THE PHOSPHOR BRONZE GO. "I The I ►Strap? $il IN SIX CLASSES. OF VARYING STRFNRTH Aun ti cp.tdip/ii mun.. ds of E." sentric Rods, afes, ire. R( f RflNKLiN Institute Li^R^R^ Class BookC jS 33 Accession 52j4: J^ " WHITE ANT " METAL, cheaper than Babbitt's and equal to some Antifriction Metals at far higher prices. Please specify the manufacture of the Phosphor Bronze Company, Limited, to prevent imposition and error. 1 THE PHOSPHOR BRONZE GO. LIMITED, 87 Sumner Street, Southwark, London, S.E. Sole Makers of "COG WHEEL" and "PHOSPHOR The Best and most durable Alloys for Slide Valves, Bearings, Bushes, Eccentric Straps, and other parts of Machinery exposed to friction and wear ; Pump Rods, Pumps, Piston Rings, Pinions, Worm Wheels, etc. (see Testimonials). A Malleable and Inoxidizable Quality in the form of Plates, Bars, Wire, Tubes, Rods and Strips. GREAT TENSILE STRENGTH, RESISTANCE TO CORROSION, AND PRACTICAL INDESTRUCTIBILITY. "DURO METAL" (registered trade mark). Alloy B, specially adapted for BEARINGS for HOT-NEOK ROLLS of IRONWORKS, TIN-PLATE MILLS, etc. Castings in PHOSPHOR BRONZE, DURO METAL, GUN METAL and BRASS, to Patterns or Drawings, in the rough or Machined if required.' the Original "VULCAN" Brands of BRONZE." Please specify the manufacture of the Phosphor Bronze Company, Limited, to prevent imposition and error. 2 A TEXT-BOOK OF PAPER-MAKING 1^ A TEXT-BOOK OF PAPER-MAKING BY C. F. CEOSS AND E. J. BEVAN SECOND EDITION ?Lontion: E. & F. N. SPON, Ltd, 125 STEAND SPON & CHAMBERLAIN, 12 COETLANDT STREET 1900 PREFACE TO THE SECOND EDITION. In preparing this present Edition we have adhered strictly to the original plan and scope of the work, which is that of a text-book of principles, and not either an exliaustive treatise or a minutely descriptive manual of the manufac- turing art. We have endeavoured at the same time to bring the matter of the bouk to the level of later developments, some of which are of such importance as to necessitate the re-writing of certain sections, notably those dealing with the Chemistry of Cellulose and the operations of Sizing, Loading and Colouring. Our aim is to present the subject to the reader according to its scientidc perspective ; to furnish a guide for the student or apprentice in acquiring his practical experience to the best advantage, and for those who have a working experience of Paper-making, in reviewing either for pleasure or profit, the multitude of facts to be observed in the daily routine of the mill. Vi PAPER-MAKING. We wish to thank our friend Mr. A. D. Lit- rLEj of Boston, Mass., for permission to make use of his work on ' The Chemistry of Paper-Making ' ; onr thanks are also due to the Publishers of this book, the Howard Lockwood Publishing Company, for the right to reproduce portions of the text. The Engineering firms of Messrs. Bertram, Ltd., Jas. Bertram and Sons, Ltd., Masson and Scott, Ltd., Mather and Platt, Ltd., have shown their usual courtesy in supply- ing particulars and drawings of machinery. To our paper-making friends, more particularly Mr. 'J'hos. Tait and Mr. C. M. King, we are obliged for suggestions and criticisms. CONTENTS. PAGE The Authors are especially indebted to Messrs. Longmans Green and Co., Paternoster How, London, B.C., for the' privilege of reprodnoing portions of the text of the wort on 'Cellnlose' (Cross and Bevan, 1806), of which they are Publishers. VIII. Sizing, Loading, CoLOURrNG, Etc. . . . .181 IX. Paper Machines, Hanb-made Papei: . . . .210 X. Calendering, Cdtting, Etc. 233 XI. Caubtio Soda, Recovered Soda, Etc 244 XII. Paper Testing 263 vi PAPER-MAKING. "We wish to thank our friend Mr. A. D. Littlk, of Boston, Mass., for permission to make use of his work on ' The Chemistry of Paper-Making ' ; our thanks are also due to the Publishers of this book, the Howard Lockwood Publishing Company, for the right to reproduce portions of the text. The Engineering firms of Messrs. Bertram, Ltd., Jas. Bertram and Sons, Ltd., Masson and Scott, Ltd., Mather and Platt, Ltd., have shown their usual courtpssv ^■n cut^^^u. CONTENTS. CHAPTER p^Qj, Intboductoby 1 I. Cellulose 5 II. Physical Stuuctdre of Fibres 78 III. Scheme for the Diagnosis and Chemical Analysis of Plant Substances ....... 90 IV. An Account of the Chemical and Physical Charac- teristics OF THB Principal Raw Materials . . 95 V. Special Teeatmbnt op Various Fibres ; Boilers, Boiling Processes, Etc. . . . . , .111 VI. Bleaching 253 VII. Beating .166 VIII. Sizing, Loading, Colouring, Etc. . . . .181 IX. Papeb Machines, Hanc-made Papeu .... 210 X. Calendering, Cutting, Etc. ..... 2.S3 XI. Caustic Soda, Recovered Soda, Etc 244 XII. Paper Testing 263 I ■ ■ yiii PAPEK-MAKING. PAGK CHAPTER nrjn Xm. Gekeral Chemical Analysis FOR PArER-MAKERS . . m XIV. Site fob Paper-mill-Water SfPPLY, Water Pubi- FiCATioN, Etc. . • • • * XV. Action of Cvpbamm .kium on Cellulose-Prepae-^tion OF " Willesden '' Paper XVI. Statistics . XVII. Bibliography Index . 299 . 311 . 313 PAPEE-MAKING. INTEODUCTORY. Paper-making is essentially a mechanical art, consisting as it does in the production of a continuous web or fabric by- aggregating together structural units of relatively minute dimensions. These, the essential components of paper, are the vegetable fibres, first isolated from the fibrous raw materials by opera- tions of both mechanical and chemical nature, and then further resolved or broken up by anechanical treatment. But, though the actual paper-making processes are of the mechanical and physical order, they involve auxiliary chemical processes of fundamental importance. Moreover, there is a special chemistry of the fibrous components and of the auxiliary agents used in the making and finishing of papers, to know which is a necessary equipment of the paper- maker. In presenting this brief outline of first principles, we cannot overrate the importance of a thorough grasp of the composition and constitution of the plant fibres, as the neces- sary foundation for the intelligent conduct of paper-making, and to this subject we will at once proceed. Careful study of a mature flowering plant will show that it is made up of structural elements of two kinds, viz. fibres and cells, which, to use a rough parallel, we may liken in func- tion to the bricks and mortar of a house. It is the former which admit of the many extended uses with which we aro 2 PAPER-MAKING. familiar in the arts of spinning and weaving, and which con- stitute the fabrics which are the most indispensable to our civilised life. Fur the most part, as we know, fibres and cells are aggregated together into compound tissues, and a pro- cess of separation is therefore a necessary preliminary to the utilisation of the former. The cotton fibre is the only im- portant exception to this general condition of distribution. Here we have the seed envelope or perisperm converted into a mass of fibres, and these, by a sjDontaneous process accom- panying the ripening, so isolated as to be immediately avail- able. Next in order, in point of simplicity of isolation, are those fibrous masses, or tissues, which, although components of complex structures, exhibit a greater cohesion of their constituent fibres than adhesion to the contiguous cellular tissues with which they go to build up the plant. Into such a tissue the " bast," or inner bark layer of shriibs and trees, more especially those of tropical and sub-tropical regions, frequently deA'elops ; and it is, in fact, this bast tissue, gra- duating in respect of cohesion of its constituent fibres, from a close network such as we have spoken of, to a collec- tion of individual fibres or fibre-bundles disposed in parallel series, which supplies the greater part of the more valuable of the textile and paper-making fibres ; we may instance flax, hemp and jute, each of which is the basis of an enormous industry. According to the degree of adhesion of the bast to the contiguous tissues, or, in another aspect, according to its lesser aggregate development, so is the difiiculty of iso- lation and the necessity of using processes auxiliary to the mechanical separation of the tissue. It is w^orthy of note here that the Japanese paper with which we are in these times so familiar, is prepared by the most primitive means from the bast of a mulberry (Brous- aonetia papyri/era) ; the isolated tissue, consisting of a close network of fibres, is simply cut and hammered to produce a surface of the requisite evenness, and the production of a web of paper is complete. In isolating the bast fibres em- ployed in the textile industries, a preliminary partial disinte- INTEODUCTOEY. 3 gration of the plant stem is brought about by the process of steeping or retting, by which the separation of fibre from fleshy or cellular tissue is much facilitated. Last in order of simplicity of distribution, we have the fibres known to the botanist as the fibro-vascular bundles of leaves and monocotyledonous stems, these bundles being irregularly distributed through the main cellular mass, and consequently, by reason of adhesion thereto, much more diffi- cult of isolation. For this and other reasons, more or less in correlation with natural function, we shall find this class of raw material lowest in value to the paper-maker. It is necessary at this stage to point out that the work of the paper-maker and that of the textile manufacturer are complementary one to the other, and the supply of fibrous raw material is correspondiugly divided : it may be said, indeed, that the paper industry subsists largely upon the rejecta of the textile manufactures. The working up of dis- continuous fibre elements into thread, which is the pui pose of the complicated operations of the spinner, is conditioned by the length and strength of these ultimate fibres. Paper- making, on the other hand, requires that the raw material shall be previously reduced to the condition of minute sub- division of the constituent fibres, and therefore can avail itself of fibrous raw material altogether valueless to the spinner, and of textile materials which from any cause have become of no • value as such. To the raw materials of the paper-maker, which we have briefly outlined above, we must therefore add, as a supplementary class, textiles of all kinds, such as rags,. I'ope and thread. During the latter part of this century, and in response to the enormously increased demands upon the sources of supply, there has come the exploitation of the fibrous woods for paper- making purposes. Having thus acquired a general idea of the sources of our raw materials, we must study more closely the substances themselves; and, first of all, we must investigate them as we should any other chemical substance, i.e. we must get to B 2 4 PAPER-MAKING. understand the nature and properties of the matter of which, the vegetable fibres are composed. "While these exhibit cer- tain variations, which are considerable, the substances pre- sent a sufficient chemical uniformity to warrant their being designated under a class name : this name is Cellulose. The prototype of the celluloses is the cotton fibre. CHAPTER I. CELLULOSE. Cellulose is the predominating constituent of plant tissues, and may be shortly described as the structural basis of the vegetable world. Constituting, as it does, the material frame- work or skeleton of the plant, or plant cell, this more per- manent function implies a corresponding resistance to the destructive agencies of the natural world : in other words, considered as a chemical individual, cellulose is extremely inert, or non-reactive. It is resistant to the hydrolysing action of alkalis and acids, to oxidants ; and, as a ' saturated ' carbon compound, has no tendency to combine directly with the halogen elements, e.g. chlorine and bromine. These main features of its negative chemical characteristics are mentioned thus early in explanation of the methods employed for its isolation in the laboratory. Cellulose never occurs in the plant in the free state, but always in admixture or combina- tion with other groups : members of the fatty series (' fat and wax ' constituents) ; the aromatic series (colouring matters, tannins) ; the pectic group of more or less oxidised, and there- fore acid, derivatives of the carbohydrates. These latter yield to the attack of one or other of the reagents towards which cellulose is inert ; and hence the following general method of removing these ' impurities ' in the form of soluble deriva- tives, and of isolating the cellulose as the resistant lesidue : (a) the fibrous raw material is boiled with a dilute solution of sodium hydrate (1-2 p.ct. NaOH), and, after thorough washing, is (&) exposed in the moist state to an atmosphere of chlorine gas ; (c) it is again treated with boiling alkaline solution. By such treatment of the majority of vegetable 6 PAPER-MAKING. tissues, the 'non-cellulose' constituents are removed, and a residue of cellulose obtained. A slight treatment with a bleaching agent, to remove residues of coloured impurities, and a final washing with alcohol and ether, completes the puri- fication, and the cellulose is obtained as a mass of 'ultimate fibres' of pure white colour, more or less translucent. Though purified by the removal of ' non-cellulose ' groups, such as above indicated, the residue of cellulose is not neces- sarily ' pure ' in the sense understood by the chemist, that is, it cannot be taken to represent a single homogeneous sub- stance. On the contrary, we shall show that the paper-makers' ' celluloses ' — obtained by a large variety of drastic treatments of fibrous raw materials — are mixtures of celluloses of different constitution. Cotton cellulose, however, when fully purified, may be re- garded as a chemically pure substance, and in setting forth the outline of the chemistry of cellulose, we shall at first confine ourselves to this typical representative of the group. Empirical Composition. — Cellulose is a compound of ■carbon, hydrogen and oxygen, united in the percentage pro- j)ortions — corresponding with the statistical formula CqH^qO^, by which also it is defined as a 'carbohydrate.' The above numbers represent the composition of the ' ash-free ' cellulose. All vegetable tissues contain inorganic or mineral consti- tuents, of which a certain proportion is retained by the cellu- lose, isolated as described, or by any of the processes practised on the large scale in the arts. The celluloses burn with a quiet luminous flame, leaving these inorganic constituents as an ash. In bleached cotton the average proportion of ash is 0-l-0-4p.ct. In the preparation of filter paper for chemical use it is important to reduce this impurity to a minimum, which is C H O 44-2 6-3 •19-5 CELLULOSE. 7 effected b}' treatment with hydrofluoric and other acids. 'Swedish ' filter paper contains 0-03-0 -05 p.ct. ash consti- tuents, representing about -jj^i^-u mgr. per sq. cm. of area ; and is the purest form of cellulose with which we can deal. Cellulose and Water. Cellulose Hydrates.— All vegetable structures in the air-dry condition retain a certain proportion of water — or hygroscopic moisture, as it is termed — which is readily driven off on heating, but regained on ex- posure to the atmof^phere under ordinary conditions. The mean percentage of this ' water of condition ' varies from 6 to 12 in the several celluloses: in any given cellulose variations of 1-2 p.ct. from the mean number follow the variations in the hygrometric condition of the atmosphere. The factor of ' normal moistui'e ' is of obvious importance in commercial dealings in celluloses. Thus, fur the ' wood pulps ' (celluloses), the 'standard moisture' commonly adopted is '10 p.ct.' — that is, 100 parts of the air-dry pulp yield on drying, at 100°, 90 parts dry cellulose. Conversely, in calculating from the basis of dry cellulose = a, to ' air-dry with 1 0 p.ct. moisture ' = h, it is clear that b = a. The proportion of water held by the celluloses in an atmosphere saturated with aqueous vapour is necessarily very much greater than in the ordinary atmosphere, partially saturated at the same temperature.* The celluloses or compound celluloses {supra) as they occur in the plant are characterised by a wide range of hydration phenomena. Plant tissues in the early stages of growth take the form of gelatinous hydrates, the proportion of water combining with the organic colloid in these hydrates being very large, e.g. 80 p.ct. of the weight of the hydrate. The re-hydration of the mature cellulofrcs to these forms is deter- mined by certain reagents as a stage in their conversion into the fully soluble form. Such proces-ses of solution of cellu- lose we proceed to consider. Solutions of Cellulose. — Cellulose is insoluble in water * See H. Muller, Pflanzenfaser. p. 3. 8 PAPER-MAKING. as in all simple solvents. In presence of certain metallic com- pounds, however, it combines with M'ater, passing, as above described, through the conditions of gelatinous hydrates, and finally disappearing to form a homogeneous viscous solution. Of such solvents of cellulose the simplest is (1) zinc chloride in concentrated aqueous solution (40 p.ct. ZnCls). The solution process requires the aid of heat (60-100°), and may be carried out as follows: 4-6 parts ZnCla are dissolved in 6-10 parts water, and 1 part cellulose (bleached cotton) stirred in till evenly moistened. The mixture is digested at first at 60-80°, when the cellulose is gelatinised ; the solution is completed by exposure to water-bath heat, stirring from time to time, and replacing the water which evaporates. In this way a homogeneous syrup is obtained. The solution is entirely decomposed by dilution, the cellulose being precipi- tated as a hydrate in combination wdth zinc oxide. On washing with hydrochloric acid a pure cellulose hydrate is obtained, the quantity recovered being approximately equal to the original cellulose taken. When precipitated by alcohol, a compound of cellulose and zinc oxide is obtained with 18-25 p.ct. ZnO, i.e. in the approximate molecular ratio 2CeHioO,ZnO. Technical Applications— (a) The precipitation of the syrupy solution by alcohol is of such a character as to permit of a continuous production of thread or film, the solution being 'squirted' under pressure from a fine glass ori6ce into the alcohol. The thread, when purified, is carbonised in closed vessels, to form the very resistant carbon constituting the fila- ment for incandescent electric lamps. (6) Vulcanised fibre is produced by treating a suitable paper (1 part) with a zinc chloride of 65-75° B (4 parts). When the constituent fibres are superficially gelatinised, the sheets are welded together under pressure into very compact macses. These are then purified, and subjected to further treatment to render them waterproof.* * German Patent 3181 (1S78): C. Hofmanii, Prakt. Handb. Papierf., CELLULOSE. 9 (2) Z';NC Chloride and Hydrochloric Acjd. — If theZnClo is dissolved in twice its weight of aqueous hydrochloric acid (40 p.ct. HCl), a solution is obtained which dissolves cellu- lose rapidly in the cold. If quickly diluted the cellulose is recovered with but little change, but on standing it is re- solved into products of lower molecular weight (dextrins, &c.) entirely soluble in water. The solution is a useful aid to investigations in the laboratory, but so iar has received no industrial applications. (3) Ammoniacal Cupric Oxide. — The solutions of the cuprammonium compounds generally, in presence of excess of ammonia, attack the celluloses rapidly in the cold, forming a series of gelatinous hydrates, passing ultimately into fully soluble forms. The solutions of the pure cupiammonium hydroxide are more active in producing these etfects than the solutions resulting from the decomposition of a copper salt with excess of ammonia. Two methods are in common use for the preparation of these solutions, which should con- tain — 10-15 p.ct. ammonia (NHj). 2 '0-2 5 p.ct. copper (as OuO). (1) Hydrated cupric oxide is prepared by precipitating a solution of cupric sulphate at 2 p.ct. strength with a slight excess of sodium hydrate, also in very dilute solution. The precipitate is washed till entirely free from alkali. The original solution in which the precipitation takes place, and the -water used in washing, should contain a small portion of glycerin, 0-05 to 0-10 p.ct. The washed precipitate is fully drained, and then mixed with a quantity of a 10 p.ct. solution of glycerin — in contact with which it may be preserved un- changed in stoppered bottles. Prior to dissolving in 15-20 p.ct. NH3 for use, the oxide may be washed free from glycerin, should the presence of the latter be objectionable.* (2) Metallic copper in the form of sheet or turnings is placed in a cylinder and covered with strong ammonia ; atmo- spheric air is caused to bubble through the column of liquid * Fassbender: Berl. Ber. 13, 1822. 10 PAPEK-MAKING. at a rate calculated to forty times the volume of the liquid used per hour. In about six hours a liquid of the requisite composition is obtained. Solutions of cellulose of 5-10 p.ct. (cellulose) strength, are readily prepared by digestion in the cold with 20-10 times its weight of the solution. The solutions are rather *ro])y ' and gelatinous than viscous. The cellulose is readily precipitated from the solutions : (a) by neutral dehydrating agents, such as alcohol, sodium chloride and other salts of the alkalis ; (6) by acids. In the latter case the cellulose is precipitated in the ' pure ' state, i.e. free from cupric oxide. It retains a large proportion of water of hydration. On drying hy heat, the gelatinous hydrate changes by molecular aggregation into compact horny masses. Technical Applications. — This property of gelatinising and dissolving cellulose has been taken advantage of in important industrial applications of the cuprammonium compounds. Vegetable textile fabrics and paper, passed through a bath of the cuprammonium h} droxide, are ' surfaced ' by the film of gelatinised cellulose, which retains the copper oxide (hydrate) in such a way that it dries to a bright malachite green colour, the ammonia of course escaping. By this treatment the fibres are further compacted together, and the fabric acquires a water-resistant character ; the presence of the copper oxide is also preservative against the attacks of mildew, insects, &c. Jf the fabrics are rolled or pressed together when in the gela- tinised condition, they become welded together on drying, and a variety of compound textures are produced in this way. The fabrics are sold in this country under the style of ' Willesden ' goods. Eecently, also, the solution has been applied to the production of an artificial thread of high lustre, a so-called * artificial silk.' Cellulose and Hydrolytic Agents. — Without attempt- ing a theory, or even an explanation, of the action of the solvents we have just described, there is one aspect which claims attention, and will be more clearly grasped from what follows in the present section : that is, that the cellulose CELLULOSE. 