid" are imitations of Eunge's original chrome ink. A " blue black ink," consisting of logwood decoction and chrome alum, gives writing which is too pale and grey. Bichromate Logwood Ink.* One hundred parts of log- wood extract are dissolved in 800 parts of lirne water, and when solution is complete, 3 parts of phenol and 25 parts of hydrochloric acid are introduced, and the whole left for thirty minutes on the hot-water bath. It is then cooled and filtered, 30 parts of gum arabic, and 3 parts of potassium chromate added, and the ink diluted to make 1800 parts. This ink is violet-red in colour, and the writing at first appears reddish-brown, but rapidly darkens, and within five minutes has a bluish-black tint. It keeps well and yields very little deposit, but, owing to the amount of free hydrochloric acid present, it has a considerable action upon steel pens. Thus in one of our experiments a pen left in the ink had lost 4.5 per cent, in weight after six weeks, whilst the ink itself had become semi-solid. * Dingier'* potyt, Journ., 1882, ccxlv. 475. 108 INKS AND THEIR MANUFACTURE The basic chloride and acetate of chromium are some- times used instead of potassium chromate in chrome inks in order to lessen the tendency to gelatinise. IHeterich's School Ink. Dieterich * recommends the fol- lowing as a cheap and effective ink for school purposes : 200 parts of a 20 per cent, solution of logwood extract are diluted with 500 parts of water and heated to 90 C. A solution of 2 parts of potassium bichromate, 50 of chrome alum and 10 of oxalic acid in 150 parts of water, is then added, drop by drop, and the mixture maintained at 90 C. for thirty minutes, and then diluted to 1000 parts and mixed with i part of phenol. After standing for two or three days it is decanted, and is then ready for use. Hsematein Inks. Inks prepared with hasmatein in place of logwood extract have more brilliant shades, but are wanting in lustre, and are readily decomposed on heating. An alkaline hasmatein ink is prepared by mixing 12 parts of hsematein with 720 parts of water for two hours at about 20 C., and then decanting the liquid, heating it to 30 C., and adding 3 parts of crystalline sodium car- bonate. When cold, 0.5 part of potassium chromate in 48 parts of water is gradually added with constant stir- ring, and lastly 12 parts of gum and 0.5 part of phenol, with sufficient water to make 960 parts of ink in all. Schmieden's Acid Hcematein ink consists of 24 parts of haematein dissolved in water (760 parts) at a temperature not exceeding 39 C. ; then acidified with 80 drops of strong sulphuric acid, mixed with a solution of 4 parts of ferrous sulphate in 48 parts of water, 1 2 parts of hydro- chloric acid, and diluted to 960 parts, a sufficient quantity of gum being subsequently added. This ink is dark red in colour, and gives dark red writing, which changes to brown and then to black within twelve hours. For other formulae of logwood inks see Copying Inks, chap. xii. Use of Logwood in Patent Inks. Logwood fre- quently occurs as an ingredient of patented inks. Thus Phar-m. Manual, 1897, p. 685. LOGWOOD.. &c., BLACK INKS 109 it is used in WhitfiMs Indelible Safety Ink (Eng. Pat. No. 7474; 1837), whilst Scott (Eng. Pat. No. 8770 ; 1840) prepared a similar indelible ink, consisting of a logwood iron ink with the adding of gum, indigo, Prussian blue, gas-black and iron nitrate. In J. Eeades patent (Eng. Pat. No. 11,474; 1846), the precipitate obtained by adding metallic salts (iron, copper, potassium bichromate) to logwood extract is incorporated with a printing ink. In 1856 (Eng. Pat. 342), C. and Gr. Swann claimed an ink prepared by adding potassium bichromate with a sufficient quantity of potassium bicarbonate, potassium chlorate, mercuric chloride and ammonia to a decoction of logwood. Underwood (Eng. Pat. No. 1112 ; 1857) patented a leg- wood copying paper (see chap. xii.). whilst logwood and haBmatoxylin ink powders were claimed by Cooley (Eng, Pat. No. 1 06; 1867), Byford (Eng. Pat. No. 974; 1876). and Grunwald (Eng. Pat. No. 963 ; 1881). Joly (Eng. Pat. No. 4484; 1875) prepared an ink by the action of tungsticacid upon colouring matters, such as those of logwood, elderberries, &c. Fonscca and Co. (Eng. Pat. No. 859 ; 1883) used logwood as an ingredient of an indelible carbon ink ; and Frusher {Eng. Pat. No. 8241 ; 1885) has patented the manufac- ture of ink from waste logwood and potassium bichromate from dyeing processes. VANADIUM INKS. The discovery of the fact that ammonium vanadate forms a black ink with gallotannic acid is attributed to Berzelius* but we have been unable to discover any refer- ence to the subject either in the Jahresberichte or Lehrbuch of Berzelius. The statement that this ink is of a very permanent character has been copied from one text-book to another, and is still found in different standard works on chemistry. In 1 889 Appelbaum ) made a number of experiments * Dlnglet-'s polyt. Journ., 1835, Ivi. 237. t Ibid. 1889, cclxxi. 