11 molecule contains OH groups of opposite function, giving it some of the charactei istics of the inorganic salts ; that it yields to the action of zinc chloride by reciprocal inter- action of its OH groups with those of the salt in solution, and that the dissolution of the cellulose is therefore due to a species of double salt formation. An incipient activity of this kind is manifested by cellulose in contact with highly dilute solutions of alkalis and acids, the active reagent being absorbed by the cellulose in per- ceptible degree. The amount though small is definite, and sufficient to allow of the establishment of a definite ratio of absorption from equivalent solutions of alkalis and acids. Thus, with typical members of the two groups, the molecular ratio of absorption is 10 NaOH : 3 HCL* The phenomenon has been more recently studied from the independent stand- point of thermal equilibrium. It has been shown that when pure cotton is plunged into dilute solutions of the acids and alkalis, liberation of heat takes place. The rise of tempera- ture was found to be slow, and, under the conditions chosen for the experiments, ceases after the lapse of seven to eight minutes. The following are typical results in calories per 100 grms. of cotton : — KOH. NaOH. HCl. H2SO4. Raw cotton 1-30 1-08 0-65 0-60 Bleached .... 2-27 2-20 0-65 0-58 L. Vignon. It would appear from these results that cellulose has the properties of a feeble acid, and of a yet feebler base. From the comparative insignificance of the ' affinities ' involved, it might be inferred that they may be neglected in practical and industrial operations. So far from this being the case * Mills. 12 PAPEK-MAKING. it must be remembered that tbe combination of cellulose with colouring matters, i.e. the dyeing properties of the fibre substance, are largely dependent upon a ]>lay of affinities of this particular order. So also the auxiliary processes of mordanting, in which the fibre absorbs both acids and basic oxides from dilute saline solutions ; these oxides, in combina- tion with the fibre substance, enabling it in turn to take up particular colouring matters from their solutions. Formerly it was much discussed as to whether dyeing phenomena were of the 'physical' or 'chemical' order. The modern view does not concern itself so much with definitions as to insist that the phenomena are moleculai'. This chemical aspect is prominent in the reciprocal play of acid and basic functions of the constituent groups of both fibre-substance and colouring matter (and mordant) ; the ' physical ' side is brought into evidence by certain properties of the fibres which are bound up with their minute structure, and which evidently play an important part in the absorption of reagents from solu- tion, viz. the phenomena of capillary transmission of liquids. Schonbein appears to have been the first to observe that strips of unsized paper of which one end is placed in an aqueous solution, e.g of a metallic salt, will absorb and transmit the water more rapidly than the dissolved salt, which is therefore ' filtered out' ; further, that to the various salts cellulose manifests varying degrees of resistance to transmission in solution. These phenomena have been further studied by Lloyd,* for metallic salts, more recently by E. Fischer and Schmidmer,| and by F. Goppelsroeder for various colouring matters ; J the results of their observations constituting the beginnings of a method of capillary analysis or separation. Without further discussing the phenomena from a theoretical standpoint, we may point out that they are of direct practical moment to the paper-maker: since, first, they have to be reckoned with in every one of his manu- facturing operations ; secondly, in one of the most important * Chem. News, 51, 81. t Liebig Ann., 272, 156. X Berl. Ber., 20, G04. CELLULOSE. 13 applications of paper, viz. for writing purposes, the penetra- tion of the paper by the ink, its indelible fixation, and the kind of press copy obtainable, are points largely affected or determined by such inter- actions as we have been considering. It will be remembered that we have followed up the various matters dealt with in this section from the initial observation of the behaviour of cellulose to typical hydrolytic agents in cold dilute solution. These are absorbed, as we have seen, to form what we may term contact compounds, and they are an index of the hydrolytic changes which are deter- mined by these compounds acting in more concentrated forms and at higher temperatures. Hydrolysis is essentially a process of resolution or decomposition : it is the loosening or undoing of a bond of union through combination with the elements of water. The agent which determines the change is the hydrolytic agent : and of such agents the most im- portant are the (a) acids and (&) alkalis on the one hand, and (c) on the other, a class of carbon compounds known as soluble or unorganised ferments, or by the more modern term Enzymes. Cellulose yields to hydrolysis of both types. (a) Acids. — The mineral acids of concentration equal to semi-normal at the boiling temperature, rapidly disintegrate the fibrous celluloses, as a consequence of molecular changes in the fibre-substance. The modified cellulose is brittle and pulverulent. Its composition is changed to that of a hydrate of the formula 2 • CgHioOg . HgO, and it is therefore termed hydro- or hydra-cellulose, the chemical properties of which are described in a later section. The time required for com- pleting this change varies with the temperature and concen- tration of the acid. The acid treatments of cellulose textiles which are necessary incidents of bleaching and dyeing opera- tions are carried out well within the limits of safety — for the most part in the cold (<20° C.) and with acids of less than 2 p.ct. strength (HCl, H2SO4). In dyeing operations requir- ing an acid bath and the boiling temperature, ' free ' mineral acids are as much as possible avoided, acetic acid being sub- stituted — an acid of low hydrolysing activity, and without 14 PAPEK-MAKING. sensible action on cotton. Paper is usually finished from the machine with a slightly acid reaction, but the utmost care is required to ensure the absence of ' free ' acid. (&) Alkalis. — To alkaline solutions of equivalent strength, e.g. solutions of caustic soda of 1—2 p.ct. Na20, cotton cellulose is extremely resistant, even at temperatures exceeding 100.° The principal operations in the process of bleaching cotton and linen textiles consist in drastic alkaline treatments of this kind, wherebj^ the non-cellulose constituents of the fiibres are hydrolysed to products soluble in the alkaline lye. The oxidation processes which follow, e.g. treatment with the hypochlorites, permanganates, &c., in dilute solutions, although they may be regarded as the bleaching processes proper, really accomplish very little beyond removing resi- dues or by-products of the alkaline treatment. It is also evident that resistance to alkaline treatment such as soaping, is a very important condition of the everyday u.ses of celluluse textiles. At higher concentration and temperature the typical cellulose is attacked by the alkaline hydrates and converted into soluble derivatives. Thus, purified cotton cellulose digested three times in succession with solutions of 3 p.ct. NagOj was found to lose * : — at 1 atm. pressure . . . . . 12-1 p.ct. 5 „ 15-4 „ 10 „ ..... 20-3 „ With solutions of 8 p.ct. Na^O under similar conditions the losses were 22-0, 28-0 and 69 -0 p.ct. The processes of isolating paper-makers' celluloses largely consist in severe alkaline treatments, the conditions of which require adjust- ment to secure the most complete removal of the non- cellulose constituents of the raw materials, with the mini- mum of action (hydrolysis) on the cellulose. A process of estimating cellulose in fibrous raw materials in the laboratory, based upon the action of alkaline hydrates *. H. Tauss, Joarn. Soc. Chem. Ind. 1889, 913 ; 1890,^883. CELLULOSE. 15 at elevated temperatures (150-180°), has been proposed by Lange * The process rests upon the assumption that the celluloses are not attacked under the severe conditions of alkaline treatment adopted. This assumption cannot be maintained in view of the results above cited ; the process has, moreover, been subjected to a careful critical investigation by Tollens, t who finds that it is subject to large and variable errors. (c) Ejtzymes.— A typical case of enzyme action is that of the conversion of the starches of cereals into fermentable sugars, in the operations (1) of malting the grain, (2) of ' mashing ' the malt with water, to the solution or wort in which the sugars formed from the starch are in a condition to yield readily to the attack of the yeast-cell, undergoing further hydrfdysis under the action of the yeast enzyme, and being finally broken down to alcohol and carbonic acid. The starches resemble the typical celluloses in the following par- ticulars : they have the same empirical formula ^[Ce H10O5] ; they are poly-anhydrides of hexose molecules, each hexose molecule losing water in the proportion of one molecule —thus m. 2C6Hi206-m. 2H2O— in condensing to form the anhydride ; the highly complex molecule of the latter takes up water under the influence of hydrolytic treatments in sucessive stages, but special to each — giving a similarly complex series of intermediate (i.e. partially hydrated) mole- cules, these breaking up finally (i.e. by complete hydration) to the typical hexose, dextrose. Starch, under the influence of the enzyme of malt, termed diastase, yields the series of dextrins, amylins, malto- dextrins, maltose, and finally dextrose. To convey some idea of the complications presented by the series, it may be noted that it is necessary to expand the original starch mole cule to o[Ci2H2oOio]2o- The first effect of hydrolysis is to split this into a resistant dextrin of the formula rC,oH„„0, 1 and a group ot amylms representing the remaining four- * Ztschr. Physiol. Chem., 14. t Suriiigar and Tollens, Ztschr. Angew. Chem. 1896, No. 23. 16 PAPEE-MAKING. fifths of the molecule which pass through further hydration stages represented Tby such formulEe as [CigH^oOjo]?! + H.,0, maltose groups heing successively formed and split off. Yeast, again, secretes an enzyme termed invertase, which hy- drates the crystallisable but still complex sugars such as maltose and cane sugar [both CigH^aOu] to the simple hexose C6Hj20e. No enzyme has yet been discovered which enables us to carry out a similar ' conversion ' or hydration of cellulose in the laboratory. That such actions take place in plant life has been abundantly established, (a) In the germination of seeds the cell-walls (cellulose) are broken down to supply nutriment to the embryo. (&) In the attack of parasitic plants, dense structures of the ' host ' are penetrated by the most delicate hyphae of the invading organism by a dissolu- tion and resolution of the original cellulose structure. («) Brown and Morris have succeeded in cultivating the excised embryos of the cereals on artificial endosperms, and. in show- ing that a cytohydrolytic enzjme is secreted which is suffi- ciently powerful in its action as to attack (i.e. hydrolyse) the typical cotton cellulose. While, therefore, we may expect in the future to be furnished with the means of studying an enzyme hydrolysis or dissection of cellulose, we are at present limited to the action of powerful reagents such as the concentrated mineral acids, which, in a later section, we shall show to resolve cellulose through a series of hydration products (amyloid, dextrins, dextrose) presenting many analogies with the starch-dextrose series above described. Cellulose, therefore, while an analogue of starch, is differ- entiated from it by an enormously greater resistance to hydrolytic actions of all kinds, which must express a corre- sponding difference in constitution. We have now to study the actions of hydrolytic agents in their more concentrated forms, (a) Alkalis. — Cold solutions of the alkaline hydrates of a certain concentration exert a remarkable effect upon the cellu- CELLULOSE. 17 loses. Solution of sodium hydrate, at strengths exceeding 10 p.ct. NagO, when brought into contact with the cotton fibre at the ordinary temperature, instantly changes its structural features, i.e. from a flattened riband with a large central canal, produces a thickened cylinder with the canal more or less obliterated. These effects in the mass, e.g. in cotton cloth, are seen in a considerable shrinkage of length and width, with corresponding thickening, the fabric becom- ing translucent at the same time. The results are due to a definite reaction between the cellulose and the alkaline hydrates, in the molecular ratio C12H20O10 : 2NaOH, accom- panied by combination with water (hydration). The com- pound of the cellulose and alkali which is formed is decom- posed on washing with water, the alkali being recovered unchanged, the cellulose reappearing in a modified form, viz. as the hydrate (Ci2H2oOio-H20). By treatment with alcohol, on the other hand, one half of the alkali is removed in solution, the reacting groui)s remaining associated in the- ratio C12H20O10 : NaOH. The reaction is known as that of Mercerisation, after the name of Mercer, by whom it was discovered and exhaustively investigated. Technical Applications. — Until quite recently, the observa- tions of Mercer remained undeveloped. They are now applied on a large and increasing scale to the production of a silky lustre in cotton textiles. It is found that if during the- action of the alkaline lye the cotton goods are kept under strain, the physical changes determined in the fibres enable it to produce that concentrated reflection of incident light which causes lustre.* These structural changes are perma- nent, persisting after the removal of the alkali by washing. Alkali Cellulose. — The compounds resulting from the union of cellulose and the alkaline hydrates, though of little stability— as we have seen — are still well-defined products. This fact is emphasised by the production from the alkali cellulose of two series of characteristic derivatives of cellu- ' Rev. Gen. d. Mat. Col. 1898. C 18 PAPER-MAKING. lose : (a) the siilpho-carbouates ; (6) the benzoatt s — whicli are described in later secti(jns. ()S) The Acids in concentrated form act in two opposite directions upon cellulose : (1) they attack and resolve the cellulose complex by processes of condensation and hydro- lysis ; (2) they combine with the cellulose (OH groups) to form acid ethers or esters. The consideration of the latter we defer until vve have completed our survey of hydrolytic actions by now describing briefly those determined by the mineral acids in their more concentrated forms. Hydrochloric Acid in presence of water rapidly converts the fibrous cotton cellulose into a friable substance, the formation of which is not attended by any visible changes ; hut, on subjecting the pioduct to pressure or mechanical action, it becomes a more or less structureless powder. This product, known as hydro- or hydra cellulose, was first investigated as a cellulose derivative by A. Girard ; * but the physical changes of the cotton fibre under the ^actions of acids which accompany the formation of this pro- duct had previously been studied by various observers. Sulphuric Acid. — Hydro-cellvilose results also from the action, at ordinary temperatures, of sulphuric acid at certain dilutions. The following points have been established by C. Koechlin.f What maybe called the critical concentration of the acid in regard to the production of hydro-cellulose lies between the limits 60-80° B. Thus, with the mixture of 3 vols, of the concentrated acid and 8 vols, water — i.e. an acid of 69° B — at the ordinary temperature, its action upon cotton does not become evident till after three hours' ex- posure. With an aqueous acid containing 100 grams H2S0^ per litre and at 80° C, the first appearances of change in the cotton are noted at the expiration of five minutes ; after thirty minutes' exposure there is sensible disintegration ; after sixty minutes the conversion is complete, i.e. into a friable mass of hydro-cellulose. * Me'moire sur I'Hydroct lluloses. Paris, 1881. Gauthier-Villers. t Bull. Mulhoiise, 1S88. . CELLULOSE. 19 The subjoined ai e analyses of specimens produced under very variable conditions of treatment of the cotton : (1) (2) (3) (4) C12H02OH Carbon . Hydrogen Oxygen . 42-10 6-30 51-60 42-50 6-50 51-00 42-04 6-70 51-26 41-80 6-70 51-50 42-1 6-4 51-5 Viz. (1) and (2) by the action of sulphuric acid of 45° B. ; (3) by the action of gaseous hydrochloric acid in presence of moisture; (4) by the action of dilute sulphuric acid (3 p.ct. H2SOJ at 60° C. The action of such acids, therefore, which do not combine with the cellulose to form definite esters (infra') is one of hydration and hydrolysis, the in- soluble product tending to a limit l epresented by the formula »i[Ci2H2oOio . H2O]. The hydration of cellulose to such a pro- duct^ is attended by a gain of weight. But the reaction studied in the mass, i.e. as an aggregate effect, is always attended by loss of weight, the hydro-cellulose weighing less, and under some conditions of action of acids, much less than the original cellulose. This is due to a further hydrolysis of a portion of the cellulose to products soluble in water — tending, in fact, to the limit of extreme hydrolysis to dex- trose. The reaction, in fact, is a complicated one, condensa- tion or dehydration playing an important part in the changes which may, in the net or aggregate effect, result in a hydra- tion of the cellulose. The properties of i:he hydro-cellulose are in some respects those of cellulose : it dissolves, though more readily, in the special solvents of cellulose ; and combines, but also more readily, with nitric acid {infra) to form a similar series of nitrates. Generally it is in all respects much more reactive. It rapidly oxidises when heated at 100°, with discolora- tion ; the brown-coloured products of oxidation are soluble in water, and reduce Fe-hling's solution. Hydro-cellulose is c 2 20 PAPEE-MAKING. attacked by dilute alkalis and dissolved more or less ; the solutions are yellow in colour and reduce Febling's solution. It is clear, therefore, that the main characteristics of the hydro-cellulose series is a progressive hydrolysis of the cellulose, with liberation of the CO groups of the constituent hexose units. These hydroly tic changes are attended by pro- gressive structural disintegration, and the products have lost the characteristics of chemical inactivity. In all properties, therefore, wbich determine the industrial value of cellulose, and notably to the paper-maker, the hydro-celluloses are inferior, or degradation products. Technical Applications.— The removal of vegetable fibres from mixed woollen refuse (shoddy), and of vegetable im- purities from raw wool, is effected by acid treatments of various degrees such as we have described, and to which the wool is resistant. To the paper-maker hydi'o-cellulose has first the negative importance that he has by all means to avoid its production in the finished web or sheet. Sulphuric Acid.— Di- and tri-hydrate — H2SO4.2H2O, H2SO4 . 3H2O — produce a special series of hydration effects. Unsized paper plunged into sulphuric acid diluted previously with I- to ^ its volume of water and cooled, is rapidly attacked, the paper becoming transparent, owing to the swelling and gelatinisation of the fibres. The reaction quickly becomes one of solution ; but if the paper be transferred after short exposure to water, the acid compound is decomposed, and the resulting gelatinous hydrate is preci[)itated in situ. The product, after exhaustive washing and drying, is obtained as parchment paper. This modification gives a tough translucent sheet. The hydrated compound itself, produced as described, from its resemblance to starch, has been termed amyloid. Its empirical composition is that of the hydro-celluloses, viz. n [C12H22O11], to which compounds also it is closely allied in chemical properties.* * See ' Soluble and Insoluble Colloidal Cellulose and Composition of Parchment Paper.' Guignet. Compt. Kend., 108, 1258. CELLULOSE. 21 Technical Application. — Tlie process as described above is carried out industrially on ' continuous ' machinery, the parchment paper being produced in endless length. The product is variously applied as a substitute for paper where resistance to water is required. Nitric Acid of sp. gr. 1*4 also produces (without oxida- tion) an effect of a similar character. A short immersion of unsized paper, e.g. filter paper, in the acid, followed by copious washing, has a considerable toughening action, attended by a shinkage in linear dimensions of about y'^-* The effect is made use of in the laboratory as a convenient method of toughening filter papers when required to stand exceptional fluid pressures. To follow a strictly logical order of treatment, we should describe here the compounds of cellulose, with acid radicals — the cellulose esters : for instance, the sulphates and nitrates which are formed on bringing cellulose in contact with the respective concentrated acids. But we defer the mention of these to a later section, for the reasons (1) that they have only an indirect bearing on the technology of paper- mating, and (2) that we have yet to continue tho dis- cussion of the relationship of cellulose to water and oxygen, upon the basis of the views which have been gradually developed to this point. Sulpho-Carbonates of Cellulose [Cellulose xantho- genic acid]. — We have shown that cellulose combines with the alkaline hydrates, and when the resulting compound, or alkali-cellulose (hydrate), is exposed to the action of carbon disulphide at the ordinary temperature, a simple synthesis takes place, which may be formulated by the typical equation : X.ONa-l-CSa = CS.gj^^. The best conditions for the reaction appear to be when the I'eagents are brought together in the molecular pro- portions : * See Journ. Cliem. Soc, 47, 183. 22 PAPER-MAKING. 2NaOH CS2 2x40 76 [.3O-4OH2O] ; 162 the second ONa group being in direct union with the cellu- lose molecule, which reacts, therefore, as an alkali cellulose. The resulting compound may therefore he described as an alkali-cellulose-xanthate. It is perfectly soluble in water, to a solution of extraordinary viscosity. The course of the reaction by which it is produced is marked by the further swelling of the mercerised fibre and a gradual conversion into a gelatinous transparent mass, which dissolves to a homogeneous solution on treatment with water. To carry out the reaction in practice, bleached cotton is treated with excess of a 15 p.ct. solution of NaOH, and squeezed till it retains about three times its weight of the solution. It is then placed in a stoppered bottle with carbon disulphide, the quantity being about 60 p.ct. of the weight of the cotton. After standing about three hours at ordinary temperatures, water is added sufficient to cover the mass, and the further hydration of the compound allowed to proceed spontaneously some hours (e.g. over night). On stirring, a homogeneous liquid is obtained, which may be diluted to any required degree. Thus prepared, the crude solution is of a yellow colour, due to by-products of the reaction (trithiocarbonates). The pure compound is obtained either by treatment of the solution with saturated brine or with alcohol. It forms a greenish-white flocculent mass or coagulum, which redis- solves in water to a colourless or faintly yellow coloured solution. Solutions of the salts of the heavy metals added to this solution precipitate the corresponding xanthates. Iodine acts according to the typical equation : CS OX , xo SNa + JSaS CS + I2 = 2NaI -f CS. OX . xo s-s CS. The compound, which may be described as a cellulose dioxythiocarbonate, is precipitated in the flocculent form ; it CELLULOSE. 