423. 110 INKS AND THEIR MANUFACTURE v:ith inks thus prepared from gall extracts and solutions of pure gallotannic acid, and found that both the ink itself and the writing faded after the lapse of a few weeks. Hence he doubted whether Berzelius had ever made any experiments with the ink. We have repeated the work of Appelbaum * and can confirm what he says about gall vanadium inks, though vve find that gallotannic acid gives an ink of somewhat greater permanency than was found to be the case by him. It has been shown by one of us (M.) that the law estab- lished by Schluttig and Neumann (p. 73), for iron salts also applies to ammonium vanadate i.e., that it yields black ink with substances containing three adjacent hydroxyl groups. Thus gallic acid, logwood extract, hsematoxylin and pyrogallol combine with ammonium metavanadate to form black inks, whilst phenol, benzoic acid, saccharin, &c., do not form such compounds. None of these inks, however, have proved satisfactory in our hands. Although they give an immediate black writing, the characters gradually turn yellow, even when protected from the light. Hence, apart from the question of expense, ammonium vanadate cannot be regarded as a suitable constituent of writing inks. It has, however, been claimed as an addition to inks in various patents. Thus Pinhney (Eng. Pat., No. 2745, of 1871) prepares an ink from an aniline salt with. a salt of vanadium or uranium and an oxidising agent ; and a similar patent was taken out by G-rawitz (Eug. Pat., No. 1620; 1875). The use of vanadium is claimed by Hickisson as a constituent of a marking ink (Eng. Pat., No. 5122; 1878), whilst it is also used \>j Just, Weiler, and Heidepriem in their patent safety ink (Eng. Pat., 16,757 ; 1890). Mitchell (loc. cit.) has described certain reactions of vanadium enabling vanadates to be readily distinguished from chromates, which are frequently very similar in colour. * Analyst, 1903, xxviii. 146. LOGWOOD, &c., BLACK INKS 111 BLACK ANILIXE INKS. The formation of aniline black in a fine state of division within the fibres of the paper was described by Jacobsen as an indelible ink for writing or marking, though it has chiefly been used for the latter purpose. Various brands of nigrosine, which are the sodium salts of the sulphonic acids of anilidophenyl- anilidodiphenyl and dianilidodiphenylsafraniu-hydrochloride, are used in the preparation of a black writing ink. They are readily soluble in water, and when dissolved in the proportion of about i part in 80 yield a solution which flows readily, dries to a good black, and has no action on metallic pens, The solution keeps well, and the writing resists the action of different chemical reagents, although it can be removed or smudged by water, and lacks the permanency of good iron gall ink. Coupler and Collins' u Indulin ink,"* which was awarded a prize in Paris, was a blue-black ink which, according to Viedt, contained nigrosine or similar aniline dye-stuffs. Solutions of nigrosine were sold under the name of " stylographic ink " when first manufactured in 1867, owing to the readiness with which they flowed from stylographic pens. Particulars of other aniline inks are given under Coloured Inks in chap. vi. * Dingier 1 s polyt. Journ., 1867, clxxxiii. 78. CHAPTER VI. COLOUKED WKITING INKS. CONTENTS. Historical Coloured aniline inks Fugitive- ness of aniline inks Patented coloured inks. Historical. One of the earliest references to the use of a coloured ink is by Plutarch, who mentions a red ink (nvppov fiapfjia) with which certain letters were marked on the doors of the dikasts in Athens. Red ink com- pounded with minium or vermilion seems to have been used for the titles of books among the Romans,* whilst JSidonius (vii. 12) states that rubrica (red ochre) was used for the same purpose. It is interesting to note that our word "rubric," which is applied to the titles of sub- sections printed in red, thus finds its origin. A reddish-purple ink was prepared by the Romans from the Murex, the mollusc which yielded the famous Tyrian dye. Montfaucon was of opinion that this was the source of the ink used by the Byzantine Emperors in their signatures to documents. In fact, the use of any red ink was forbidden to any one excepting those of royal blood.f If the Emperor was still a minor, his guardians signed for him in green ink, the general use of which was pro- bably also interdicted to some extent. According to Astle,\ ink of this colour was frequently used in Latin manuscripts, though rarely found in charters ; but his remarks apply to later ages. The same authority stated that blue and yellow ink seldom occurred in old manuscripts, and that he knew of no instance of the latter being used later than 1200 A.D. * Ovid, Trist., I. i. 7. t Cod. Justin. I. [23], I Origin of Writinff, 1803, p. 209. COLOURED WRITING INKS - 113 Gold and silver inks were used by both Greek and Roman Emperors at later periods. These probably consisted of the finely divided