23 is redissolved by alkaline solution, in presence of reducing agents, to form the original compound. The most characteristic property of the cellulose xanthates is (a) their spontaneous decomposition into cellulose (hydrate), alkali, and carbon disulphide — or products of in- teraction of the latter. When this decomposition proceeds in aqueous solution, at any degree of concentration exceed- ing 1 p.ct. cellulose, a jelly or coagulum is produced, of the volume of the containing vessel. These highly hydrated modifications of cellulose lose water very gradually, the shrinkage of the ' solid ' taking place symmetrically. The following observations upon a 5 p.ct. solution (cellulose), kept at the ordinary atmospheric temperature, will convey a general idea of the phenomena attending the regeneration of cellulose from the alkali xanthate. The observations were made upon the solution kept in a stoppered cylinder ;. after coagulation, the solution, expressed from the coagulum of cellulose by spontaneous shrinkage, was removed at inter- vals. Original volume of solution, 100 c.c. Coagulation . First appearance I of liquid . ./ Time in 8 th days, day Vol. of cellulose hydrate. nth 16th „ . . 98-0 . . 2- 20th „ . . 83-0 . . 16- 25th „ . . 72 0 . . 28- 80th ,. . . .'58-0 . . 42- 40th „ . 42-8 . . 57- 47th „ . . 38-5 . . 61- DilT. fi-om 100 c.c. = vol. expressed. 5-0 i-5 !-0 !-0 '•2 •5 The shrinkage from a 5 p.ct. to a 10 p.ct. coagulum of cellu- lose hydrate is therefore extremely slow and fairly regular ; from 10-12 p.ct. there is considerable retardation ; and at 12-15 p.ct. the coagulum may be considered as a hydrate, stable in a moist atmosphere. It follows from these observa- tions that if a 10-12 p.ct. solution be allowed to coagulate spontaneously, the resulting cellulose hydrate will undergo very small shrinkage if kept in a moist atmosphere. T hese observations indicate the uses which can be made of the solution in preparing cellulose casts and moulds. 24 PAPER-MAKING. As regards the problem of hydration and dehydration of the cellulose there are, of course, other methods of approxi- mately determining the ' force ' by which the water mole- cules are held. It is a problem of wide significance, by reason of the important part played by such hydrates in the economy of plant life. Further investigations of the problem, therefore, by the various known methods are being prosecuted. (6) Coagulation by heat. — The solution may be evaporated at low temperatures to a dry solid, perfectly re-soluble in water. If heated at 70-80°, however, the solution thickens ; and at 80-90° the coagulation (i.e. decomposition) is rapidly completed. If the solution be dried down at this tempera- ture in thin films, it adheres with great tenacity to the surface upon which it is dried. On treatment with water, Jiowever, the cellulose film may be detached, and when freed from the by-products of the reaction the cellulose is obtained as a homogeneous transparent colourless sheet or film, of great toughness, which, on drying, hardens somewhat, in- creasing in toughness and preserving a considerable degree of elasticity. From the properties of the solution and of the cellulose regenerated from it, it will be seen that both are capable of extensive applications. Quantitative Eegeneration of Cellulose from Solution AS Thiocarbonate. — Very careful experiments have been made to determine the proportion of cellulose recovered from solution as thiocarbonate. Weighed quantities of Swedish filter paper were dissolved by the process, and the solutions treated as follows : (a) allowed to ' solidify ' spontaneously at 15-18°; (&) coagulated more rapidly at 55-65°; (c) sul- phurous acid was added in quantity sufficient to combine with one-third of the alkali present in the solution — the resulting solution being colourless : this was then set aside to coagulate spontaneously. The regenerated celluloses were exhaustively purified, by boiling in sodium sulphite solution, digesting in acid, digesting in water, &c., and, repeating the treatments until pure, they were finally dried at 60° and finished at 100°. CELLULOSE. 25 The following results were obtained : — Weight of Weight of original cellulose. regenerated cellulose. (a) . . 1-7335 .... 1-7480 (6) . . 1-7415 .... 1-7560 (c) . . 1-8030 .... 1-8350 The results show a net difference of 1 • 1 p.ct. (increase), a quantity which, for practical purposes, may be neglected. As, however, the empirical composition of the regenerated cellulose indicates hydration to 4C6H10O5.H2O (infra), and a corresponding gain of 2-7 p.ct., it appears that there is a slight hydrolysis of even this very pure form of cellulose to soluble products. From subsequent observations (p. 38) it will appear that the hydrolysis falls upon an oxycellulose, probably present in all bleached celluloses. The cellulose regenerated from the thiocarbonate differs from the original cellulose, so far as has been ascertained, in the following respects : — (1) Its hygroscopic moisture, or water of condition, is some 3-4 p.ct. higher, viz. from 9-10 • 5 p.ct. (2) Empirical composition. — The mean results of analysis show C = 43*3, H = 6-4 p.ct., which are expressed by the empirical formula, 4:GqHiqO^ ' H2O. (3) General properties, in the main, are identical with those of the original, but the OH groups of this cellulose are in a more reactive condition. Thus, this form of cellulose is acetylated by merely heating with acetic anhydride at its boiling point, whereas normal cellulose requires a tempera- ture of 180° C. {Vide Cellulose Acetates.) As regards reaction in aqueous solution, we may notice that it has a superior dyeing capacity, and also combines with the soluble bases to a greater extent : e.g. if left some time in contact with a normal solution of sodium hydrate, it absorbs from 4 '5-5 '5 p.ct. of its weight in combination. Towards the special solvents previously described it behaves similarly to the normal or fibrous cellulose ; the solu- tions obtained are, however, more viscous and less gelatinous. 26 TAPEK-MAKING. The thiocarbonate reaction throws light on that some- what vague quantity, the ' reacting unit ' of cellulose. We use this term in preference to that of molecular weight ; for the latter quantity can he determined only for bodies which readily assume the simplest of states, and which can be ascertained by physical measurements to be in that state ; whereas in the case of cellulose the ordinary criteria of molecular simplicity are quite inapplicable. We have formulated the synthesis of the thiocarbonate as taking place by the interaction of CgHjoOg : 2NaOH : CS., ; or in approximate percentage ratio — Cellulose: Alkali: Carbon disulphide = 100 : 50: 50; or, again, in terms of the constituents estimated in the analysis of the product- Cellulose: Alkali (Na^O): Sulphur = 100: 40: 40. If, now, the crude product be precipitated from aqueous solu- tion by alcohol or brine, and again dissolved and reprecipi- tated, the ratio changes to 100: 20: 20; and, through a suc- cession of similar treatments, the ratio of alkali and sulphur to cellulose continually diminishes the product, however, preserving its solubility. In fact, no definite break has been observed in the continuous passage from the compound as originally synthesised to the regenerated cellulose (hydrate). It is clear, therefore, that the reacting cellulose unit is a con- tinually aggregating molecule; and if in the original syn- thesis it appears to react as CgHioOg, so in a thiocarbonate containing, e.g. only 4 p.ct. NaaO, the unit is lOCgHigOg. There being, moreover, no ascertainable break in the series, we have no data for assigning any limiting value to the reacting unit under these conditions. All we can say is, that the evidence we have points to its being of indefinite magnitude ; and we can see no a priori reason why it should not be so. In discussing this reaction we have left out of considera- tion the part played by the water. It may be noted that a CELLULOSE. 27 1 p.ct. solution of cellulose (as thiocarbonate) will ' set ' to a firm jelly of hydrate, of the volume of the containing vessel ; and that even at 0-25 p.ct. cellulose, gelatinisatiun of the liquid occurs in decomposition. We have also pointed out that a hydrate containing only 10 pet. cellulose is a sub- stantial solid which gives up water with extreme slowness. Cellulose, therefore, affords conspicuous illustrations of the property which the ' colloids ' have, as a class, of ' fixing * water, and of the modes in which this property takes effect. In regard to the causes underlying this peculiar relatioH- ship to water, we know as yet but little. It is to be noted that the group of colloids comprises bodies of very various chemical function, acids, bases, salts and compounds of mixed function, as in the complex carbohydrates and proteids ; the only possible feature common to so varied a group would be that of molecular arrangement, favouring the aggregation of the molecules, together with those of water, to groups of indefinite magnitude. On this subject, however, conjectures must, for the present, do duty for a theory which can only be shaped by further investigation. Technical applications. — The solutions of cellulose in the form of sulphocarbonate are prepared for industrial use at concentrations of 10-20 p.ct. (cellulose) according to require- ment. The product in this form is known as ' viscose.' Its uses depend upon the ease with which the cellulose can be regenerated from the solution and in any desired form, viz. compact solid (' viscoid '), sheet or film, powder, or lastly thread (' lustra-cellulose ') — retaining in these forms the essential properties of the oi iginal fibrous cellulose. Viscose is used in the engine sizing of jDulps for working up into paper and boards ; in the coating of papers as the vehicle, chiefly for white pigments such as barium sulphate and China clay ; in the preparati(jn of filmed fabrics such as ' leather cloths ; ' aiid for a number of other similar purposes. These applications all depend upon the relatively low cost of production of this particular solution of cellulose, together with the obvious advantages of an aqueous solution 28 PAPEE-MAKING. from whicli cellulose is directly regenerated by a variety of simple methods. Cellulose and Acid Radicals: Cellulose Esters. — For the reasons previously given, these compounds will only be briefly described, i.e. only so far as is necessary to com- plete our review of the general chemistry of cellulose. While these derivatives are many of them of enormous industrial importance, they have only an indirect connection with the work of the paper-maker, and for further information we must refer him to special treatises. An ester may be generally defined as a species of salt formed by union of an alcoholic OH group with an acid, water being formed and eliminated. In these reactions cellulose may be represented by its unit group G^K^^O^. In this unit there are several OH groups, each capable of such combination. On general grounds, in fact, it might be expected that four of the five atoms of O represent OH groups, and the formation of a tetr acetate has supplied the proof that the formula may be written CgHgO (OH)^. The highest nitrate which has been obtained is, on the other hand, the tri-nitrate, CgH^Og (ON02)3. Certain of these esters may be formed without destroying the fibrous form of the original cellulose (nitrates, ben- zoates) ; in others the combination is attended by solution of the cellulose (acetates, sulphates). Similarly, from the amorphous modifications of cellulose, nitrates and benzoates may be formed without solution of the product. Nitric Esters: Cellulose Nitrates. — These are the best known of the synthetical derivatives of cellulose, their discovery dating back over fifty years, during which period they have come into ever-increasing industrial application. They are formed by the inter-action of cellulose, in any form, and concentrated nitric acid. Water being formed simul- taneously, it is usual to carry out the reaction in presence of sulphuric acid, which combines with and removes the water from the sphere of reaction. A concise view of the entire series of these nitrates is CELLULOSE. 29 given by Vieille in a paper entitled ' Sur les degres de la Nitrification limites de la Cellulose.' * From the title of this author's communication, it maj^ be concluded that it is a study of the nitrations of cellulose (cotton) under the con- dition of progressive variations, with the view of deter- mining the maximum fixation of the nitric group corre- sponding to such variations. The most important factor of the pi'ocess is the concentration of the nitric acid, which was the variant investigated. The temperature was kept con- stant — 1 1 ° C. — and the nitrating acid (nitric acid only) was employed in very large excess (100-150 times the weight of cellulose), so as to avoid disturbance of the results by rise of temperature or by dilution of the acid. The products were analysed by Schloesing's method, and the analyses are ex- pressed in c.c. NO (gas) (at 0° and 760 mm.) per 1 grm. of substance. In regard to the time factor, or duration of exposure to the acid required to give the maximum number, this was in cases controlled by observation. Thus with the acid HNOg, ^ HgO (1-488 sp. gr.), after 48 hours the product was still blued by iodine, and gave 161 c.c. NO ; whereas after 62 hours'" exposure the iodine reaction was not obtainable, and the maximum number (165*7 c.c. NO) was obtained. At the slightly lower gravity 1*483, an exposure of 120 hours was necessary. At the still lower gravity when the cotton (nitrate) passes into solution, the maximum is very rapidly attained (5 minutes). The highest nitrate obtained as above, with nitric acid only, is somewhat lower than when sulphuric acid is present. Under these latter conditions the author regards the highest nitrate obtainable as G2iB.is (J^^3^)ii^9- The following is a brief account of the usual methods of preparation, and of the properties of the various products. In the case of this series the cellulose unit group is taken as C12H20O10, by which device a nomenclature in fractional proportions is avoided. * Compt. Kend. 95 132. 30 PAPER-MAKING. Compofiiion (approximate). I NO3H.IH I N03H.|H,0 I NO3H.IH2O I [ NO,H.H..O ! Analysis of I'roduct c.c. NO per 1 grm. Properties of Products. Structural features of cotton pre- serveil ; boluble in acetic ether ; not in ether-alcohol a,H,„(N03H),<,0,„ Appearances unchanged ; soluble in ether-alcohol ; collodion cotton 0,,H,,(N03H),.0„ C,,H,,(N03H),0„ Fibre still unresolved ; soluble as above, but solutions more gela- tinous and thready 0,,H,„(N03H);0,3 Dissolve cotton to viscous solution ; products precipitated by water ; gelatinised by acetic ether ; not ether-alcohol C,,H,,(N03H),0„ Friable pulp ; blued strongly by iodine in KI solution ; insoiubfe in alcoholic solvents C,,H3„.(NO,H)A5 C,,H3.XN03H),0,e Several well chai-acterised citrates have been formed, but it is a very diflficult matter to prepare any one in a state of purity, and without admixture of a higher or lower nitrated body. The following are known : — Hexa-nitrate, Ci2Hi404(N03)g, gun-cotton. In the for- mation of this body, nitric acid of sp. gr. 1 • 5, and sulphuric acid of sp. gr. 1 • 84 are mixed, in varying proportions — about 3 of nitric to 1 of sulphuric. Sometimes this proportion is reversed, and cotton immersed in this at a temperature not exceeding 10° C. for 24 hours: 100 parts of cellulose yield about 175 of cellulose nitrate. The hexa-nitrate so prepared is insoluble in alcohol, ether, or mixtures of both, in glacial acetic acid or methyl alcohol. Acetone dissolves it very slowly. This is the most explosive gun-cotton. It ignites at 160°- CELLULOSE. 31 170° 0. According to Eder, the mixtures of nitre and sul- phuric acid do not give this nitrate. Ordinary gun-cotton may contain as much as 12 per cent, of nitrates soluble in ether-alcohol. The hexa-nitrate seems to he the only one quite insoluble in ether-alcohol. Penta-nitrate, CigHigOg (N03)g. This composition has been very commonly ascribed to gun-cotton. It is difficult, if not impossible, to prepare it in a state of purity by the direct action of the acid on cellulose. The best method is the one dcA^sed by Eder, making use of the property discovered l>y de Vrij, that gun-cotton (hexa-nitrate) dissolves in nitric acid at about 80 or *J0° G., and is precipitated, as the penta- nitrate, by sulphuric acid after cooling to 0° C; after mixing with a larger volume of water, and washing the precipi- tate with water and then with alcohol, it is dissolved in ether-alcohol, and again precipitated with water, when it is obtained pure. This nitrate is insoluble in alcohol, but dissolves readily in ether-alcohol, and slightly in acetic acid. Strong potash solution converts this nitrate into the di-nitrate O12H18O8 (m,),. [Eder.] The tetra- and tri-nitrates (collodion pyroxyline) are generally formed together when cellulose is treated w ith a more dilute nitric acid, and at a higher temperature, and for a much shorter time (13 to 20 minutes) than in the formation of the hexa-nitrate. It is not possible to separate them, as they are soluble to the same extent in ether-alcohol, acetic- ether, or wood spirit. On treatment with concentrated nitric and sulphuric acid, both the tri- and tetra-nitrates are converted into penta-nitrate and hexa-nitrate. Potash and ammonia convert them into di-nitrate. [Eder.] Cellulose di-nitrate, Ci8Hj308 (N03)2 formed as a product of partial saponification by the action of alkalis on the higher nitrates, and also by the action of hot dilute nitric acid on cellulose. The di-nitrate is very soluble in alcohol-ether, acetic-ether, and in absolute alcohol. Further action of 32 PAPEE-MAKING. alkalis on the di-nitrate results in a complete decomposition of the molecule, some organic acids and tarry matters being formed. [Eder.] By the graduated action of alkaline hydrates on the nitrates previously dissolved, hydroxypyruvic acid is obtained as a main product. [W. Will.] Technical Applications. — The nitrates of cellulose, as such, are the basis of extremely important industries, and of very opposite character, viz. (1) the production of explosives both fur military and industr ial use, and (2) as a structural material in the production of xylonite and celluloid, both in the form of compact solids and in sheet or film. The latter uses depend upon the ease with which the nitrates are brought into a plastic condition, or entirely dissolved in various ' neutral ' solvents, e.g. alcohol-ether, acetone, amyl acetate. In association also with camphor and vegetable oils (castor oil) their plastic qualities are considerably heightened. In later years the nitrates have been used as a means of obtaining an artificial cellulose fibre : the solutions of the nitrate being drawn or spun into water as a precipitating solution or into air and the thread is afterwards ' denitrated ' by treatment with ammonium sulphide. The resulting pro- duct is a practically nitrogen-free cellulose. It is known as artificial silk, or by the more appropriate term, lusti a-cellulose. Cellulose Acetates. — Certain of the polyhydric alcohols, glycerin, unite directly with acetic acid to furm acetic esters. Cellulose, however, is indifferent to acetic acid under any conditions of action so far investigated. Neither does it react with the anhydride of the acid at its boiling point ; but at 180°, in contact with six times its weight of the anhydride, it is converted into the triacetate (Schutzenberger). With twice its weight of the anhydride, on the other hand, a mixture of lower acetates is obtained, insoluble in glacial acetic acid. The triacetate is freely soluble in the acid. The solutions are highly viscous and are filtered with ex- treme diflSculty. Passage through filter paper is facilitated by dilution with chloroform and benzene. The acetate is CELLULOSE. 33 soluble, as are all the esters of cellulose, in nitro-benzene. The cellulose regenerated from the thiocarbonate solutions, it is to be noted, reacts with the anh^ dride directly, passing gradually into solution in the anhydride at its boiling point. The product appears to be the tetracetate. The fibrous celluloses of straw and esparto are also partially acetylated under these conditions. Reaction of the normal cotton cellulose with acetic anhydride, at the boiling point of the latter, is determined by the addition of zinc chloride in relatively minute pro- portion (Franchimont) ; probably, however, as the result of a previous hydrolysis of the cellulose. Iodine has also been found to determine the reaction. The most characteristic acetylation of cellulose, however, is the following : — The cellulose regenerated from the thio- carbonate is, after purification, mixed with the molecular proportion of zinc or magnesium acetate [CqK^qO^ : Zn (O.C2H30)2] in concentrated solution, the mixture dried down on the water-bath, and finally dehydrated at 105°. It is then moistened with acetic anhydride, and acetyl chloride is added in the proportion of 2-C2H3O-01. Keaction ensues at 30-50°, the inass liquefies, and the formation of cellulose tetracetate results. The pure product is obtained as an opaque, white, voluminous powder, soluble in acetic acid and chloroform, to solutions of high viscosity which, on evapora- tion in thin layers, leave the acetate in the form of transparent coherent film. On boiling these films with normal sodium hydrate diluted with an equal volume of alcohol, the ester is resolved into acetic acid (soluble as sodium acetate) and cellulose (insoluble). The cellulose is obtained as a trans- parent coherent film. This is an important criterion of a true cellulose acetate, as distinguished from acetates of deriva- tives of lower molecular weight ; the latter giving brittle films which are more or less disintegrated by the process of saponification, and also yield a proportion of soluble carbo- hydrates, reducing cupric oxide in alkaline solution. Technical Applications. — The acetates have only recently 34 PAPEC-MAKING. been prepared on the manufacturing scale. Tlieir physical properties are very similar to those of the nitrates ; they are, however, non-explosive and withstand a temperature of 200^ It appears, therefore, that they are capable of substituting the nitrates in many of their useful applications other than as explosives, and with greatest advantage when the proper- ties of an explosive or liigh inflammability are not merely superfluous but prejudicial. Cellulose Benzoates.