metals incorporated with some adhesive medium, such as gum. Metallic writing of this character was sometimes burnished or coated with wax. Wecker (De Secretis, 1582) gives details of the compo- sition of ink of different colours, and refers to gold and silver inks. In the work of, Canneparius (1660), to which we have frequently referred, various formulas for coloured inks appear, such as solution of verdigris in vinegar for green ink, &c. The use of both indigo and logwood as dye-stuffs was forbidden in England in the reign of Elizabeth, and the Act was not repealed until the reign of Charles II. (vide supra). After that they gradually came into use as con- stituents of writing inks, and are now widely employed in the manufacture of black writing inks. Inks can be made of any desired tint, for a variety of pigments and dye-stuffs are at the manufacturers' dis- posal ; and the discovery of the coal-tar colours, to which the main credit is due to Perkin, has increased their resources almost indefinitely, for they are now able to match any ray of the solar spectrum. Before the time of alizarine and aniline (1858) the maker of coloured inks had recourse to the various vegetable and mineral pro- ducts which have been used from time immemorial for dyeing fabrics. Thus for red he would employ Brazil wood and cochineal, the latter having a disadvantage in the circumstance that caustic ammonia in considerable quantity is necessary to dissolve it, so that it shall remain in solution and now freely from the pen. But cochineal or carmine inks were expensive, and they, together with Brazil wood and tin-salt red inks, ceased to be manufac- tured to any great extent when the more brilliant coal- tar colours became available. Older Formulae for Coloured Writing Inks. The following recipes, taken from various sources, are typical of the kind of coloured ink prepared from pigments other than aniline dye-stuffs : H 114 INKS AND THEIR MANUFACTURE Eed Inks. (i) Cochineal, i oz. ; ammonia, i oz. ; and water, i quart ; the infusion being decanted after three days, diluted with water to the required intensity of colour, and a little antiseptic added. (ii) Brazil wood (powdered), i Ib. ; acetic acid (5 per cent, strength), i gallon, boiled until of sufficient colour, and the extract mixed with 8 oz. of gum, 8 oz. of alum, and a little antiseptic. Green Inks. (i) Cream of tartar, 1 part ; verdigris, 2 parts, boiled with 8 parts of water. (ii) Copper acetate, i oz. in i pint of water, (iii) Potassium chromate, 10 parts; hydrochloric acid, 10 parts; alcohol, 10 parts; water, 30 parts. Neutralised with sodium carbonate after reduction to the chromic salt, mixed with 10 parts of gum, and decanted ( Winckler). (iv) Indigo ink mixed with a 1.25 per cent, solution of picric acid (Stein). Blue Inks. (i) Freshly precipitated Prussian blue triturated with a tenth of its weight of oxalic acid, and water gradually added. (ii) Indigo carmine, 10 parts; gum, 5 parts; in 75 parts of water. Purple Ink. Infusion of logwood mixed with copper acetate, gum arabic and alum (Normandy). Violet Ink. Indigo blue ink mixed with cochineal ink. Yellow Inks. (i) A decoction of 25 parts of .^Persian berries (Rhamnus amygdalinus, &c.) in 100 parts of a 3 per cent, solution of alum mixed with 4 parts of gum. (ii) A solution of gamboge in alcohol (10 : 10) mixed with 5 parts of gum and diluted to 30 parts with water, (iii) A 10 per cent, solution of picric acid containing 2 per cent, of gum. In a Report to the Science and Art Department in i888/ Dr. Russell and Sir W. Abney summarised the results of their experiments on the stability of various water-colour pigments exposed for two years to the action of light and dry air. In each case a wash of 8 tints was applied to paper of the same size and quality, and the slips enclosed in glass cylinders, so arranged that free circulation of air took place whilst dust was excluded. COLOURED WRITING INKS 115 In the following list, based on these results, the different pigments are arranged in the order of their instability, whilst those showing a distinct change in hue or depth of colour are marked with an asterisk: Carmine,* ciimson lake,* purple madder,* scarlet lake,* Naples yellow,* olive green,* indigo,* brown madder,* gamboge,* vandyke brown,* Indian yellow,* cadmium yellow, sepia,* aureolin, roc- madder, permanent blue, Antwerp blue, madder lake, vermilion, emerald green, burnt umber, yellow ochre, chrome yellow, raw sienna, Indian red, Venetian red, burnt sienna, chromium oxide, Prussian blue, cobalt, ultramarine ash. When exposed to the action of moist air very few of the pigments remained unaffected, and none of those of organic origin, whilst the Prussian and Antwerp blues were completely destroyed. Experiments were also made in which the washes of pigment were protected from the action of atmospheric oxygen and moisture. The vermilion turned black, but this change was attributed to a physical and not to chemi- cal alteration. In