— Esters containing the radical of benzoic acid are obtained by inter-action of the alkali celluloses with benzoyl chloride in presence of excess of the solution of alkaline hydrates. (a) Mercerised Cellulose. — The benzoates obtained fiom the alkali cellulose in this form retain the fibrous structure of the original cellulose. A mixture of products is obtained, varying in composition from a mono- to di-benzoa.te. (&) Soluble Alkali Celluloses— These react in a more definite way: the products are purified by dissolving in glacial acetic acid, filtering from residues of unattacked celluose, and reprecipitating by water. Thus isolated, the ben- zoates approximate in composition to G&^s^3> (a) The oxycellulose gelatinised on washing and was similar to the product obtained by the action of nitric acid * Collie, Journ. Chem. Soc. 65, 262. 44 PAPEU-MAKING. (/3) The oxidised carbohydrate in solution resembled ' caramel ' in appearance. The compound or mixture was precipitated by basic lead acetate, and isolated by decom- posing the precipitate with hydrogen sulphide, filtering and evaporating. On distillation from hydrochloric acid, furfural was obtained in large proportions. (3) Extreme action of alkaline hydrates. — -When fused at 200-300° C. with two to three times its weight of sodium or potassium hydrates, cellulose is entirely resolved, the charac- teristic products being hydrogen gas and acetic (20-30 p.ct.) and oxalic (30-50 p.ct.) acids. Generally the reaction takes the same course as with the simpler carbohydrate-^, resolution of the cellulose into molecules of similar constitution, no doubt, preceding the final resolution, which appears to be an exothermic or explosive reaction. C. Dkstructive Eesolution by Ferment Actions. — This group of decompositions of cellulose is necessarily a very wide one. In the 'natural' world of living organisms, of course, no structures are pei manent ; and although cellulose distinguishes itself by relative permanence and resistance to the disintegrating actions of water and oxygen, the differen- tiation in this respect is only a question of degree, and all cellulosic structures are subject to the law or necessity of redistribution. The directions of redistribution are chiefly three: viz. (1) In the assimilating process of the plant a cellulosic structure is broken down, re-absorbed into the supply of plastic nutrient material, and re-elaborated. (2) Structures which have ceased to play a part in the general organisation of the plant are cast off, and then exposed as 'dead' matter to the play of the redistributing agencies of the natural world. The processes of ' deca}^ ' take various forms, according to the conditions to which they are exposed. The humus of soils, peat, lignite, amd all forms of coal present various forms of the residual soli d products of the decay of cellulosic structures, the remainder having been dissipated and restored to the general fund oi CELLULOSE. 45 matter in cixculation, in the gaseous form — viz, as CO2 and CH,. (3) In the processes of animal nutrition plants and vege- tahle suhsta&ces are, of course, most important factors. In the course of animal digestion the vegetable substances are attacked by the fluids of the alimentary tract and resolved into proximate constituents fulfilling the requirements of the organs of assimilation ; and in addition to these decomposi- tions, which are largely hydrolytic in character, more funda- mental resolutions are observed in which the carbohydrate molecules are completely broken down, i.e. with formation of gaseous products. The Cellulose Group. — Thus far we have been dealing mainly with one member of the very numerous class of plant constituents comprehended in the term ' cellulose.' While the properties and characteristics of cotton cellulose are in such wise representative that this substance maybe regarded as the typical cellulose, the differentiation of this, as of every other group of tissue constituents, in conformity with functional variation, necessarily covers a wide range of divergencies. The celluloses of the plant world, so far as they have been investigated from the point of view of chemical constitution, group themselves as follows : — (a) Those of maximum resistance to hydrolytic action, and containing no directly active CO groups. (&) Those of lesser resistance to hydrolytic action, and containing active CO groups. (c) Those of low resistance to hydrolysis, i.e. more or less soluble in alkaline solutions and easily resolved by acids, with formation of carbohydrates of low molecular weight. Group (a). — In addition to the typical cotton cellulose — which, it is to be noted, is a seed-hair — there may be included in this group the following fibrous celluloses which constitute the "bast of exogenous flowering annuals : viz. the celluloses of Flax (Linum usit.). Hemp (Cannabis sativa), China Grass (Rhea and Boehmeria species), and of the lesser known 46 PAPEE-MAKING. Marsdeilia tenacissima, Caloiropis (gigantea), Sunn Hemp (Crotalaria juncea). As in the case of cotton, tKe celluloses of the fibres are considered in the form of the white (or bleached) and purified residues, resulting from the treatment of raw materials by- processes of alkaline hydrolysis and oxidation more or less severe in cliaracter. For the purification of the celluloses in the laboratory, the methods usually practised consist in (1) alkaline hydrolysis, i.e. treatment with boiling solutions of sodium hydrate, carbonate or sulphite; (2) exposure to bromine water or chlorine gas ; or, when oxidation alone is sufiicient for the removal of the ' impurities,' to solutions of the hypochlorites or .permanganates (in the latter case fol- lowed by a treatment with sulphurous acid to remove the MnOa deposited on the fibre-substance); (3) repetition of (1) for the removal of products rendered soluble by (2). The celluloses of this group thus purified may be taken as chemically identical with cotton cellulose, investigation having so far failed to differentiate them. It must be noted, however, that the several members of the group present dis- tinct morphological characterisiics, and differ also in such external properties as lustre and ' feel.' These are correlated with the differences in minute structure, but they are no doubt in part due to differences of substance. So far, how- ever, we have no knowledge of the proximate constitution of these substances, and can therefore say nothing as to the causes of difference in this respect. On the other hand, the essential identity of these cellu- loses is established in regard to ultimate composition and in reference to the following properties and reactions : — (1) Eesistance to hydrolysis and oxidation, and other negative characteristics, indicating a low reactivity of the CO and OH groups. (2) The relationships to the special solvents previously described, including the thiocarbonate reaction. (3) Formation of esters, nitrates, acetates, benzoates. Of the above, it is sufficient in general laboratory practice CELLULOSE. 47 to examine cellulose in regard to ultimate composition, re- sistance to alkaline hydrolysis, behaviour with solvents, and reactions with sulphuric acid (solution without blackening) and nitrating mixture (H2SO4 and HNO3) 5 ' nitration ' proceeds without oxidation, and gives a higher yield of product, 160-180 p.ct., according to the conditions. Group (&). — These celluloses are diflferentiated from the former group (1) by ultimate composition, the proportion of oxygen being higher; (2) by the presence of active CO groups ; (3) in certain cases by the presence of the O'CHg group. The general characteristics of the group are those of the oxy celluloses. It has recently been shown that these oxidised derivatives of the normal celluloses are further characterised by yielding /Mr/wraZ as a product of acid (HCl) hydrolysis. The yield of this aldehyde is, in certain cases, increased by previous treatment of the oxycellulose with a reagent prepared by saturating sulphuric acid of 1 • 55 f^p.gr. with HCl gas. In this reagent the oxycelluloses dissolve ; and on then diluting with HCl of 1-06 sp.gr. and distilling, maximum yields of furfural are obtained, the yield being an indirect measure of the increased proportion of oxygen beyond that correspond- ing with the formula CgHmOs. Celluloses of this class are much more widely distributed in the plant world than those of the cotton type ; they appear, from recent observations, to constitute the main mass of the fundamental tissue of flowering plants, in which they usually exist in intimate mixture or combination with other groups more or less allied in general characteristics. It appears, from a survey of the contributions of investigators to the subject of cellulose, that research has been very much confined to the fibrous celluloses, more particularly to such as receive extended industrial use. The time has come, how- ever, when systematic research is much needed to establish at least a preliminary classification of the 'cellular' cellu- loses upon the lines of chemical constitution. Constitution, taken in relation to physiological function, is an attractive 48 PAPER-MAKING. subject of research ; and it is in the plant cell, where syn- thetical operations are predominant, that we have to look for the foundations of a ' new chemistry,' which shall definitely investigate the relation of matter to life. It is to be noted that the differentiation of many of these celluloses from the typical cotton is, in regard to empirical composition, only slight. There appear, on the other hand, to be more important differences of constitution. Thus pine- wood cellulose dissolved in sulphuric acid, the solution diluted and boiled, and further treated for the isolation of crystallisable carbohydrates, yields these (i.e. dextrose) in only small proportions.* Investigation has stopped short at this negative result. It would be of interest, therefore, to isolate the products formed in the reaction with the concentrated sulphuric acid, so as to characterise them, at least generally. Until this is done, or some other method of proximate resolution is worked out in detail, we can only say that the constitution of these celluloses is in some important feature radically different from that of the typical cellulose. An account of recent investigations of these ' celluloses ' will be found in Berl. Ber. 1893, and a more special treat- ment of the subject, ibid. 1894, and Journ. Chem. Soc. 1894 (C. Smith). Of this group of the natural oxycelluloses, the following have been more particularly investigated : — (1) Celluloses from woods and lignified tissues generally. — Lignified tissues are made up of compound celluloses, to Ibe subsequently described (see ' Ligno-celluloses,' p. 53), frcm which the celluloses may be isolated by a number of treait- ments, all depending upon the relative reactivity of the sio- called 'non-cellulose' constituents, which in combination with the celluloses make up the compound cellulose, ligno- cellulose or wood-substance. These non-cellulose constituen ts are readily attacked and converted into soluble derivativess ; >and there are various industrial processes for preparing * Liudsey and Tollens, Lieb. Ann. 267, 370. CELLULOSE. 49 celluloses (paper pulp) from raw materials of this class, depending upon the direct conversion of the former into such soluble compounds. The isolated celluloses show the following general characteristics : * — Elementary composition ^-^^"^^[^^^^'^l Yield of fur- fural, by solution and hydrolysis (HCl), 2-6 p.ct. Beactions with phenylhydrazine salts and magenta-sulphurous acid, indicating the presence of active CO groups. These cellu- loses are necessarily less resistant to oxidation and hydrolysis, but show in all other respects a close general .agreement with the normal cotton cellulose. (2) Celluloses from cereal straws, from esparto, etc. — These celluloses are isolated from the matured stem or haulm, by digestion with alkaline lye at elevated temperatures. They are also of considerable industrial importance, being largely used in the manufacture of the cheaper kinds of writing and printing papers. Eecent investigation has shown that these celluloses are strongly differentiated from the normal, and are, in fact, pro- nounced oxycelluloses. The following are the characteristics of difference : — Ultimate composition, after treatment with hydrofluoric, acid, to remove siliceous ash constituents : — Oat straw cellulose. Esparto cellnlose. (1) (2) (1) (2) 0 . . . 42-4 42-4 41-78 41'02 H . . .5-8 5-8 5-42 5-82- Yield of furfural by solution and hydrolysis (HCl) : Oat straw cellulose. Esparto cellulose. 12-5 12-2 Reactions. — In addition to those with Fehling's solution, phenylhydrazine salts, and magenta sulphurous acid indicat- ing the presence of active CO groups, the celluloses give a characteristic rose-red colouration on boiling with solutions * Berl. Ber. 27, 161. E 50 PAPER-MAKING. of aniline salts. This reaction serves to identify their presence in papers, and from the depth of the colouration the percentage may he approximately estimated. Investigation has also established the ibllowing jDoints in regard to the oxidation and deoxidation of these oxy- celluloses. They are gradually oxidised in dry air at the temperature of the water-oven, undei going discoloration ; the jield of furfural by hydrolysis showing a progressive increase. They are deoxidised, on the other hand, by neuti'al and alkaline reducing agents. Thus, after lengthened exposure to solu- tions of zinc-sodium hyposulphite, prepared by the action of zinc dust upon sodium bisulphite, the yield of furfural — which is a measui-e of the degree of oxidation — was reduced, in the case of esparto cellulose, from 12*6 to 8-9 p.ct. A still further deoxidation results from solution of those oxycelluloses as thiocarbonate, and regeneration of the cellu- lose by heating the solution at 80-100°. The regenerated • cellulose approximates to the normal, yielding only 2 p.ct. furfural on hydrolysis. It is to be noted, however, that esparto cellulose, in common with all the celluloses of this group, is partly hydrolysed to soluble derivatives by this treatment; the regenerated cellulose amounting to 80 p.ct. ■ of the original weight dissolved. The soluble portions yield furfural on hydrolysis, amounting (in a typical experiment) to 4*0 p.ct. of the original. The celluloses of this group are dissolved b} concentrated ;sulphuric acid to dark-coloured solutions. On diluting and boiling, they are resolved into carbohydrates of low molecular weight; dextrose appears to be invariably formed, and in many cases also mannose ; but only very small yields of >either carbohydrates have been so far obtained. Group (c). — This includes the heterogeneous class of non-fibrous celluloses which we have defined as of low re- sistance to hydrolysis, being easily resolved by boiling with dilute acids, and being also more or less soluble in dilute alkaline solutions. This group has been but little studied CELLULOSE. 51 and therefore can only be generally characterised. Physio- logical research has shown that there are a large number of cellular, as distinguished from fibrous ' celluloses,' which are readily broken down (hydrolysed) by the action of enzymes within the plant iiself, whether as a normal or abnormal incident of growth. Thus, in the germination of starchy seeds, the cell walls (cellulose) of the btarch-con- taining cells are broken down, as a preliminary to the attack upon the starch granules themselves, to form the supply of nutrition to the embryo. In an exhaustive investigation of the germination of the barley, Brown and Morris have thrown a good deal of light upon this particular point, which they emphasise in the following words : ' that the dissolution of the cell wall invariably precedes that of the cell contents during the breaking down of the endosperm is a fact of the highest physiological importance, and one which for the most part has been strangely overlooked.' A similar, but abnormal, dissolution of cell walls is that which occurs in the attacks of parasitic organisms upon the tissues which they invade. These processes are well known to physiologists, who, however, generally regard 'cell wall' and 'cellulose' as sub- stantially identical terms. The chemical differentiation of the substances comprising cell walls is, on the other hand, an entirely new field of research ; but, although investiga- tion has not gone very far, the results are sufficient to show that the celluloses of this order are enormously diversified. The variations already disclosed are (1) those of the carbo- hydrates yielded by ultimate hydrolysis, and (2) those of molecular configuration or condensation. We have already seen that the celluloses of the cotton group (a) yield dextrose as the ultimate product of hydrolysis; those of group (h) yield, in addition to dextrose, mannose and probably other bodies ; and the group we are at present discussing yield, in addition, galactose, and the pentoses xylose and arabinose. In illustration we may cite a few examples. Thus Galactose has been obtained as a product of hydrolysis of the cell E 2 52 PAPER-MAKING. walls of the seeds of Lupinus luteiis, Soja liispida, Coffea arabica, Pisum sativum, Cocos micifera, Phoenix dactylifera, etc. Mannose is obtained in relatively large quantity from the ' ivory nut,' and from a very large number of other seeds ; and Pentoses, from the seeds of the cereals and of leguminous plants. It appears, therefore, generally that a large number of plant constituents which have been denominated by the physiologists as ' cellulose' have little more title to be con- sidered as such than has starch. However, external resem- blances count for something, at least in the beginnings of classification, and substances of the type we are considering may be conveniently grouped with the celluloses ; but we should propose to apply to them the term Pseudo-celluloses, or Hemicelluloses— as has been proposed by E. Schulze. Our group (c) of pseudo-celluloses may therefore be defined as substances closely resembling in appearance the true cellu- loses, but easily resolved into simple carbohydrates by the hydrolytic action of enzymes, or of the dilute acids and alkalis. Compound Celluloses. — In dealing with the isolated celluloses, it has been shown that the processes by which they are isolated or purified are based upon the relative reactivity of the compounds with which the celluloses are combined or mixed, in the raw or natural products of plant life. These natural forms of cellulose are, of course, multi- tudinous. Eemembering the infinite variety of the vegetable world, the endless differentiation of form and substance of the tissues of plants, it might be presumed that the chemical classification of these products would present unusual com- plications. Investigation, however, has shown, and continues to show, that this great diversity of substance, as revealed by pro:xi- mate analysis, exists upon a relatively simple chemical basis. The compounds constituting the fundamental tissue of plamts may, in fact, be broadly classified in correspondence wath the three main types of differentiation of the cell wall lomg CELLULOSE. 53 recognised hy the physiologists, viz. lignification, stiberisation and conversion into mucilage. That is to say, in addition to the celluloses proper and hemi- or pseudo-celluloses — which may be defined as polyanhydrides of the normal carbo- hydrates, ketoses and aldoses — there are tliree main types of compound celluloses in which the celluloses as thus defined exist in combination with other groups, as follows : — LiGNOCELLULOSES. — The substance of lignified cells and fibres, notably the woods — of which the characteristic non- i'ellulose constituent is a R. hexene derivative. Pectocelluloses and Mucocelluloses. — Comprising a wide range of tissue constituents — of which the non-cellulose con- stituents are colloidal forms of the carbohydrates, or closely- allied derivatives, easily converted by hydrolytic treatments into soluble derivatives of lower molecular weight, and belonging to the series of ' pectic ' compounds, or hexoses, etc. Adipocelluloses and Cutocelluloses. — The substance of cuticular and suberised tissues — in which the cellulose is associated with fatty and waxy bodies of high molecular weight. Of the above groups, the first only has any direct interest to the paper-maker. The ligno-celluloses as such are in the forms of the jute fibre and ' mechanical ' pulps, actual components of papers of certain classes ; the pectic and cuticular consti- tuents of fibrous raw materials are almost entirely eliminated by the chemical and mechanical methods of treatment of such materials, of separating and preparing the actual paper- making fibres. LiGNOCELLULOSES. — As the chemical prototype of the sub- stance of lignified tissues, we select the jute fibie. This fibre is the isolated bast tissue of plants of the species Cor- chorus (order Tiliacese) an annual of rapid growth, attaining a height of 10-12 feet in the few months required in the Indian climate for the maturing of the plant. The textile fibre or bast is obtained in long strands. It is of a brown to silver-grey colour in the finer sorts. The jute fibre sahsfance differs strikingly in composition 54 PAPER-MAKING. and reactions from the celluloses. Its ultimate composition is represented by the percentage numbers : — J Calc. for Ci2llx80g. The above numbers are calculated to dry ash free sub- stance. In the air-dry condition the fibre contains on the average 10 '0 p.ct. moisture; the normal ash amounts in the aggregate to about 1 • 0 p.ct. Beaciions. — It is diflferentiated from the celluloses by the following characteristic reactions. With solutions of aniline salts (and other coal tar bases) it gives a deep yellow colour- ation. With pMoroglucinol dissolved in hydrochloric acid (1'06 J'p.gr.) it gives a rose-red colouration : the red solution of ferric ferricyanide is reduced with production of ' prussian ' blue which is deposited in the fibre-substance ; a solution of magenta discolorised by sulphurous acid dyes the fibre a deep magenta colour, a reaction which is characteristic of aldehydes and ketones. The jute fibre is also dyed directly by a large variety of the ' coal tar ' dyes. But the most characteristic of its reactions is that of direct combination with the halogen elements, and notably chlorine. The combination is attended with change of colour to bright yellow ; the yellow quinone chloride (infra) is dis- solved by a solution of sodium sulphite, with development of a magenta-red colouration. While serving thus as a qualitative reaction for the identification of the lignocellulose, it is also one which takes place in definite quantitative proportions. Further, when so carried out, it afibrds a sharp separation of the lignocellulose into its constituent groups : viz. (1) the lignone (ketone) which combines with the chlorine and may then be dissolved away by sodium sulphite solution, leaving (2) the cellulose constituents of the fibre-substance, having the properties of the celluloses of group (2) of the classification on p. 49. 