a later series of experiments, to quote Sir W. Abney's words : " We took exactly similar tubes, dried the papers very carefully indeed, dried the tube, inserted the papers, put a Sprengel pump to work, and made a vacuum, and then when the vacuum was very complete, sealed off the top and exposed them." Under these stringent conditions only five colours were acted upon in the very least, and the amount of change was almost imperceptible. The five that were changed were vermilion, raw sienna, Prussian blue, purple madder, and sepia. COLOURED ANILINE INKS. Aniline Inks. Any conceivable kind of red tint, from magenta to the most brilliant scarlet, can now be obtained from the makers of coal-tar colours, and the writers have to acknowledge their indebtedness to the Badisclie Company, Ltd., for their courtesy in supplying full infor- mation as to the most suitable dye-stuffs, and sending 116 INKS AND THEIR MANUFACTURE specimen samples for testing. Some of these colours are more fitted for ink manufacture than others, those which are the more readily soluble in water being naturally the best. The red, known as eosine, which was discovered by Caro in 1874, was early recognised as a valuable material for the purpose, and appears to be more used than any other dye-stuff. In aqueous solution eosine is subject to the formation of a fungoid growth, so that a small quantity of an antiseptic must be added to the ink to keep it in good condition ; otherwise its rich colour is liable to change. As in the case of red inks, the manufacturer has a number of different tints of blue to choose from in the coal- tar colours. They are most tempting substances to employ, for the suitable ones form a true solution with water, and as a general rule nothing beyond water is required to convert them into serviceable inks. The first aniline colour which was tried for the purpose was Hofmann's violet, discovered in 1863, a dye-stuff of such high tinctorial value that an ink composed of one part of it in 200 parts of water not only gives a most vivid colour, but will afford by pressure three or four good copies. Fugitiveness of Aniline Inks. Instances of the instability of ink of this character, which is largely employed for typewriting, are given in a letter to the Scientific American (Ap. 18, 1903). The writer states that typewritten documents, after being stored for six months in a slightly damp place, were illegible, with the exception of the gall ink signatures. In another case a letter-book was wetted with water used to extinguish a fire, and the signatures (in gall ink) were all that remained of IOO pages of correspondence. It has also been shown by Cross and Bevan * that all the aniline colours when dyed on fabrics fade more or less on exposure to sunlight, whilst eosine and methylene blue are specially fugitive. The following table of certain aniline dye-stuffs suitable for coloured writing inks is based on information supplied by the Badische Company, Ltd. See also Aniline Black Inks, chap. v. ; Copying Inks, chap. xii. ; and Ink Powders, chap. xvi. * Journ. Soc. Arts, 1891, xxxix. 152. COLOURED WRITING INKS 117 Aniline Dye-stuffs suitable for Writing Inks. Colour. Trade name. Scientific. Bed. Eosine, erythrosine and phloxine. Ponceau scarlet. Cotton scarlet. Alkali salts of bromine and iodine compounds of fluorescein and of dichlorfiuorescein. Alkali salts of xylidin-azo- and cumidiu - azo - naphtholdisulphonic acids. Sodium salt of amido-azo-benzol-azo- naphtholdisulphonic acid. Green. Neptune green S.G. Light green S.F. (yellowish). Light green S.F. (bluish). Diamond green G. &B. A triphenylmethane dye-stuff. Sodium salt of diethyl-dibenzyl-dia- midotriphenyl carbinol trisulpho- nic acid. Sodium salt of the dimethyl com- pound. Salts of tetra-ethyl- and tetra-methyl- di - para - amido - triphenyl carbi- drides. Blue. Indigo carmine. Soluble blue T. Indigotin sodium disulphonate. Salts of triphenylrosanilin and tri- phenylpararosanilin - trisulphonic aeids. Violet. Acid violet 46. L. Sodium salt of tetraethyldibenzyl- pararosanilin-disulphonic acid. Yellow. Fast yellow. Tartrazine. Mixtures of the sodium salts of amidoazobenzoldisulphonic, mono- sulphonic, and amido-azotoluoldi- sulphonic acids. Sodium salt of diphenyl-parasulpho- nic acid or of azo-dioxytartaric acid. The usual proportions that we have found to yield suit- able solutions for writing inks are about I gramme in 50 to 80 c.c. of water, according to the tinctorial power of the particular dye-stuff. The inks thus made are very fluid, and in this respect particularly suitable for stylo- 118 INKS AND THEIR MANUFACTURE graphic and other descriptions of fountain pens. If a suitable dye-stuff be used there will be no precipitation, and therefore no suspension of particles in the liquid. There is, therefore, no need to add gum to inks of this description ; indeed, such an addition would tend to counteract one of their most valuable properties their fluidity. It is neces- sary to mention this point, because we have found many published formulae for aniline inks in which sugar or gum is erroneously included as a necessary constituent. Patent Coloured Inks. Reade, in his patent (No. 1 1,474, 1846), claimed the use of inks containing a soluble Prussian blue," prepared in a specified manner, and of a red ink prepared from cochineal. A lake of cochineal extract and alum dissolved in ammonia solution was also claimed by Wood in 1885 (Eng. Pat., No. 1676). The use of aniline dyes was first claimed in this country by Crocin 1861 (Eng. Pat., No. 2972), and in the following year by Annaud (Eng. Pat., No. 675, 1862). Pigments from aniline waste were proposed as the source of writing inks by de la Rue (Eng. Pat., No. 2235, 1862), whilst aniline dye-stuffs were again patented by Jefferies in 1 879 (Eng. Pat., No. 3391). For or/her patents in which coloured pigments are claimed see Copying Inks, Sympa- thetic Inks and Ink Powders. CHAPTER VII. EXAMINATION OF WKITING INKS. CONTENTS. Fluidity of ink Penetration through paper- Stickiness of writing Composition of commercial inks Schluttig and Neumann's stripe test Acidity, action on steel pens Stability on keeping Examination of handwriting Old manuscripts Palimpsests Forged handwriting Bleaching agents Differentiation of writing done with different inks Photographic methods Mechanical erasure Chemical removal of writing Destruction of sizing Alterations and additions to writing Photographic distinction between dif- ferent inks. THE number of substances entering into the composition of ink is very large ; but since the influence of many of these on the permanency of the writing is unknown, a full analysis of an ink would not afford much information, at any rate as compared with the results of practical tests. One of the rules of the German Versuchsamt is that an ink for documentary purposes shall contain a certain mini- mum proportion of tannic or gallic acids derived from galls (vide infra). It has, however, been shown conclusively by Schluttig and Neumann, who submitted samples of dif- ferent inks to the office for examination, that the chemical tests employed are quite incapable of identifying an ink prepared (e.g., from chestnut bark), and they therefore contend that an ink prepared from tannic or gallic acid derived from any source should be permitted in inks of Class I. The requirements of a good ink are : (i) It must yield permanent writing which becomes black within the course of a few days ; (2) It must flow readily from the pen, and penetrate well into the fibres of the paper, without passing right through the paper; (3) It must not gelatinise or become mouldy in the ink-pots ; (4) It should have a 120 INKS AND THEIR MANUFACTUEE minimum corrosive action upon steel pens ; (5) The writing must not be sticky (except in the case of some copying 1 inks). Fluidity of Ink. Schluttig and Neumann recommend their stripe test (p. 121) as a means of determining the capacity of an ink to flow readily from the pen without spreading too freely on the paper. At the point where the glass pipette touches the paper in their test, an oval head to the stripe is formed, whilst the remainder of the stripe is nearly as wide. Of 8 1 inks examined by Scliluttig and Neumann, the majority gave the same results as the typical ink, whilst the copying inks yielded somewhat narrower stripes. Inks flowing too readily, however, produced a much wider head, whilst the lower part of the stripe was contracted to a narrower band than the others. We have found a simple viscosimeter, consisting of a 5oc.c. pipette to give concordant results when used in the follow- ing manner. The pipette is standardised on distilled water at 15.5 C., the time required for the liquid to run down to a given mark on the lower stem being taken as unity. The ink is then brought to the same temperature, and the time taken for the same volume to run out determined in the same way. Thus, in a pipette from which the water ran out in 40 seconds, we found that different writing inks required from 42 to 55 seconds, whilst copying inks in some cases required 70 seconds. Penetration through Paper. This is best determined by a practical test on standard paper, under the same con- ditions as used with the typical standard ink. The ink should penetrate into the fibres, but should not come through the paper. ; Stickiness of Writing. Here, again, the best results are obtained by comparison with a typicalink, as recom- mended by Scliluttig and Neumann. Composition of Commercial Inks. The following table gives the results of partial analyses of certain well-known commercial inks. It will be seen that the blue-black inks of the three different manufacturers are very similar in character. EXAMINATION OF WKITING INKS 121 Efflux Ink. Specific gravity Water. Total solid Ash. Iron. viscosity (Water at r * ,- f^ matter. at 15.5 C. I 5-5 c - = 40 seconds). Per cent. Per cent. Per cent. Per cent. Seconds. i. Blue-black I. 96.21 3-79 0.764 0.32 2. I. 1.0206 96.42 3-58 0.767 0.32 5 2 3- ' II. 1.0214 96.44 3-56 0.90 027 49 4. Chrome ink 98.70 1.22 0.26 42 5. "Japan" ink . 