0 H O •160-47-0 6-1- 5-8 47 -9-47 -2 47-0 60 47-0 CELLULOSE. 55 From its critical importance the process of chlorination of the jute lignocellulose will be described in detail. The fibre is prepared by a preliminary treatment with a boiling solution of sodium hydrate (1 p.ct. NaOH). From this it is well washed and squeezed to retain not more than its own weight of water. In this condition it is exposed for one hour to an atmosphere of washed chlorine gas. The fully chlorin- ated fibre is washed to remove the hydrochloric acid formed in the reaction, and transferred to a solution of the sulphite (1-2 p.ct. ISaaSOg) in which it is boiled. To complete the isolation of the cellulose, the solution is made alkaline with sodium hydrate (0-2 p.ct. NaOH), after which the boiling is continued 2-3 minutes. The fibrous mass is thrown on a cloth filter, and thoroughly washed. This residue is the almost pure cellulose ; and the entire elimination of the non- cellulose components of the original fibre, is shown by failure to react with the various reagents previously cited The following are important quantitative data. The yellow chlorinated derivative has the empirical formula CigH^gCl^Og ;. to form this derivative the fibre-substance (purified and prepared by the boiling with alkali) takes up 8 p.ct. of its weight in combination ; at the same time an equal amount goes to form hydrochloric acid. The proportion of cellulose yielded is approximately 80 p.ct. This cellulose is a mix- ture of two celluloses, a cellulose resembling the normal cotton cellulose, and {3 cellulose, a furfural-yielding cellulose, and much less resistant to hydrolytic treatments, both acid and alkaline. In other processes of attacking the fibre for the isolation of cellulose, this latter is hydrolysed and dis- solved, and the yield falls to 62-66 p.ct. Bromine acts similarly, but its action is relatively slow and incomplete. Thus if the fibre prepared by boiling in weak alkalis is submerged in saturated bromine water, it requires some hours at the ordinary temperature for the fibre-substance to combine with the maximum proportion of the halogen. The brominated product is similarly attacked and dissolved by alkaline solutions ; but it will be found 56 PAPER-MAKING. that the treatment, requires to be once or twice repeated before the isolation of cellulose is complete. In certain cases, however, the method is useful as an alternative pro- cess of isolating and estimating cellulose, not only in jute, but in other fibrous materials. Iodine is absorbed by the Hgno-celluloses from aqueois solutions, and they are coloured a deep brown. The absorption is constant, under constant conditions of treatment, and as a result of quantitative investigations, it was established thit in digesting the fibre-substance at 18° C, with twenty timss N its weight of the ^ iodine solution (in potassium iodide) the absorption is constant at 12- 9-13 '3 p.ct. No definite compound is formed, however, and the halogen is easily removed by solvents. These reactions with the halogens, and notably with chlorine, are dealt with thus prominently, as they are speci- fically characteristic of the lignone or non-cellulose group, the presence ot which in combination with the cellulose is the distinguishing feature of this class of compound cellu- loses. The lignone group is sharply differentiated from the carbohydrates by its reactions, which are those of an unsaturated compound. Investigation of the chlorinated derivative has established it as a quinone chloride and related to the trihydric phenol pyrogallol, yielding as it does tri- chloropyrogallol on reduction. Subordinate characteristics of the lignone complex are the presence of methoxyl groups (0 • CHg), and the absence of reactive hydroxyl (OH) groups. It must be noted that the union of cellulose and lignone to form the ligno-cellulose complex is very intimate, and in several characteristic reactions both of decomposition and synthesis, the ligno-cellulose behaves as a homogenous com- pound. Of such we may cite more particularly the actions of hydrolytic agents under ordinary conditions on the one hand, and reactions of ' nitration ' and ' acetylation ' on the other. The highly reactive nature of the lignone group exposes it, on the other hand, to selective attack not only by the halogen CELLULOSE. 57 as^described, but by dilute nitric acid at temperatures of 50° and upwards, and by sulphurous acid and the bisulphites at elevated temperatures, all leading consequently to the isolation of the more resistant cellulose. We must briefly note here the structural relationship of the jute-cellulose to the jute-fibre. The cellulose is obtained in the form of single, ultimate, bast fibres or cells, of short length (2 mm. = inch) ; whereas, the jute-fibre occurs in apparently long strands. The latter are, however, in effect complex, and a cross-section viewed under the microscope reveals a number of the unit cells in close apposition. It would appear therefore that these fibre-bundles owe their cohesion to the Ugnone constituent; and hence the term 'encrusting and intercellular suhstavce' by which it is fre- quently designated. This view is at variance with the undoubted chemical homogeneity of the fibre-substance; and the student must avoid any so crude a view of a ligno- cellnlose, as of pure cellulose 'overlaid' or 'encrusted' with its lignone or non-cellulose components. The remaining features of the chemistry of the typical lignocellulose we note only briefly. Action of Cellulose Solvents. —The lignocellulose is attacked and dissolved similarly to the celluloses, by zinc chloride in concentrated solution, whether in water or hydro- chloric acid : so also by solutions of cuprammonium. The solution is attended by hydrolysis, which is greater or less according to the conditions and duration of the action : by precipitation of the solutions the recovery of the lignocellu- lose is therefore incomplete. The lignone groups are found in both soluble and insoluble fractions,. Action of Hydrolytic Agents.— (1) Alkaline solutions at temperatures of 60-100° attack the lignocellulose as a whole, dissolving a proportion of the fibre-substance from 10-25 p.ct., according to the nature of the alkali, the con- centration of the solution and the temperature and duration of its action. With caustic soda solution (of 3 p.ct. NajO) at elevated temperatures (150-170°), the ligmme and cellu- 58 PAPER-MAKING. lose groups are selectively atfaclcecl, and a cellulose is isolated approximating in composition to the normal celluloses. Concentrated solutions of caustic soda (12-25 p.ct. NagO) in the cold determine a • mercerising' action, attended by a remarkable change in appearance due to partial resolution of the fibre-bundles, and the swelling of the individual fibres. These hydration effects are accompanied by a partial con- version into soluble derivatives (10-15 p.ct.) but there is no essential change in the composition of the lignocellulose. (2) Acids. — The dilute 'mineral' acids, e.g. HCl and H2SO4 of semi-normal strength rapidly disintegrate the fibre at temperatures exceeding 60°. As the action continues the fibre takes a mahogany-red colour and falls to a mass of brittle fragments, harsh to the touch. The acid solution acquires a reddish-yellow colour and smells strongly of acetic acid : this acid is in effect produced in some quantity. The furfural-yielding groups of the lignocellulose yield more readily to these actions, and when the lignocellulose is digested with 1 p.ct. sulphuric acid at 3 atm. pressure, they are entirely hydrolysed and dissolved. The insoluble fibrous residue gives the characteristic reactions of the original fibre, i.e. the essential features of its lignone constituent are not changed. Dilute Nitric Acid at 50-80° has a specific and selective action upon the lignone group ; from which soluble yellow coloured nitroso-ketones are formed, the y8 cellulose is simul- taneously hydrolysed, and finally the a cellulose is isolated. This reaction is characteristic, and has been more closely studied with the related group of lignocelluloses, the hard woods.* Sulpho-carbonate Reaction.— The joint and simul- taneous action of carbon disulphide, and the alkaline hydrates e.g. solutions of 15-20 p.ct. NaOH, produces only a partial converbion of the constituents of the fibre into sulpho-car- bonates, the proportion actually passing into solution on subsequent treatment with water varying from 25-50 p..ct. » E. 0. 0. Baly and J. 0. Chorley, Berl. Ber. 195, 9 22. CELLULOSE. 59 according to the conditions of treatment. The reaction is remarkable for the extreme degree of hydi ation and gelatini- sation of the portion which does not actually dissolve: in the swollen and gelatinised condition it has been observed to occupy 100 times the volume of the original. In this reaction it is the group we have designated as )8 cellulose which yields most readily, and is, in fact, dissolved. The undissolved lesidue reacts with chlorine as does the original lignocellulose. Compounds with Acid Radicals. Esters. — Of these, the nitrates are alone of suflS.cicntly definite composition to be considered as derivatives of the entire ligno-cellulose. The nitrates are obtained by short exposure to the action of the mixed acids (HNO3 + HjSO^) ; with concentrated nitric acid alone, rapid and destructive oxidation supervenes. The highest yield of nitrate obtainable has been determined at 145 p.ct. of the original lignocellulose : this increase of weight would correspond with the production of a tri-nitrate of the formula C^JI^^Oq . (N03)3. The percentage of nitrogen in the products is from 11* 8 to 12-2, and this indicates the production of a tetra-nitrate. This discrepancy is accounted for by secondary reactions — with production and removal of water molecules. The nitrates are of a brilliant yellow colour and are homogeneous in composition, fractional solu- tion failing to resolve them, as would be the case if they were mixtures of cellulose nitrates, with derivatives of the lignone group. In explosive properties, and in their be- haviour towards solvents, they closely resemble the cellulose nitrates. Compound with other acid radicals, e.g. acetates and benzoates, have been obtained, but they are of an ill-defined character. Action of Oxidising Agents.— As ' unsatuiated ' com- pounds, i.e. containing unsaturated groups, the lignocelluloses are relatively greedy of oxygen. Also it is evident that the lignone groups yield more readily to such actions than the cellulose constituents of the fibre. Thus by the action of 60 PAPER-MAKING. oliromic acid in dilute solution and in presence of sulphuric acid the former may be entirely eliminated. The most important oxidations, however, are those of the ' standard ' bleaching agents — the permanganates and hypochlorites. By a careful regulation of the conditions, the jute fibre may be bleached by these reagents to a very high point, but by no means to the pure white of a cellulose. The lignocellulose is essentially a coloured compound, and in attempting to exceed the limit which this imposes, we merely remove the Hgnone groups and approximate to the condition of cellu- lose. This implies loss of weight, and for the reasons given above, loss of strength by actual disintegration of the fibre- bundles. As the products of oxidation of the fibre substance are acid in character, precautions must be taken in employing the hypochlorites, that the bleaching solutions do not become acid : in that case hypochlorous acid is set free and the lig- none groups are chloiinated. The presence of the resulting chlorinated derivatives in paper or textiles is extremely prejudicial. The lignocelluloses, of which we have briefly described the typical jute-fibre, are of course widely distributed in the plant world, and are the components of structures of various function. We may instance as ligno-celluloses of closely related composition and constitution, the hard or stony con- cretions of the pear: and, again, the most important con- stituent of the juice of the white currant, a type of soluble and hydrated lignocelluloses, of which group it has all the essential characteristics. This latter is a further proof that the lignocelluloses, though complex, are homogeneous. But the most widely distributed of the ligno-celluloses are the woods : in this group a surprising unity of chemical type is preserved ; and though a woody stem is made up of cells and vessels of various structure and function, we may regard them as composed of an identical chemical substance. Im the woods proper, all those features of the typical ligno- cellulose, which it has in contrast to cellulose, are more CELLULOSE. 61 strongly developed. This is readily seen by a comparison of a typical wood, viz. beech- wood, with the jute-fibre in regard to the essential chemical constants : — Eli'mentary Composition. Proximate resolution. Quantitative Reactions of Non-cellulose. Non- Cl. Carbon. Cellulose. cellulcise. Methoxyl. Furfural. Combining. Jute . . 46-5 75 25 4-0 S-2 8-0 Beech 49' 1 55 45 6-2 12'8 12-0 Kemembering that the woods are the substance of pe- rennial stems, whereas the lignocelluloses previously dealt with occur in or as 'annual' structures, these numbers contain the suggestion that the process of lignification con- sists chemically in the production of the unsaturated lignone groups at the expense of cellulose. This view is emphasised by the progressively 'condensed' character of thermion of the groups, which shows itself in the I'elatively strong re- sistance of the woods to all processes of simple hydration and hydrolysis. This is most clearl}^ seen in relation to the solvents of cellulose : for while jute and the ' annual ' ligno- celluloses are similarly dissolved, the woods yield to no process of solution as a whole. More definitely characteristic is their complete resistance to the thiocarbonate reaction; which argues the absence of reactive OH groups. This, again, is- confirmed by the results of conversion into nitrates by the action of nitric acid in presence of sulphuric acid, which are so imperfect and unsatisfactory that this group of ligno- celluloses cannot be said to be ' nitrated ' as such, but yield only as the result of partial resolution into constituent groups. A little reflection will show that inertness in these directions is an important condition of such ' permanent ' functions as the woods are destined to fulfil. But in all other respects there is the closest similarity between the two groups of lignocelluloses. The specific colour reactions are shown by the woods in greater intensity 62 PAPER-MAKING. proportioned to their higher proportion of unsaturated (lig- none) groups, thus: — (a) Golden yellow, with solutions of salts of aniline, etc. (b) Carmine-red, with solutions of phloroglucinnl in HCl (1-06). (c) Magenta, due to liberation of this colouiing matter from its solutions decolorised by sulphurous acid. ( general treatment of the celluloses. So far no reactions have been brought to light in which it is differentiated from cottoni cellulose, with perhaps one exception, viz. its lesser resistance to hydrolysis. Thus H. Miiller mentions * that flax cellulose^ isolated by the bromine method, lost, on boiling five times; with a dilute solution of sodium carbonate (1 p.ct. NagCOg),.. 10 p.ct. of its weight. The statements of E. Godefi'roy,| that flax cellulose is distinguished from cotton cellulose by its; reducing action upon silver nitrate in boiling neutral solution,, are eironeous, the reaction resulting from residual impurities,, which, for the reasons given, are extremely difficult to isolate. * Pflanzenfaser, p. .38. t Abstracted in Journ. See. Chem. Ind. (1889), 575. CELLULOSE. 71 Flax cellulose may therefore, for the present, be regarded as chemically indistinguishable from cotton cellulose. The oil and wax constituents of the raw fibre will be described under the group of cuto-celluloses. Other Pecto-celluloses. — As far as investigation has pro- ceeded, it appears that pectose, or pectose-like substances, are associated with all fibrous tissues of the tinlignified order. And indeed in the lignocelluloses themselves pectous sub- stances make their appearance with increasing age. Thus the lower portions of the isolated jute bast— jute cuttings or butts — when boiled in alkaline solution yield products which cau&e the solution to gelatinise on cooling ; and the gelatinous pro- duct is insoluble in alcohol, distinguitihing it, as pectic acid, from the products of hydrolysis of the lignocellulose itself, which are dissolved, after precipitation, by alcohol. It must be remembered, however, that in the 'jute cuttings ' the ad- hesion of the bark and tortical parenchyma to the true bast fibre is such that we are dealing with a complex tissue, and the source of the pectic acid may be in the parenchyma of the tissue and not in the bast fibre. On the other hand, it has been shown that in the spontaneous decomposition of jute, lying in the damp state, gelatinous acid bodies are formed indistinguishable from pectic acid. It would not be difficulty therefore, to account f*.r the pectic constituents of the bast tissue towards the root end as products of degradation of the lignocellulose itself. Eeverting, however, to the non-lignified fibres such as China grass, or Ramie (Bohmeria species), and the 'nettle fibres' generally, hemp, and even raw cotton— these all con- tain pectic bodies associated with the cellulose, which are hydrolysed and dissolved by treatment with boiling alkalis. But these pecto-celluloses have not been sufficiently investi- gated as compound celluloses to admit of any useful classifi- cation on the basis of particular constitutional variations of their non- cellulose constituents. The mono-cotyledonous fibre-aggregates, whether fibre- vascular bundles (Phormium, Aloe fibres, Musa, etc.) or entire 72 TAPEE- MAKING. planls (Esparto, Bamboo stems, Sugar Cane), are largely made up of pecto-celluloses, with a greater or less proportion of lignocellu loses. But the constitution of these non-cellulose constituents is as yet quite unknown, and we have therefore none but the general basis of classification. In the same way also the parenchymatous tissue of fruits, fleshy roots, etc. — the typical pecto-celluloses— must be, for the present, dismissed with the bare mention. The investigation of these substances belongs rather to the province of general carbohydrate chemistry than to the narrower cellulose group ; and the problems involved are in many respects rather morphological and physiological than purel}'- chemical. These same considerations apply also in great measure to the mucilaginous constituents of plant tissues, though certain of these have been investigated by modern chemical methods. The relationship of these substances to cellulose is indicated (a) hj the histology of the tissues, which shows them to be associated with the cell wall rather than with the cell contents ; (b) by their tmpirical composition, which is approximately that of cellulose ; (c) by their reactions with iodine, by which they are coloured variously from blue to violet, as are the hydrated modifications of cellulose.* Beyond superficial ob- servations of reactions (iodine) and gelatinisation with water, these compound celluloses — which may conveniently be termed muco-celluloses — had been but little investigated | until the systematic work of Kirchner and Tollens, and Gans and Tollens, I upon the mucilages and gums. Cuto-celluloses [Adipo-celluloses]. — The plant repre- sents, in the one view, an assemblage of synthetical opera- tions carried on within a space enclosed and protected from the destructive influences of water and unlimited atmospheric oxygen. The protecting external tissues are those which we are about to describe as constituting the third important group of compound celluloses. These tissues contain, in * Sachsse, Farbstoffe, etc., p. 161. f ?acbsse, loc. at. t Anualen, 175, 205 ; 249, 245, CELLULOSE. 73 admixture with the tissue-substance, a variety of oily and waxy products (easily removed by mechanical solvtnts), the presence of which adds very considerably to the water- resisting property of the tissue. It will be seen as we proceed, however, that the tissue-substance, after being entirely freed from these adventitious constituents or oily excreta, yields a large additional quantity of such products when decom- posed by ' artificial ' processes of oxidation and saponification. By this and by its empirical composition {infra) the tissue- substance will be seen to contain ' residues ' of high carbon percentage and molecular weight, and closely allied in chemical structure to the oil and wax compounds found in the ' free ' state in the tissue as it occurs in the plant. These groups are associated in combination in the tissue with cellu- lose residues, and hence the description of such complexes as adipo-cel luloses. There are two main types of these compound-celluloses, viz. cork and the cuticular tibsues of leaves, stems, etc. CoKK in its ordinary form is a complex mixture contain- ing not only oils and waxes, but tannins, lignocelluloses, and nitrogenous residues. The following are the results of elementary analysis : (a) of cork purified by exhau .stive treatment with ether, alcohol and water ; (h) of cork {Quer- cus suher') without purification ; (c) of the cork tissue of the cuticle of the potato (tuber) purified by exhaustion with alcohol : — c (a) {h) (c) . 67-8 65-7 62-3 H . 8-7 8-3 71 <) . 21-2 24-5 27-6 N . 2-3 1-5 3-0 The analyses are calculated on the ash-free substance.* These investigators succeeded in isolating cellulose from cork, but by complicated and drastic methods of treatment, such as would break down the greater proportion of the cellulose into soluble derivatives. These treatments were ; * Dopping, Annalen, lo, 286 ; Mitscherlich, Annalen, 75, 305. 