1.0413 92.74 7.26 2.18 0.84 55 6. Blue-black III. 1.0141 97-44 2.56 0.58 0.23 43 7. Black ink (log-- AVOOd) . 1.0115 97.86 2.14 1.04 0.14 43 A qualitative test that sometimes enables one to dis- tinguish between inks of different makers is their behaviour on titration with a saturated solution of bromine water. Thus, the blue-black ink I. became first dirty grey, and then greyish-black ; whilst No. II. first changed to violet and became turbid brown on. the further addition of the reagent ; and No. III. first became violet and then dirty green, the liquid remaining clear all the time. This last ink was also characterised by the ash being very difficult to burn white, and being then extremely insoluble in hydrochloric acid. The presence of logwood in an ink is readily identified by the colour changing to bright red on the addition of hydrochloric acid (see p. 104). When indigo is also present the hydrochloric acid gives a purple colora- tion. Indigo increases the stability of an ink towards bleaching agents such as bromine water. An ordinary iron gall ink is rapidly decolorised on tbe addition of strong hydro- chloric acid, but if indigo be present the liquid remains blue, even after being boiled with the reagent. Schluttig and Neumann's Stripe Test. For comparison in their colorimetric method, Schluttig and Neumann make use of a standard ink containing the following con- stituents : Gallotannic acid, 23.4 grms. ; gallic acid, 7.7 grms. ; gum, 10 grms. ; hydrochloric acid, 2.5 grrns. ; and ferrous sulphate, 30 grms. in a litre of water. After 122 INKS AND THEIR MANUFACTURE standing for not less than four days, this ink is decanted and kept in a well-corked bottle. In testing an unknown ink, 10 to 15 c.c. are compared with the standard ink, and should there be a difference in shade a small quantity of one or more suitable aniline dye- stuffs is added to the standard, so as to make the colours match. The apparatus required consists of a frame over which is tightly stretched a piece of best writing paper. This frame is fixed at an angle of 45, and the unknown ink is allowed to run down this from a special pipette delivering exactly 0.6 c.c. A similar stain is made with the ink after dilution with an equal quantity of water, and also with the standard ink before and after dilution. When the ink is dry the paper is set aside in a place exposed to light and air, and after eight days the stains are compared with the standard stains with regard to their colour, whilst their shape also gives an idea of the fluidity of the original ink. The paper is also to be cut in strips horizontally and one piece immersed in water, another in alcohol of 85 per cent, strength, and a third in alcohol 50 per cent, strength for several days. If the stripe of any particular ink becomes paler than that of the typical ink under these conditions, ScJiluttig and Neumann conclude that the former is either too poor in gallic or tannic acid, or contains too much acid. The objects of repeating the test after dilution are that the added aniline colours have less disturbing influence in the diluted ink, that differences of intensity are more pro- nounced, and that differences in the breadth of the stripes given by copying inks are reduced to a minimum. Schluttig and Neumann do not contend that all inks must have the composition of their typical preparation, but assert that while containing at least 0.6 per cent, of iron, they must give equally satisfactory results in the different tests. In place of the somewhat complicated apparatus devised by Schluttig and Neumann, we have found that sheets of Bristol board placed on a wooden stand fixed at an angle of 45 give satisfactory and concordant results in this test. EXAMINATION OF WRITING INKS 123 Acidity : Action on Steel Pens. It is not an easy matter to determine the amount of free acid in an ink by titration on account of the colour of the ink itself masking the change of colour of the indicator. It is possible, however, to obtain approximately correct results by diluting the ink with a very large volume of water and titrating with standard alkali, using phenol-phthalein as indicator. Thus in the case of a well-known blue-black ink we found that the acidity of 10 c.c. of the ink corre- sponded to 3. 1 c.c. of normal alkali. Another practical test is to immerse a steel pen in the ink for a given period, and to determine the loss in weight. Thus in the case of the ink referred to above we found that a pen had lost 5.18 percent, of its weight after being kept in 10 c.c. of the ink for a month, whilst the ink itself had become nearly solid. It is thus evident that the contention of Schluttig and Neumann * that free acid does not act upon metals in the presence of tannin is not justified by the results of ex- periments. Schluttig and Neumann consider that it is not possible to fix a maximum amount of acidity. They recommend their stripe test as the best means of determining whether too much acid is present, since ink darkens more slowly the greater the proportion of acid. Thus, if the stripes are as intense a black in as short a time as those given by their standard ink. they consider the amount of acid as not too great. Stability on Keeping. A good ink will frequently keep as long as a year without throwing down any in- soluble deposit on the sides of a vessel provided air be excluded. At the same time, if a sample bottle contain some deposit, the ink is not necessarily of bad quality, since this may be due to changes of temperature. Should there be a pellicle, however, in a sample bottle the ink should be rejected as inferior. Schluttig and Neumann have devised the following test for determining the stability of an ink in a comparatively short time : The bottle containing the ink is allowed to * loc. cit. 124 INKS AND THEIR MANUFACTURE stand for three days at a temperature of 10 to 15 C., after which 50 c.c. are withdrawn from the centre by means of a pipette without shaking the contents. This is filtered and 25 c.c. of the filtrate placed in a cylindrical glass vessel about 185 mm. high and 72 mm. in diameter, the mouth of which is then covered with paper to exclude dust. ^ ';.;"- Sclduttig and Neumann's typical ink when thus tested remained unchanged for three weeks, after which a pellicle began to form on the surface, and small flecks to separate Oilt- 'This ink was obviously purer than ordinary com- mercial inks, many of which, however, remained unchanged for fourteen days or longer, and in Scliluttig and Neu- mann's opinion this period should be fixed as the minimum keeping time for an ink of the first class under these conditions. EXAMINATION OF HANDWRITING. Old Manuscripts. We have already pointed out that, although in many ancient manuscripts the writing is as distinct as when first written, there are also numerous cases in which the characters have faded to such an extent as to be almost illegible. Numerous methods have been suggested for restoring tte intensity of the original writing, but many of these are open to the objection that they injure the surface of the material. The oldest and best known of these methods was to sponge the writing with an infusion of galls, the tannin of which would once more combine with the iron left in the paper, thus forming a fresh ink. This method was described in the book of Canneparius in 1660. A much more reliable method was that first proposed by Blagden in 1787, which was based on the formation of a blue compound by the action of a solution of potassium ferrocyanide and dilute hydrochloric acid on the residual iyon in the paper. This affords an easy means of dis- tinguishing between carbon inks and iron gall inks, and Blagden was thus able to show that the writing on certain EXAMINATION OF WRITING INKS 125 vellum manuscripts of the ninth and fifteenth centuries consisted of iron ink. Lenher's method of applying this process of restoration is to immerse the paper for a few seconds in one per cent, pure hydrochloric acid, and to allow it to dry spontaneously. The writing is then dusted over with powdered potassium ferrocyanide, and covered with a glass plate on which is placed a weight. After being left for a few hours the paper is thoroughly dried, and the excess of ferrocyanide removed by means of a soft brush. A more modern and less drastic method is to use a solution of ammonium hydrosulphide, which is applied foV a few seconds until the writing becomes darker, and is then sponged off as rapidly as possible after the desired effect is obtained. In this method the basic ferric salt into which the ink in the paper had decomposed is converted into the black ferrous sulphide. Writing thus restored may speedily be reoxidised so as to nearly disappear again, and the method can only be regarded as a temporary expedient. Lenher * has devised the following means of keeping writing thus restored for a longer period : The damped paper is sup- ported on a frame of threads fixed half way up in a shallow box (four inches deep), whilst ammonium sul- phide is placed in a small dish beneath. The box is closed by a glass cover, and after a short time the vapours of ammonium sulphide act on the writing, which becomes first brown and then black, and retains its intensity so long as the manuscript is left in the box. Palimpsests. The name palimpsest is derived from the Greek words 7ra\iv = again, and Tro-ftrroc = rubbed. It is applied to old manuscripts, the parchment of which had been previously used for a similar purpose. This practice was doubtless due to the cost of new material. The first writing on the skins was obliterated by means of pumice or some other abrading substance, but the mechanical action was insufficient to remove characters, possibly written three or four centuries earlier. In cases where the iron constituent of the ink had sunk deeply into * Die T'uiten Fabrikation, p 144. 