74 PAPER-MAKING. (1) drastic oxidation with nitric acid ; (2) alternate treat- ments with boiling dilute hydrochloric acid and 10 p.ct. solution of potassium hydrate. The proportion thus isolated amounted to 2-3 p.ct. only. The authors, on the other hand, have observed that the non-cellulose of cork is entirely converted into soluble deri- vatives by the process of digestion at high temperatures with solutions of the alkaline sulphites, as described, p. 66. In this way a residue is obtained preserving the form (i.e. cellular structure) of the original cork, and amounting to 9- 12 p.ct. The details of a particular experiment were as follows : 10- 995 grms. coik, 20 grms. NaaSOg-THoO, 2 grms. NaaCOg, dissolved in 500 c.c. water. Digested 3 hours at 75 lbs., and 4 hours at 125 lbs. pressure. Residue bleached with sodium hypochlorite solution. Yield of cellulose, 1 * 34 grmsi. ; 12-1 p.ct. These relationships have been more definitely made owt by Fremy, in his investigations of the closely allied cocn- pound wliich constitutes the epidermal or cuticular tissue of the leaves, stems, etc., of Phanerograms. As the character- istic constituent of cork is termed suberin, so Fremy terims this cuticular tissue-substance cutin or cutose. To prepa:re this substance, cuticular tissue ('peel') of the apple, e.g., is treated with boiling dilute acids, followed by digestion with the cuprammonium reagent (p. 9) ; then again wilth boiling acid and dilute alkali (KOH) ; finally the residue is treated with alcohol and ether. In this way a nitrogen-frt=C[ B ? n fa n p c -r Fig. 29. Fig. 30, form of small blocks is kept in contact with the stone by a number of hydiaulic presses P. A somewhat similar arrangement, but placed horizontally, is shown in front and side elevations in Figs. 29 and 30. The same letters correspond. Fig. 31. The ground and sorted pulp is made into thick sheets by nueans of the board machine shown in Fig. 31. The pulp mixed with water passes down the shoot D into the vat B in which the cylinder K revolves. This cylinder is covered with wire-cloth, and as it revolves it takes with it a cer- tain quantity of. pulp in the form of a continuous sheet. TKEATMElSfT OF VAKIOUS FIBEES. 151 This steet is taken on to tlie endless travelling felt F ty the small couch roll E. When it reaches the rolls C C it is wound round the upper one, from which it is removed when a sufficient thickness is obtained. Obtained in this form the pulp is readily transported. The woods commonly employed are white pine and aspen. The latter yields a pulp of a better colour, but of inferior strength than the former. Paper containing mechanical wood pulp is very liable to become discoloured by the action of air and light, the lignocelluloses being much more readily acted upon than the celluloses isolated from them. Such fibre is, moreover, devoid of much felting power; it has, in fact, little to recommend it but its comparative cheapness. It is never- theless used in large quantities, some cheap papers being made almost entirely from it. In reference to the alkaline treatment of fibres dealt with in the early part of this chapter, a Table of Strength of Caustic Soda Solution will be found useful for reference^ for which see the next page. 152 PAPEE-MAKING. Table op Strength of Caustic Soda Solutions (15° C. = 59° F.) (Tunnerman). Specific Gravity. Degrees Twaddle.* Per cent. Na20. Equivalent Percentage of 60 per cent. Caustic Soda. 1 • 0040 0-80 U OvO 1-0081 1 62 U Dlf 1 1 .0163 3-26 I Ma 1 0246 4-92 1 - 81 Q 1 olo ^ • 091 1 ■ 0330 6 60 9 "4.1 8 4.-n>a2So37H20. More concentrated forms are prepared by the action of sulphurous acid upon soda ash. It is obvious also that calcium and magnesium sulj)hites may be used to replace the more expensive sodium salt. Sulphites are converted by the action of bleach into sulphates, thus ; — Ca(0Cl)2 + 2Na2SO, = CaSO^ -f -f 2NaCL Calcium Sodium Calcium Sodium Sodium hypochlorite. sulphite. sulphate. sulphate. chloride. SIZING, LOADING, COLOURING, ETC. 183 ' From this equation it will be seen that 252 parts of sodium sulphite will neutralise 143 parts of calcium hypo- chlorite, or 204-3 parts of bleaching powder. Assuming that crystallised sodium sulphite contains 50 p.ct. NajSOg, the same amount of bleach would require 504 parts. Comparing these numbers with those given above for sodium hypo- sulphite, it will be seen that 204-5 parts of bleach require for neutralisation 129 parts of sodium thiosulphate and 504 parts of crystallised sodium sulphite, or proportionately less of stronger products. Sodium sulphite is preferred to sodium thiosulphate by m')st paper-makers, notwithstanding the fact that even in its most concentrated form nearly three times as much is required to produce a certain result. It is found that when it is used the wire- cloth of the machine is preserved for a longer time than if sodium thiosulphate is employed. This may be due to the fact that with the latter a certain amount of free acid is always formed, which of course would act injuriously on the wire ; whereas, when sodium sulphite is used, the products of decomposition are neutral salts without any action upon metals. (See the above equations.) Hydrogen ])eroxide has been used to some extent as a,n * antichlor ' with satisfactory results. The only objection to its use is its relatively high cost. Whichever variety of 'antichlor' is used, an excess should be carefully avoided, as all act more or less upon the size and colouring matter added to the pulp subsequently. The quantity is easily adjusted, either by an actual estimation of the ' chlorine ' in the pulp and calculating therefrom, or the ' antichlor ' may be added gradually, testing the stuff from time to time with the iodide test papers (p. 156). Sizing. — If the celluloses, reduced by the beating process to the condition of ultimate fibres, or further disintegrated by tearing and cutting to 'artificial' units of suitable length, are put together into a sheet or web, the product will be a ' blotting ' or bibulous paper. Such papers have a well known use ; and ' blottings ' are in effect made from cellulose 184 papp:e-making. pulps with the addition of a small proportioa of starch, added to the beater in the form of starch paste, and for the purpose of adding a certain 'binding' quality to the fibres. Such papers, in addition to their absorbent properties for liquids, are ' soft,' relatively weak, and altogether unsuited to uses involving mechanical wear and tear. Writing papers on the other hand have, as is well known, a totally different texture and ' handle,' and they have a certain water and ink-resisting quality : that is, they are required to oppose as little resistance as possible to the me- chanical action of writing, but to resist the penetration of the ink in such a way that the written characters sink into the substance of the paper in the one direction only, i.e. at right angles to the surface. These properties are in large part added by means of sizing agents. While we shall consider the sizing process from this point of view, it must not be forgotten that the texture of the paper is an important factor, and that this is influenced by everything which tends to ' close ' the sheet : for instance, the mineral loading agents (to be dealt with in the next section) have the effect of filling up the interspaces between the fibres : or the effect may be produced by suit- able beating of the pulp. Thus, though the ' sizing ' of pa[)or may appear to be a simple process of adding substances of certain water-resisting quality, which quality they commu- nicate to the paper, the experienced paper-maker recognises it as a really complicated result, influenced by a great num- ber of factors. This is more particularly the case with the pulp or engine-sizing processes — which, in the sequence of treatments will be considered here. The alternative method of sizing the finished sheet or web, ' tub-sizing,' as it is called, is a simpler process which in the case of machine- made papers is carried out on the paper-machine, and will be dealt with in that connection. Engine sizing. (Eosin-alum-sizing) — Fur the specific water- or ink-resisting quality specially required in writing and wrapping papers, we are practically limited to rosin or SIZING, LOADING, COLOURING, ETC. 185 colophony. This well-known substance is the residue from turpentine distillation. It consists of an acid body, abietic acid (C44HS4O5), or rather of its anhydride (GnHg^O^). By virtue of its acid character and the solubility of the sodium salt (C44H|3205Na2) of the acid, rosin is rapidly dissolved on heating with solutions of the alkalis, e.g. carbonate of soda, caustic soda. The solution of the rosin-soap is similar in properties to solutions of the ordinary soaps, which are soda compounds of the higher fatty acids, and in fact the simi- larity of properties is such that mixtures of fat and resin acids, the latter in large proportion, are the basis of the ' yellow ' soaps of commerce, which, outside the slight differ- ence in colour and smell, are not distinguishable from fatty acid soaps. The reaction between abietic acid and carbonate of soda, with expulsion of carbonic acid (gas), may be formulated as follows : — C,,\^,,0, + Na2C03 = G,JI,,0,, Na^ -f CO^ -f H^O ; [672] [106] [Beilstein] that is, 100 parts of the pure acid require 16 "7 of the pure carbonate to form, with the sodium oxide which it contains, the ' neutral ' resinate. Commercial rosin being a mixture, and containing other acids which require more soda per unit, and, on the other hand, hydrocarbons and some other com- pounds devoid of acid properties, the ' equivalent ' of feoda (carbonate) will vary from 15 to 18 p.ct. with various rosins. The equivalent is readily determined for any sample by warming with excess of a standard solution of caustic soda in alcohol till completely dissolved, and titrating the excess of soda with a standard alcoholic acid in presence of phenol phtalein, as indicated (see p. 282.) To make the neutral rosin size on the large scale, the cal- culated quantity of soda ash is dissolved in about twelve times its weight of water, heated preferably in a jacketed pan : the rosin, previously crushed, is gradually added to 186 PAPER-MAKING. about six times the weight of the ash, at)d the boiling con- tinued 3-4 hours until a perfect solution results. The frothing due to the escape of large quantities of car- bonic acid is controlled by the form of the boiling vessel, and by limiting the steam jacket to the lower portion, so that the sides are not ^uperheated. The boiling being finished, the strong 'size' is run off into iron tanks, where in cooling and standing a separation of the ' pure ' soda resinate takes place as a semi-solid mass surrounded by a dark coloured liquor, which is a solution of various by-products. In ihe older practice of size-making this solution was drained away and the size again boiled up with water. On cooling a further separation of coloured products takes place, with a consequent further purification of the 'size.' These operations requiring considerable time, it was customary to boil the supply of size many days in advance, the interval being reqnired for the series of cleansing operations. The quantity of this size per 100 lbs. paper is that representing 3-4 lbs. of the original rosin. In later practicej however, a rosin size is largely used, differing from the above in containing 15 to 25 p.ct. of rosin over and above the 'equivalent' of the soda, the excess of Tosin being dissolved in the strong size in ihe free or uncom- bined state. For an acid size of this composition, finishing with 25 p.ct. of ' free rosin,' the following formula is given by C. Beadle: — 170 lbs. soda ash in 200 gallons of water are boiled with gradual addition of 1800 lb. rosin; the boiling is continued 7 hours, and the volume finally made up to 225 gallons. The acid size on dilution gives a milky liquid, due to suspension of the ' free ' rosin in a finely divided con- dition. More care is required in diluting the size prepara- tory to adding to the beater, since with undissolved rosin there is a danger of agglomeration to particles of such dimen- sions as would cause rosin-specks in the paper. The danger is lessened by adding starch to the diluting water. The fol- lowing working formula is given by Wurster (infra) : — In a 8IZING, LOADING, COLOURING, ETC. suitable vessel water is heated to 50° C. A quantity of starch calculated to 15-20 p.ct. of the rosin in the size to be diluted, is stirred with warm water to a cream and added to the vessel, which is then heated to 80'' until the starch is entirely 'dissolved,' The temperature is lowered to 50-60° by adding water, and the strong size is then stirred in. Finally the temperature is gradually raised to 65°. The extent of dilution has a considerable effect on the efficiency of the size. Wurster recommends a 3 p.ct. limit, i.e. 3 lbs. rosin per 10 gallons size. The quantity of the acid size usually required is that representing 2-3 lb. of the original rosin per 100 lb. paper. The sizing operation consists, first, in adding the size, prepared by either of the above methods, to the pulp m the beater, passing it through a fine sieve or a coarse cloth filter- bag. It is usually added when the pulp is one-half or two- thirds beaten. But following the addition of the size is a second operation by which the decomposition of the soda- resinate is determined, or perhaps rather completed, and the resin acids fixed upon the pulp in the insoluble condition. The reagent used for this purpose, almost universally, is alum or sulphate of alumina. In elucidating this reaction we must consider some of the properties of the alkali salts of the resin and fatty acids. These salts are easily decom- posed by the weak acids, even carbonic acid, the action of which, expressed by the equation CO^ -f- H^O + C^^He^O^. Na^ = Na^COg + C^.Hg.Os, is the exact reverse of that formulated on p. 185, and it takes place in dilute solution at the ordinary temperature. We note also that 44 parts of carbonic acid (CO2) set free 682 parts of the resin acid. Other acids act similarly, and in proportion to their equivalents. Such acids, when com- bined with metallic bases, also pass over to the soda of the resinate when the two salts are brought together in aqueous 188 PAPER-MAKING. solution, with sinuiltaneous formation of the resinate of the metallic base. Thus, with calcium salts, CaCla 4- NaaU = Ca . R + 2 NaCl ; Calcium Chloride. R = Resin Acid. CaSO^ + NaaK = Ca . E + NaaSO^. Calcium Sulphate. Those reactions are all of practical significance : (1) because nearly all natural waters contain carbonic acid, and many contain the sulphate of lime ; and (2) calcium chloride is contributed to the pulp by the bleaching solution, and is present in the beater, less or more, according to the degree of washing of the pulp after bleaching. These resinates are curdy insoluble precipitates. Sulphate of alumina reacts according to the equation — 3 . NaaR 4- AI2 . 3S0^ = 3Na2SO, + A12R3 ; i.e. for 3 mol. abietic acid (= 2016) in a neutral size, we require 342*8 of the sulphate; or in terms of alum, which contains 36 "1 p.ct. sulphate of alumina, 949 parts. In the acid sizes, the proportion cf soda being less, and in the sul- phates of alumina of commerce, the proportion of Al23SO^' varying from 40 to 60 p.ct., it is necessary in every mill to construct a table of ' alum ' equivalents for the size, accord- ing to the composition of each. Thus, as an example, the size is made with 15 p.ct. soda ash per 100 lbs. rosin, and the sulphate of alumina to be used contains 50 p.ct. AI23SO4. The 'alum equivalent' per 100 lbs. rosin is 106 X 15 = 23 lbs., or 0'23 sulphate per 1 lb. rosin. As a matter of fact, however, the quantity used in practice is much in excess of this proportion — ■usually about three times, but often very much more. This ' optimum * proportion has been determined rather hy experience than by quantitative calculations based upon scientific grounds. SIZING, LOADING, COLOUEING, ETC. 189 In 1878, C. Wurster published a treati&e on the Eosin-Alum Sizing Process,* which may be regarded as the first attempt to reduce the process to a scientific or exact basis. Wurster combats the then current view that the actual sizing agent is a resinate of alumina formed as above represented. It appears that ordinary engine-sized papers treated with ether a,nd other solvents of the resin acids, entirely lose their water-resisting quality, losing in weight from 2 to 6 p.ct. ; these dissolved sizing components are isolated by evapora- tion of the solvent, and found to be almost entirely composed of the free rosin acids. This observation is also confirmed by determining the alumina in the paper before and after extraction ; which is found to be but slightly affected. On these results Wurster concludes that the free rosin acids are the effective sizing agents, and hence his advocacy of the acid sizes with a maximum proportion of free or uncombined rosin ; hence also the function of the excess of alum, i.e. to ■decompose the neutral resinate of alumina with liberation of the free acid, thus : — AI2R3 + 2(Al2 S0,)3 + 3 H2O = 3 R . H2 + Al2(Al A) 3 (SOJ. Free Resin Acid. Basic Sulpbate of Alumina. These views require complementing by a careful considera- tion of the chemical function of the cellulose itself. It is well known that cellulose takes up alumina from solutions of the sulphate and chloride ; still more easily from the acetate. It is therefore more than probable that cellu- lose would decompose the neutral resinate by combining with the alumina. It must also be borne in mind that cellulose contains OH-groups of basic function, and that these would combine simultaneously with the rosin acids; nor is this view negatived by the fact that ether and other solvents dissolve away the rosin acids from the finished paper. On the whole, therefore, it is to be recommended that =* «Le Collage des Papiers,' Bull. Soc. Ind. Mulhouse, 1878, pp. 726-801. 190 PAPER-MAKING. all these contributing factors should be borne in mind : (1) the consumption of alum by the basic constituents of the cellulose half-stuff and of the water used in furnishing ; (2) the calculated proportion referred to the soda ui-ed in the particular type of rosin size adopted ; (3) the proportion required by the cellulose itself. For a further discussion of this somewhat complicated problem, see 'Society of Arts Journal ' (Cantor Lectures), Feb. 22, 1897. Auxiliary Sizing Agents. — Under this heading we may mention certain colloidal substances which are or may be used in the engine-sizing process, chiefly for the purpose of closing and compacting the web of fibres. Two of these we have already had occasion to introduce to notice, viz. Starch and Alumina. Starch is used by many makers in large proportion, especially in papers heavily ' loaded ' with china clay (p. 200). It is sometimes added in the form of transparent starch paste, prepared by swelling and dissolving the starch at 80-90° ; it is a common practice also to boil the clay with its complement of starch. In many mills, again, the starch is added in the raw or Tinswollen state, the granules become imprisoned in the network of fibres, and are swollen and burst under the heat of the drying cylinders. Alumina is precipitated as a gelatinous hydrate by the interaction of the soda of the rosin size with snlphate of alumina. Passing through the intermediate form of ' resi- nate,' it is no doubt finally fixed as the oxide, and aids in cementing the fibres together and filling the interspaces. Both the alumina and starch also aid in keeping the rosin acids in the state of minute subdivision during the beating and making processes. Cellulose Hydrates (' Viscose ' Sizing). — In the earlier part of this work attention has been called to the specific charac- teristics of the celluloses of parenchymatous or ' cellular ' as distinguished from fibrous tissue. Under chemical treatment these are gelatinised to hydrated forms wldch retain a rela- tively large percentage of water. Such forms of cellulose no SIZING, LOADING, COLOUKING, ETC. 191 doubt play an important part in paper-making. Certainly the special characteristics of straw cellulose are largely due to the presence of such ' cellular ' celluloses, and also to the general capacity for hydration of its constituent celluloses. The various solutions of cellulose dealt with in Chapter I. would appear to be available for a similar purpose, but for practical use they are mostly precluded by cost of produc- tion. The sulpho-carbonates, however, being prepared at rela- tively low cost, and readily decomposed with regeneration of a normal cellulose endowed with powerful cementing proper- ties, are successfully used in engine sizing. The solution, which goes by the name of ' viscose,' is added to the beater in the earlier stages of the beating, and when thoroughly mixed the decomposing salt is added. The salts best adapted to this purpose are the sulphates of zinc and magnesium, the proportion required being calculated to the equivalent of the soda present in the viscose. Thus, as an example, the wood celluloses are converted into ' viscose ' with a proportion of caustic soda (NaOH) equal to 40 p.ct. of the weight of the cellulose. The molecular weight of NaOH being 40, and its equivalent of the salts being respectively ZnSO.-7H,0 = 287 ; ^ ^^3.^ and MgS0.-7H,0 = 246 123-0, each lb. of dissolved cellulose will lequire nearly 1^ lb, of the zinc and 1^ lb. of the magnesium salt. The decomposition of the viscose has the effect of coating the fibres with a film of gelatinous cellulose, which exerts a powerful cementing action in the subsequent processes of making and drying, the effects reaching a maximum when the paper is 'loft dried,' i.e. dried in the air without heat. By viscose sizing the contents of . the beater are still left in the basic condition, and it is necessary to add alum in 192 PAPER-MAKING. slight excess in order to facilitate the working on the machine. The quantity of viscose used is usually from 1 to 5 p.et. (calculated as dissolved cellulose) of the weight of the paper. The effects produced are a considerable increase of tensile strength and resistance to the action of water, though the sizing in the latter respect is not of the character of rosin sizing. To produce the quality of resistance to penetration by water and ink the addition of rosin is necessary. It is found, however, that very much less of the latter is required, viz. from ^ to |- the usual quantity, or 1 to 2 p.ct. on the paper. These advantages of the viscose- or cellulose-sizing are realised at only a small additional cost to that of xhe ordinary plan ; but certain disadvantages have to be reckoned with, resulting from the introduction of alkaline sulptur compounds into the beater. These, by interaction with ihe fibre constituents, and also by action upon iron compounds unavoidably present, cause a certain lowering of colour ; and although the sulphur introduced is for the most part elimi- nated (as CS2) by the special action of the decomposing salts, and the alkalinity is of course easily neutralised by alum, there is a residual permanent discoloration suflSciently marked, to prevent the process being used in the ordinary production of" white papers. Aluminate of Soda has been used to a certain extent as a sizing agent in conjunction with the rosin-soap. Hydrated alumina dissolves in caustic soda by virtue of its acid function — a characteristic which it is important to bear in mind in considering the probable chemical condition of the oxide in the finished, paper. The alumina is precipitated from its alkaline solution by the weakest acids, e.g. carbonic acid — and therefore a a /or/2ori by alum itself. It has been proposed to use the aluminate of soda as the solvent for :he rosin ; hut it is difficult to see any advantage resulting from the practice. The addition of hydrated alumina to the neutral rosin size, during the later stages of the boiling, is rather to be recommended. SIZING, LOADING, COLOURING, ETC. 193 Silicate of Soda. — Silica, or SiOg, is a solid body of weakly acid properties which is precipitated from solutions of its alkaline salts on the addition of acids and acid salts. It takes the form of a gelatinous hydrate which, when formed in the beater, enables the pulp to carry more water in the machine, and produces a hardening effect on the paper. In using this compound the more 'acid' silicates should be selected, e.g. the silicates of formula Na20.2Si02, NagO.SSiOa. Gelatin is little used in engine sizing. Used in this way it fails to give results proportional to cost. Its compound with the sulphonated lignone derivatives, the by-products of the bisulphite processes, has been already mentioned. It is soluble in weak alkaline solutions, in which form it is added to the beater, and on being decomposed with the equivalent of alum is reprecipitated as a tough caoutchouc-like mass, which, as a coating upon the fibres, produces a considerable toughening effect upon the paper. Casein. — This albuminoid substance, obtained from milk — of which it constitutes nearly 5 p.ct. by weight — is largely used as a sizing agent in the coating of papers. Its use in the engine depends upon the fact that it can be added in the form of a strong solution, from which it is completely pre- cipitated by alum. It imparts toughness and a good ' handle ' to the finished paper. Its use is limited by cost, which at present rates is between 4d. and 5d. per lb., a cost consider- ably above the selling price of the bulk of ordinary paj)ers. Particulars of Materials used in Sizing. The following are the more important points which determine the value of the materials, their preferential selec- tion, mode of use, etc. Soda (carhonote) occurs commercially in various forms : — Crystal Soda (NaaCOglOHgO) is the ordinary ' washing 106 180 soda.' The crystalline form is a guarantee of purity. From . 