126 INKS AND THEIR MANUFACTURE the vellum, it would be almost impossible completely to eliminate it, and the use of the reagents described above would reveal the original characters. Morides' Process. This consists of softening the skin by leaving it in distilled water, then treating it with a one per cent, solution of oxalic acid to bring the residual iron in the paper into a soluble state, again rinsing it in water and immersing it in a one per cent, solution of gallic acid, which will again form ink with the iron. Finally, the parchment is washed in water and pressed between blot- ting paper. Care must be taken to avoid an excess of oxalic acid, which might completely destroy the writing, and the method has the further drawback that the parch- ment is sometimes blackened all over by the gallic acid. Lenher advocates exposing the parchment to steam and then to acetic acid vapours before applying the gallic acid. It may also be noted here that photography affords a ready and most efficient means of deciphering such parti- ally obliterated writing, the pale yellow colour of the iron oxide from the old ink appearing black in a photographic copy. FORGED HANDWRITING. The police reports abundantly prove that the crime of forgery frequently engages the attention of our magis- trates. This is not surprising when we remember that the means are in the hands of every one, and that bankers' cheques are a common medium of exchange. The usual evidence called by the prosecution is that of the hand- writing expert, who bases his opinion on the form and peculiarities of the caligraphy. We are inclined to think that much more stress might in many cases be laid upon the chemical aspect of the question. What we mean is that in a case where additional letters or figures have been added to, say, a banker's draft, it would be comparative!}* easy to ascertain whether the interpolated characters had or had not been written with the same ink as that used for the body of the document. When there is any objec- tion to bringing chemical reagents in contact with the original paper, the camera can be employed in the way EXAMINATION OF WRITING INKS 127 first brought into prominence by Dr. Paul Jeserich, which we describe at length in a subsequent page. Bleaching Agents. It is possible to remove completely all traces of an ink stain (not containing carbon) by the use of suitable bleaching reagents, among which may be mentioned solutions of chlorine, bromine, acidified bleach- ing powder, &c. Traill, in the course of his research into the permanency of inks, tested the effect of various reagents on the writing done with ordinary iron gall ink, and classified them into the following groups : I. Those completely effacing the writing : Solutions of chlorine ; chloride of lime with weak acid ; antimony chloride ; dilute aqua regia ; and oxalic acid. v IL Those effacing the writing to a large extent : Dilute nitric, sulphuric, and hydrochloric acids. III. Those rendering the writing faint : Solutions of potassium and sodium hydroxides and ammonia. To the first group we may add the following sub- stances : Bromine, sulphur dioxide, sodium nitrite with hydrochloric acid, and citric acid ; whilst hydrogen per- oxide solution and potassium bisulphate speedily render the writing very faint. As has been already mentioned, the addition of indigo to iron gall ink renders it much more resistant to the action of these reagents. Differentiation of Writing done with Different Inks. If writing from various sources be subjected to a system- atic series of tests differences in that done with different inks, or even at different periods, will generally be observed. For instance, in the case of the reagents used by Traill (supra), fifty envelopes were tested, and th& writing on no two of them gave identical results. Robertson and Hoffmann * employed the following reagents for distinguishing between writing done with different kinds of ink. In each case a feather was dipped in the reagent and a note taken of any change occurring at the junction of the ink and paper. * Eisner, Die Pi-axis der Chemiker, p. 598. 128 , INKS AND THEIE MANUFACTURE Resorciu ink. Bright red. Bleached. oj 02 g 03 1 1 d-Sd 'o o | D | s 1 ^1^ 03 Unaltered. Brown. M 1 . 'd tf ~| I 1* a- flJ 2 i-Q 3 S IllfS 1 02 ^ S >, pQ 02 M "Is a p ^ fcJD S pQ ^ -IS > 02 H> Q a g -d ; i\ d ,2-d '. fcJO r2 iD l _^* r ^ < 0-3 j^ O "I? 03 o as a ft a rO Pi 5 1 "5^ p Q ^ = M |l 1 Is a P -< be S >r& ^ ^ 3 S |. 'It p. o 1 'Hi p g f P 44 1 ^03 it'L *! 5 5 o 5 3 PH cl 1 P 02 'OJ H O -d ^ d -d- . -- * ^J 03 S ^ 9^ d 03 1! 1 * JO . 3, O-i 1 **! II 73 i 1 1 PP ,2 PQ g 1 "03 a 4 ^d O U 1 1 fi si a to i 6 6 6 ^ 'o ^ ppp .g>j s j* pq eS pq ^^P p S 1-1 . 3 o "^ ** 03 PH ^ o^~ o "o O oT M co- ^.ai^ e3 C 13 N I 1 1 is .2 1 I Hfl | ll Jf XI 03" O r ^ H C3 j^ O C^ ^ -2 QC r5 "^ ^ r^ ._ Q PH .2 o S 1 - 1 o w -n' 73 ^ -g 'a rH ^3 s F"^ -I S ^ 1 a t& "3 8 Ft" I is O2 O2 02 PH '5'i i + - g f d ^ 1-J 4"