106 the formula it is seen to contain — - x 100 = 37 p.ct. only 0 194 PAPER-MAKING. of the carbonate, the remainder being water. It is there- fore a relatively costly form of soda to handle. Crystal Monohydrate (Na2C03 . HgO) was introduced many years ago by Messrs. Gaskell, Deacon & Co. It has one point of advantage over the anhydrous carbonate or soda ash, viz. that it dissolves readily without tending to aggre- gate to lumps. Soda Ash (NagCOg, anhydrous and nearly pure). — Before the introduction of the Solvay process and the so-called ' ammonia ' soda, soda ash was an impure form of the car- bonate. The degree of impurity was such that the percent- age of carbonate in the ordinary forms varied from 83 to 90. Such products were objectionable for use in the making of size. Now, however, we have in the aforesaid ' ammonia * soda a remarkably pure product, the percentage of carbonate amounting to 97-98. This answers all requirements for size making. Solutions of Sodium Carbonate. Specific Gravities of Solutions of Sodium Carbonate, At 60° F. = 15° C. Pegrees Twad. 1° = 0-005 Sp. Gr. Percentage by Weight. i Degrees j Twad. 1 I° = 0-005 1 Sp. Gr. Percentage by Weight. NagO. NaaCOs. HagO. 1 0-28 0-47 16 4-42 7-57 2 0-56 0-95 I 17 4-70 8-04 3 0-84 1-42 18 4-97 8-51 4 1-11 1-90 19 5-24 8-97 6 1-39 2-38 20 5-52 9-43 6 1-67 2-85 21 5-79 9-90 7 1-95 3-33 i 22 6-06 10-37 8 2-22 3-80 23 6-33 10-83 9 2-50 4-28 24 6-61 11-30 10 2-78 4-76 25 6-88 11-76 11 306 5-23 26 715 12-23 12 3-34 5-71 27 7-42 12-70 13 3-61 6-17 28 7-70 13-16 14 3-83 6-64 29 7-97 13-63 15 4-16 7-10 30 8-24 14-09 SIZING, LOADING, COLOURING, ETC. 195 Alum : Sulphate of Alumina. — The basis of the alums i8 the normal aluminium sulphate Al^SSO^; the pure sulphate is acid to litmus. Solutions 'of Aluminic Sulphate.* — The solubility of the anhydrous sulphate (Alg.S.SO^) in water has been deter- mined as follows : — Temperature 0° C. 0° 10° 20° 30° 45° 50° ^^ofT^sTso''^!?''^!?} 3^"^ 40-4 45-7 52-1 The following are the specific gravities at 15° of solutions of varying concentration : — Percent. Al 2.3. SO^ .. 5 10 15 20 25 Specific gravity .. .. 1-0569 1-1071 1-1574 1-2074 1-2572 Crystal Alum {Al^ . . 4S0, + 24H2O) is a double salt of aluminium and potassium sulphates, the latter being inert and serving only for procuring a ready crystallisation of the sulphate of alumina. Of this, the 'active' sulphate, the double salt contains 36-1 p.ct. The double salt may be regarded as a pure but diluted form of sulphate of alumina. It is still used in certain mills, but is for the most part superseded by the ' alum cakes.' The following is a table of strengths of alum solution at 17° 0. :— Percentage of Specific Degrees Al2K2CS04)4.24H20. Gravity. Twad. 1 1-0065 1-3 2 1-0110 2-2 3 1-0166 3-3 4 1-0218 4-4 5 1-0269 5-4 6 1-0320 6-4 Alum Cahes. — Many years ago an aluminous cake or ' concentrated alum ' was introduced into the English paper made by Messrs. Pochin. Such products have now almost entirely displaced the crystal alums. Tliey are made by the action of sulphuric acid upon bauxite or iron-free clays. * Die. Chem. Sol., A. M. Coney. London, 1896, Macmillan. o 2 196 PAPER-MAKING. The subjoined analyses of commercial products show the standard of purity which they now maintain : — Alumina (AI2O3) .. 14-84 14 70 14-95 14-85 16-00 16-20 Ferric oxide (Fe^Oj) •06 12 -05 trace nil trace Sulphuric acid (SO3) 35-00 34 60 36-09 34^94 38-00 38-00 Free sulphuric acid . •32 40 nil •29 nil nil Lime(OaO) .. .. •11 11 •17 •14 •16 •14 Water 49-42 49 95 48-72 49-60 45^50 45 43 99-75 99 88 99^98 99-82 99^66 99^77 These forms of the sulphate of alumina have this advan- tage over the crystal alums: (1) in 'concentration' or per- centage contents of the sulphate ; and (2) in being much more readily soluble in water; the sulphate, in fact, is * soluble in twice its weight of cold water, whereas potash alum requires 18 parts for complete solution. In addition to the purer forms of the sulphate there are a number of products in the market containing more or less of insoluble matters, representing residues of the original or aluminous mineral not attacked by the sulphuric acid. The amount of alum or sulphate of alumina added to a pulp is largely in excess of the quantity necessary to pre- cipitate the rosin soap ; as a matter of fact, in the case of esparto or straw pulps, for the bleaching of which considerable quantities of bleach have been employed, and which therefore contain a certain amount of basic lime, together with calcium chloride, complete precipitation of the size is eflfecfed without the addition of alum. A certain amount is also required to precipitate the starch. The excess of alum appeals to be necessary, however, not only to brighten the colour of the paper, but also to render it capable of resisting the action of ink. From experiments made by the authors, it appears that one part of rosin requires 2 • 9 parts of alum for complete precipitation from its solution in soda. One part of starch requires 0 • 40 part alum. Mineral Filling and Loading Agents.— It is usual, except in the case of papers of the very highest quality, to add to the pulp a quantity of some relatively cheap mineral SIZING, LOADING, COLOURINJ, ETC. 197 loading material, such as china clay, or for certain qualities of paper, ' pearl-hardening,' or sulphate of lime. The addi- tion of clay in moderate quantity can hardly be looked upon as an adulteration, since it serves to fill up the pores of the paper, thus giving a sheet of closer texture, with a smoother surface and more absorbent of printer's ink, and also enables it to take an improved surface in the subsequent operations of calendering. It also adds to the opacity of the paper— a very important point in book papers. Moreover, it enables the manufacturer to meet the demand for cheap papers with some chance of remuneration to himself. If added largely, it of course tends to weaken the paper. China clay, or kaolin, is sold in the form of large lumps of a white or yellowish- white colour. It is formed by the gradual disintegration of felspar under the action of air and water, and consists essentially of a silicate of aluminium. Its quality depends upon its whiteness, and its freedom, from the coarficr micaceous particles. ' It is usually pre- pared for admixture with the pulp by making it into a fine cream with water in a vessel provided with stirrers ; it is then passed through a fine sieve in order to remove any impurities it may contain, and is then run into the beater. - The clay or other filling material is usually run into the beater as soon as the latter is charged with pulp, so that by the time the beating operation is concluded a perfect ad mixture of pulp and clay is effected. Sulphate of lime, or ' pearl-hardening,' is usually suf- ficiently pure to put direct into the engine. It is made by decomposing a solution of calcium cliloiide with sulphate of soda, and is precipitated as a fine brilliantly white powder consisting of CaSO^ -|- 2H2O. Two distinct forms of precipitated calcium sulphate are met with in commerce, differing from each other by their microscopical features, the one consisting of flat tabular crys- tals (Fig. 42), the other of fine needles (Fig. 43). Another form, erroneously called 'precipitated pearl-hardening,' is also sold: it consists of the finely ground native mineral. 198 PAPER-MAKING. The use of these sulphates is attended by the disadvan- tage of their solubility in water — which occasions loss on Fig. 41. Fig. 42. the machine. This is, of course, minimised in long luns on the machine, when the ' back water ' becomes saturated with the sulphate. SIZING, LOADING, COLOURING, ETC. 199 Some of the finer qualities of paper are made without addition of any loading material whatever, though such papers are of course the exception. The proportion of clay or other material that can be retained by a fibre depends to a certain extent upon the nature of the fibre, and upon the 200 PAPER-MAKING. degree of fineness to which it is reduced in the beater. The amount added by different makers varies considerably, from 2 or 3 p.ct. to 20, and even in rare cases to 30 p.ct. Other mineral filling agents having a certain fibrous character have been from time to time introduced, possessing certain advantages over china clay or calcium sulphate. Agalite is a mineral of the nature and chemical properties of asbestos : it consists of nearly pure magnesium silicate. Its structure is more or less fibrous, like that of asbestos, which, as is well known, can be spun and woven and even made into paper, and therefore, when added to a paper, it forms a ptirt of the fabric itself. It is even claimed that it assists in keeping back some of the finer fibres that invariably find their way through the meshes of the wire-cloth, and it is said that 90 p.ct. of the amount added to the engine is found in the paper. In the case of china clay it is well known that only from 40 to 60 p.ct. is often actually ' carried ' by the pulp. Figs. 41, 42, 43 and 44 show the appearance of china clay, pearl-hardening and agalite when viewed under the microscope, magnified 200 times. The nature of agalite is such that it assists the paper in taking a high finish. This is probably due to its ' soapy ' nature, a feature which is characteristic of asbestos, French chalk, 'soap-stone,' and other magnesium silicates. The following analysis of china clays may be cited : — Alumina 39-74 33-44 I nm oxide 0-27 1-04 Calcium 0-36 l-Ol Mfigneaum 0-44 0-16 Silica 4G-32 42-72 Water 12-67 20-93 Sp. gr. varying from 2-5 to 2- 8. The following is an analysis of ' agalite' : — • Singnesia 32-12 30-70 Irou oxide O'lO — Alumina 0-31 — Silica C2-01 61-90 Water 4-30 — Sp. gr. varying from 2*6 to 2-8. SIZING, LOADING, COLOURING, ETC. 201 When writing papers contain such excessive quantities as 15 or 20 p.ct. of clay, it cannot be to the advantage of the consumer, and should be looked upon as an adulteration. Analytical Determination of Loading. It is a matter of some importance to be able to determine rapidly and accurately the amount of mineral matter in a paper. The usual method is to ignite a weighed quantity of the paper in a platinum crucible until the ash so obtained is either white or a very pale grey. From the weight of the ash, the percentage of mineral matter is easily calculated. In certain cases it may be necessary to take into account both the normal ash of the paper and that due to other added mineral matter. This requires an analytical investigation of the total ash obtained, and a careful interpretation of the results. We can only note here that the processes of treat- ment add chiefly lime salts, which are obtained in the ash mainly in the form of carbonate. Tlie ignition may be performed in an open platinum dish, and to hasten the complete combustion of the carbonaceous residues the dish should be covered with a piece of platinum foil, bent in the middle to allow free access of air. A cage of platinum gauze or a closely wound spiral of platinum wire is sometimes substituted for the dish or crucible. The weighings may be perforuied upon any balance of sufficient delicacy for the degree of accuracy required. A spiral balance of the form devised by Joly, and fully described in the First Edition of this book, is a convenient instrument fur rapid work. In brown and coloured papers various coloured mineral loading agents are used ; thus the ochres, which are chiefly oxides of iron, chrome yellows, chieiiy chromate of lead. Ultramarine in heavy blue papers is used in sufficient quantity to have a certain ' filling ' eflfect. These substances or their products of decomposition are 202 PAPER-MAKING. obtained in the ash. If clays or other loading agents are also present, the proportion of pigment-loading may he approximately estimated by separating its characteristic constituent hy suitable analytical treatment of the ash. If the paper contains calcium sulphate, the ash obtained may consist partly of calcium sulphide, due to reducing action of the carbon found on ignition, and the amount will therefore not represent the true amount added. The ash should be moistened with a few drops of sulphuric acid, and again ignited, in order to reconvert it into calcium sulphate. It shouLl also be borne in mind that the sulphate of lime as present in the paper is combined with two atoms of water CaS04+2H20, and therefore that every part of calcium sulphate obtained represents 1-26 parts of ' pearl-hardening ' actually in the paper. Colouring. — The colouring of paper in the paper-mill is effected by the addition of pigments or solutions of dyestuffs to the pulp in the beater. Colouring with pigments is, of course, a purely mechanical operation. At the same time it must be remembered that many of the paper-makers' pigments are reactive chemical compounds, and when acted upon by substances present in the beater, e.g. rosin-size and alum, may be destroyed or modified, with attendant loss of colour or alteration of shade. Attention must be paid, therefore, to the chemical composition of pigments, in order that their full colouring power may be realised and permanently retained in the paper. The soluble dyes are for the most part artificial products of well-defined chemical constitu- tion. An intelligent application of these products by the paper-maker presupposes an intimate knowledge of the general principles of the dyer's art. We may note, however, at this point, that the makers of coloured papers are com- paratively few ; whereas, on the other hand, there are few papers made which do not receive some addition of colouring matter. Thus all bleached pulps retain a residual yellow tint more or less pronounced, and for white papers the yellow requires to be 'neutralised' or 'corrected' by the SIZING, LOADING, COLOURING, ETC. 203 addition of blue and red, producing a 'neutral' or dyed white. In effect these dyed wliites are greys, but the balance or neutralisation of colour produces an illusive impression of whitenes.-. In view of their more general importance we shall deal first with colouring matters and methods used in producing these dyed whites, 1. The following are the more irjportant colouring mattei s : — Blues. — Ultramarines of various shades, tending on the one side to red, on the other to green tones ot blue. The modern ultramarines are artificial products consisting essentially of silica, alumina, soda and sulphur, and are prepared by fusing together mixtures of raw materials furnishing these com- ponents. The coloured body itself may be regarded as a sodio-aluminic silicate, united either with a polysnlphide of sodium or with a polysulphide and polythionate. These relationships are expressed by the following formula : — 2A1203 : SSiOg : AI2O3 : 4Si02 : Na2S203 : SSa-J^. The compound is attacked by the mineral acids, hydrogen sulphide being liberated ; it is also rapidly attacked by chlorine and oxides of chlorine. The decomposition is synonymous with destruction of colour. Smalts. — This is a pure blue pigment obtained by reducing a cobalt-glass to the finest state of division. The glass is essentially a potassio-cobaltous silicate, K20-CoO-6Si02, with varying proportions of alumina and iron oxide as incidental impurities. The deep-coloured varieties contain 6-7 p.ct. CoO; 70-72 p.ct. Si02 ; 17-21 (with NaaO). This pigment is extremely permanent and resistant to chemical action. On account of its high price it is used only in papers of the highest quality. Soluble Dye-Stuffs. — The blues and violets used in tinting for whites are chiefly 'aniline' derivatives. The ^soluble blues ' are the sulphonic acids of phenylated rosanilines. The 204 PAPEE-MAKING. ' methyl violets ' are methylated rosanilines. The induline class of blues are greyish blues of various tones, obtained as by- products in the manufacture of rosaniline. They are oxidation products of aniline. This group includes the nigrosines, Blackley blue, etc. Methylene Blue is a useful fast blue, of pure tone, and is obtained from dimethyl-paraphenylene-diamine by the action of sulphides and subsequent oxidation. It contains sulphur as an essential component. Reds, of the pigment or lake class, are seldom used in tinting for whites. Soluble Bye-Stuffs. — Of these, one of tlie most important is the colouring matter of the cochineal insect. From the raw product a dye liquor may be prepared, either by exhaustion with boiling water or by long steeping of the material with strong aqueous ammonia in a closed vessel. The latter is the more convenient method. After steeping, the mass is thrown on to a conical flannel bag, which is then suspended in a vessel of water. By diffusion, followed by squeezing and washing, the exhaustion of the colour is completed, o lbs. of material treated in this way will make 10 gallons of ' cochineal liquor ' of ordinary working strength. Of the coal-tar dyes, the paper-maker uses (1) of the rosaniline group, the various magentas ; (2) the safranines which are poly-aniline derivations ; (3) azo-dyes, derived from benzidine, especially benzo-purpurine ; and (4) the eosines, which differ essentially from the three preceding groups in containing no nitrogen : they are derivatives of the coal-tar phenol resorcin, and not of nitrogenous bases. Applications. — The use of the above colouring matters in white papers involves, for obvious reasons, very small propor- tions. Of the blue pigments an average proportion is 0*1 p.ct. of the weight of the ' furnish ' ; of cochineal 0*01-0 -02 p.ct., calculated as original cochineal per 100 lbs. furnish. The coal-tar dj es, having greater tinctorial power, are used in still less proportion. There are no general rules which can be laid down as to the stage in the beating process SIZING, LOADING, COLOURING, ETC. 205 when the colour should be added. Where, however, the colouring matter is not affected by the rosin size, it is better to add it while the contents of the beater are alkaline from its addition. The soluble dyes should not be added in the form of the ordinary stock solution, which may contain from 2 to 5 p.ct. of the dissolved dye stuff. This should be diluted considerably before adding to the I eater, and well distributed in adding it. 2. Coloured papers, as distinguished from whites — which are dyed whites — and self-coloured papers are produced with larger proportions of colouring matters. To those used in tinting for whites we have to add other blues, and reds, and also those giving yellows and browns, and also greens. The following are the more important : — Prussian hlues of various grades. These are the ferroso- ferric ferrocyanides formed by decomposing a mixture of ferrous and ferric salt in solution, with a soluble ferro- cyanide. These blues are sold in paste form, under various trade names. They are recognised by being at once de- colorised by the caustic alkalis, the basic iron oxide being left as a brown residue. These blues are often formed by bringing together the above-mentioned salts in the beater, i.e. a mixture of ferrous and ferric sulphates (copperas and 'red iron liquor') added first and followed by the equivalent quantity of potassium ferrocyanide (yellow prussiate). Dyeing with prussian blue in this way is carried out after the sizing operations are complete. It may be noted here that all paper ' furnishes ' contain iron compounds as an impurity, in greater or less degree, and in some cases this is taken advantage of for con- version into the blue ferrocyanides by addition of the potas- sium salts to the beater in correspondingly small proportion. Logwood. — The aqueous decoction, or the concentrated extracts sold in the trade, are used in producing the darker shades of blue, usually in admixture with Prussian blue • also for a larger number of heavy mixed shades and for blacks. 206 PAPEK-MAKING. Alizarin (madder). — This most important red dye is not used directly in colouring papers. It occurs in the form of pigments or insoluble lakes, which are compounds of the colouring matter with alumina, etc. Alizarin printed cottons and 'Turkey red' rags may be beaten after the alkali treatment to a coloured pulp. Iron oxide pigments cover a wide range of gradations of colour from the reds (rouge) to browns (the brown ochres), and from these to the yellows of the ' yellow ochres.' Yellows. — Of the pigment class the most important is the brilliant chromate of lead. This is usually formed in the beater by the interaction of lead acetate and the alkaline bichromates. Of coal-tar dyes the more important yellows are Auramine, a basic dye-stuff, a derivation of dimethyl-aniline ; naphihol yellow, which is the alkali salt of di-nitro-naphthalene ; primuline, a sulphur-containing derivative of toludine ; and a large number of azo and amido-azo derivatives, such as the tropaeolins, chrysoidines. Greens. — The mineral green pigments are but little used by the paper-maker. Mixtures of blue and yellow pigments, e.g. Prussian blue and lead chromate, are however employed. The more important coal-tar dyes are malachite green, a derivative of rosaniline, and ' acid ' greens, which are the sulphonic acids of the group of the ' malachite ' or benzal- dehyde greens. Browns. — In dyed browns the paper-maker uses a number of pigments of the iron oxide group. Of soluble dye-stuffs an important group is furnished by the ' natural ' astringents, or tannin derivatives, e.g. catechu and gambir. The catechu liquor is prepared by boiling the finely powdered raw material for 2-3 hours with 3-5 times its weight of water. When cold, a solution of sulphate of copper, calculated to 6 p.ct. of the weight of the catechu, is added. In using these tannin colouring matters, bichro- mate of soda is used to develop or intensify the brown, iron salts for deepening or blackening the shades. SIZING, LOADING, COLOURING, ETC. 207 A medium brown is obtained with 4 p.ct. catechu on the furnish, ' developed ' with one-sixth its weight of the bichro- mate. Of coal tar dyes there are a large number of browns. The well-known Bismarck browns are salts of tri-amido-azo- benzene. Applications. — The dyeing of paper pulps in the engine is an art which can only be said to be grasped when the whole of the factors involved are intelligently followed. To do this it is necessary to take into account all that has been advanced in the earlier chapters as to the constitution of the celluloses, their general inertness, accompanied by graduations in chemical reactivity such as obviously affect their relation- ships to colouring matters, the combination with which is undoubtedly a chemical phenomenon. Therefore also the constitution of the colouring matters themselves must be taken into account. Lastly, owing to the inertness of the celluloses and more especially those of the cotton group, the cotton dyer is limited to comparatively few dyes or groups of dyes which combine directly with the fibre : in the major- ity of his processes he requires the assistance of intermediary substances known as mordants — compounds which combine directly with the fibre-substance and the colouring matter, and which therefore in combination with the former enable it to take up and fix the colour in the insoluble condi- tion. Owing to the development of the industry in coal-tar colours, and the enterprise of manufacturers, the paper-maker is certainly saved much trouble by being supplied not only with a practically infinite range of colours, but with work- ing specifications of methods of using them, including par- ticulars as to the mordants required. In view of this exten- sive specialisation, we make no attempt to treat the subject in detail. The student must be referred to such special treatises as the following : — (1) On the general principles of dyeing : ' The Dyeing of Textile Fabrics,' by J. J. Hummel (London, Cassell & Co.). (2) On the constitution of the colouring matters : ' The 208 TAPEE-MAKING. Chemistry of the Coal Tar Colours,' by Benedikt and Knecht (London, G. Bell & Sons); and (3) For working formula? for the colouring of pulp in the engine, those contained in J. Dunbar's ' The Practical Paper-Maker' (London, 1887, Spon) are useful and typical of a wide range of effects. The paper colourist may either work on purely empirical lines, i.e. by means of recipes and specifications supplied to him ; or he may by his own investigations devise endless combinations to produce special effects, or produce the usual effects by specially economical means. In either case it may be pointed out that since uniformity is of the utmost import- ance, all conditions of preparation and treatment must be kept rigidly uniform. In matching colours, allowance must be made for the con- dition of the stuff in the beater, and alterations of shade which take place during the process of making into paper — (1) by incidental variations in the composition of the back- water in the machine ; and (2) by the action of the heat of the drying cylinder. The usual practice of matching is to make up a sheet of paper in a hand-mould from the stuff in the beaters, which is then pressed and dried on a hot cylinder or steam pipe. The examination of the colour by compari- son with the sample requires care and attention. A good ' white ' light must be chosen, preferably daylight in a north aspect. Matching at night requires that the artificial light shall be of exceptional purity. During the time that the loading, sizing and colouring processes have been going on, the pulp has been continually acted upon by the roll, and if these operations have extended over a considerable time, it is probably in a proper condition for making into paper. The amount of 'beating' depends, as has been stated before, upon the nature of the fibre, and also to some extent on the nature of the paper for which it is intended. The 'beaterman' examines the pulp from time to time by taking a portion from the engine and placing it in a hand-bowl containing water : from its appearance when so SIZING, LOADING, COLOURING, ETC. 200 diluted he is able to judge of the time during which it may he necessary to continue the disintegration. As soon as this is completed, the pulp is ready to he let down to the stuff- chests, usually placed at a lower level than the heaters, so that the pxilp can flow into them hy gravity. For this purpose the valve at the bottom of the engine is opened : to remove the last portion of pulp it is necessary to rinse out the engine with water. p 210 PAPER-MAKING. CHAPTEE IX. PAPER MACHINES; HAND-MADE PAPER. The pulp as it comes from the beaters is now ready to be made into paper. We will first consider briefly tlie manu- facture of hand-made paper. It is made on a mould of wire-cloth, with a movable frame of wood, called the ' deckle,' fitting on to the outside of the mould, and extending slightly above its surface. The wire-cloth is generally supported by a much coarser wire-cloth, or by pieces of thick wire, and these again by wedge-shaped pieces of wood, the thin end being next to the wire. To form a sheet of paper the workman dips the mould with the ' deckle ' in position, into a vat containing the pre- pared pulp diluted with water, lifting up just so much as will make a sheet of the necessary thickness. As soon as the mould is removed from the vat, the water begins to drain through the wire-clotli, and to leave the fibres on the surface in the form of a coherent sheet of paper. The felting or inter- twining is assisted by lateral motion in every direction given to the frame by the workman. The movable deckle is then removed, and the mould, with the sheet of paper, eiven to another workman, called the 'coucher,' who turns it over and presses it against the felt, by this means transferring the sheet from the wire to the felt. In the meantime the ' vat-man ' is engaged in the formation of another sheet with second mould. A number of the sheets thus formed are piled together, alternately with pieces of felt, and when a sufficient number PAPER MACHIJTES; HAND-MADE PAPER. 211 have loeen obtained, the whole is subjected to strong pressure, to expel the water. The felts are then removed, and the sheets again pressed. They are then sized, if required, by dipping them into a solution of gelatine : again slightly pressed, and hung up on lines or poles to dry. Such paper is called loft-dried. When dry the sheets of paper are calendered. (See Chapter X.), plate-glazed or otherwise j&nished. The making of paper by hand involves considerable dex- terity on the part of the workman ; on account of the expen- sive labour necessary, in comparison with paper-machines,, it is comparatively little practised in the present day ; cer- tain kinds of paper, however, such as bank-notes, various drawing papers, and printing papers intended for the pro- duction of very elaborate editions are made in this way. Any pattern or name required on the paper is obtained by means of a raised pattern on the wire-cloth mould: con- sequently, less pulp lodges there, and the paper is propor- tionally thinner, thus showing the exact counterpart of the pattern. Such devices are known as ' water-marks.' (See also p. 222.) The Paper Machine.— paper machine of the present day, with all its ingenious improvements, differs but little in principle from that originally constructed by Four- drinier. It consists essentially of an endless mould of wire- cloth, on to which the prepared pulp flows, and on which a continuous sheet of paper is formed. The sheet of paper then passes through a series of rollers and over a number of heated cylinders, where it is completely dried. A modern Fourdrinier paper machine is shown in side elevation and plan in Plates I. and II. The pulp, after leaving the beaters, passes into a large vessel called the stuflf-chest, of which there are one or more to each machine. In emptying the beater water is run in to thoroughly rinse out the remaining pulp, the washings also going into the stuff-chests. These may be made either of wood or iron, and should be provided with arms fixed p 2 212 PAPER-MAKING. on a vertical shaft, made to revolve by suitable gearing. The arms are for the purpose of keeping the pulp thoroughly mixed, and should only work at a moderate speed, otherwise they are liable to cause the fibres to form into small knots or lumps. The pulp is drawn from the stuff-chests by means of the pump A, and is discharged into a regulating-box (not shown). The object of this box is to keep a regular and constant supply of pulp on the machine. It consists of a cylindrical vessel, having two overflow-pipes near the top, and a dischnrge-pipe near the bottom. The pulp is pumped in thi'ough a ball-valve in the bottom, in larger quantity than is actually needed, the ex- cess flowing away back into the stuff-chests, through the two overflow-pipes. By this means, the box is always kept full, and therefore the stream of pulp issuing out of the bottom pipe is always under the same pressure. It flows from this pipe, the quantity being regulated by means of a cock, ac- cording to the thickness of paper required, directly on to the sand-tables. These may be of various sizes and shapes, but should be so large that the pulp takes some little time to travel over them. They consist of long shallow troughs, generally of a sinuous form. The bottoms are sometimes . covered with* woollen felt, or with thin strips of wood placed across the direction of the flow of the pulp, and at a slight angle. These serve to retain any particles, such as sand and dirt, that may have escaped removal in the previous treatment of the pulp, and that are heavy enough to have sunk down during the passage of the pulp along the sand-tables. In some mills, where great care is exercised, the pulp is caused to flow over sand-tables 200 yards in length.- As the pulp, when it leaves the 8tuff"-chests, does not contain sufficient water for the purpose of making paper, it is mixed, where it enters the sand-tables, with a quantitj' of water from the ' save-all ' (see p. 221), flowing from the box B placed at a higher level. In some mills, instead of being pumped into the regulating- box, the pulp flows into a small vessel below the stufl'-chest, Back of Foldout Not Imaged Back of Foldout Not Imaged PAPER MACHINES ; hand-:made papek. 213 and is lifted on to the sand-tables by means of buckets fastened on the circumference of a wheel. The pulp, after leaving the sand-tables, passes on to the strainers. These consist of strong brass or bronze plates, having a large number of very fine V-shaped slits cut in them, the narrowest end being on the outside. Strainers. — The strainers are for the purpose of removing from the pulp all lumps formed by^the intertwining of the fibres, and all pieces of unboiled fibre, which, if allowed to pass on, would show in the paper as inequalities in the sur- face, or as dark specks. The slits are made narrow at the top, and gradually increasing in width, so as to prevent them from getting choked up. These slits allow only the individual fibres to pass through, and their width varies ac- cording to the quality of the paper. They are from 2 to 3 inches long, and they vary in width from -007 to -05 inch. They are put at distances of about ^ inch apart. Several plates, each containing about 500 slits, are bolted together and form a strainer. The whole strainer receives a violent shaking motion, to assist the passage of the fibres through the slits. In the machine represented, two of these strainers are shown at C. The shaking motiun is produced by the ratchet-wheel or cams a acting on the hammer h. An im- proved form, called the ' revolving strainer,' has of late years been introduced. The pulp generally passes first through one of these, and then through the ordinary or ' flat ' stiainers, as they are called. A revolving strainer is shown at D. It consists of a rectangular box, the sides of which are formed of plates perforated with slits. Inside this box, a slight vacuum is formed by means of an indiarubber bellows worked by a crank on the shaft d. The vacuum is intended to serve the purpose of the shake in the ordinary form. The box revolves slowfy inside a vat con- taining the pulp, and the strained pulp flows into the box D\ and thence on to the flat stJainers. Various patents have been taken out from time to time for flat strainers worked by means of a vacuum underneath the 214 PAPER-MAKING. plates caused by the motion of discs of indiarubber or thin metal. Fig. 45 shows in plan a set of strainers, as manu- factured by Messrs, G. & W. Bertram, similar to those in Plate II., but illustrated somewhat more in detail. The pulp first passes through the flat strainer B, and from Fig. 45. thence to the two revolving strainers A. From these it flows along the shoots placed at the side on to the paper machine at E. Fig. 46 gives a view of a patent flat vacuum strainer made by the same firm, which may also be used for cleaning straw or esparto previous to its passage over a presse-pdte. The pulp flows on to the strainer at a, and passes away by PAPER MAQHINES; HAND-MADE PAPEK. 215 the cast-iron pipes /. The valve g is for running off waste pulp. The plates are placed at a slope of about 1 in. in their length ; those nearest the supply of pulp are provided with coarser slits, as the impetus carries the knots forward. The vacuum pumps are woi'ked by the rods d from the shaft e. By means of the tube c water can be directed on to the plates, whereby the coarser particles of fibre are carried forward, and the slits- are kept clean. The plates can be removed in a few minutes. Fig. 46. Figs. 47 and 48 show in side and end elevation, Messrs. Masson, Scott & Bertram's'patent self-cleaning strainer. [The novelty consists in an arrangement by means of which the upper surfaces of the plates are continnally freed from those p6rtions of the pulp which cannot pass through the slits. The scrapers d are made of vulcanised indiarubber, and are continually carried forward by an endless chain; the knots, &c., collect in a heap at the end of the strainer. The pulp flows on at a, and passes away through the pipe* b. The pumps are driven from the shaft e. 216 TAPEE-MAKING. Messrs. C. H. & F. L. Eoeckner's patent strainer (No. 7932, 1885) consists of a series of cylindrical tubes, open tit one end and perforated with slits. They are placed in a vat into which the pulp flows. Inside the cylinders are placed two, three, or more plates fixed to the shafts on which the cylinders are supported, and extending to the circumference. These plates form a kind of fan, which, together with the cylinders, are caused to oscillate by means of a rod and cranks. This oscillating motion serves to draw the fibres through the slits, and at the same time to keep the outsides of the cylinders clean. The cylinders are easily removable. I , 1,1. .11, I III I Fig. 47. Fig. 48. After a time, the slits in the plates get too large, owing to the plate having been worn away by the constant friction of the fibres, and as they are very expensive, various attempts have been made to invent plans for partially closing them again. Hammering will efiect this, but is liable to break the plates. Annandale of Beltonfoid has introduced a method of closing the plates, by means of heavy pressure acting on small steel rollers moving on each side of the slit, in which is placed a small sheet of metal the exact thickness of the width desired. Another method of closing the plates consists in filling PAPER MACHINES ; HAND-MADE PAPER. 217 them up by means of electrically deposited copper or other metal. They can then be recut in the usual way. In the case of revolving strainers, all that cannot pass through the slits falls to the bottom of the vat, in connection with which it is customary to have an auxiliary strainer, or 'patent knotter,' as it is called, shown at E. All fibre that passes through this one, which is of the ordinary flat kind with shaking motion, goes into a box near E', called the ' low box ' for ' save-all ' water (see p. 221). The pulp, after passing through the strainers, should be perfectly free from knots and impurities, and in a fit condi- tion for making paper. In the machine shown, it passes from the last strainer directly on to the wire, its flow being regulated by a movable gate e. In f-ome cases, however, it first flows into a small vat, in the centre of which revolves a rod carrying paddles, with the object of keeping the pulp well stirred up. It is carried right on to the wire by means of the apron, a piece of canvas, oil-cloth, or sheet rubber, one end of which is fastened to the breast-board e', the other end resting on, and covering the wire to the extent of about 15 inches. The edges of the apron are rolled up to prevent the pulp from overflowing. After leaving the apron, it passes under a gate, or ' slicer,' as it is sometimes called, made of two pieces of brass, overlapping each other in the centre, and bolted together. It is made thus to enable it to be lengthened or shortened according to the width of the paper ; its height from the wire-cloth can be altered by means of screws, and should be equal at all points, in order to ensure a uniformly thick sheet of paper. The ends of the two pieces i'orming the slicer are fastened to the frame / or 'deckle,' as it is called, and this again is carried by two or more rods stretching right across the wire, and fastened by small upright supports on both sides to the frame g. The deckle-i'ranie also carries the grooved pulleys 7*, along which the deckle-straps i, endless square bands of indiarubber, move. The object of the deckle-straps is to regulate the width 218 PArER-MAKlNG. Fig. 49. of the paper ; they form, together with the wire- cloth, a kind of mould into which the pulp flows, thus corresjiond- ing to the mould used in producing hand-made paper. The width of the paper can be altered by shifting the position of the frame /, and also the deckle-fetraps, which are carried on it as de- scribed, the pulleys h being so arranged that they slide along the rods on which they revolve. In order to alter the width of the paper it is necessary to stop the flow of pulp on to the wire, and it not only consumes a con- siderable amount of time, but generally necessitates a partial cleaning up of the machine. Vaiious at- tempts have, therefore, been made to devise an arrangement whereby the change can be effected while the paper is being made. Several contrivances have lately been introduced, all similar to the one shown in Figs. 49 and 50. PAPER MACHINES; HAND-MADE PAPER. 219 la it the frame / carrying the deckle-strap is made to slide along the rods e by means of the small wheel 6, and by a similar arrangement on the opposite side and geared with it. The movable apron I, Fig. 49, is wound round the spindle g, and is kept taut by the cords m connected with the springs n. As the deckles approach each other, the excess of apron is wound up, when they are separated it unwinds ngain. The flow of pulp is regulated by two slices a a, which are kept in position by the screws h. The whole arrangement is securely bolted to the frame of the machine; c (Fig. 50) represents the bieast-roll, and corresponds to F, Plates I. and II. The thickness of the paper is regulated by altering the Fig. 50. supply of pulp to the wire-cloth, and by the speed at which the machine is working. This speed may vary from about 60 feet to as high as 500 feet per minute. The ' wire ' is an endless cloth made of very fine wire, the fineness depending much on the quality of the paper required. The mesh varies from 60 to 70 and even more threads to the inch. It is not woven endless, but is joined by very careful sewing with wire. Its width varies considerably, some being made as wide as 126 inches; the length is generally 35-40 ft. It is carried by the breast-roll F, the lower couch- roll G, and the small rolls /, and by a large number of small rolls f. The latter and the breast-roll are supported by the fiame g, while the small rolls /' are supported by 220 PAPEE-MAKING. brackets attaclied to it. The course of the wire is indicated by the arrows. The frame g works on two pivots cj, and le- ceives a shaking motion from side to side from the rod in connection with a crank worked by two conical drums H. The supports g" are also pivoted at their lower ends to allow for the shaking motion. This shaking motion is given for the purpose of weaving or intertwining the fibres. One or more of the rolls /' can be moved up or down on the support which carries it, for the purpose of stretching the wire. There is usually a large number of the small rolls /", as it has been found by experience that, probably owing to capillary attraction, they cause the water to leave the pulp. Though a large quantity of water thus passes through the wire-cloth, it is necessary to assist it by artificial means. This is done by means of the suction-boxes 1 connected by pipes with the vacuum-pumps I'. This part of the machine, which is called the " wet-end," is placed at a slight slope of about 1 in. in its entire length, the lowest end being nearest the strainers. Underneath the wire-cloth is placed a box called the ' save-all ' K, connected with the box E'. The water that passes through the wire-cloth contains a considerable quan- tity of very fine fibres, together with size, alum, clay and colouring materials, that have passed through the wire, and which would be lost but for the arrangement now universally adopted. It flows into the box E', and is pumped, together with the pulp that has passed through the knotter E (see p. 217), into the high box B, whence the mixed stuff flows on the sand- tables, to be again used to dilute fresh pulp from the stuff-chests. The following numbers will give some idea of the nature and amount of fibres, &c., which pass through. The paper was made from esparto and straw, sized with rosin and starch. It contained 12 p.ct. of clay. Grains per Gallon of Waste Water. Fibre Clay.. Starch .S4-37 37-10 1-40 PAPER MACHINES ; HAND-MADE PAPER. 221 It is almost impossible to utilise the whole of the back- water passing through the wire-cloth in the way described. In some mills a portion of this water is made to pass through a ' pnlp-saver,' such as is shown in Fig. 51 and 52. It consists of a conical drum A, the circumference of which is covered with wire-cloth, and it is caused to revolve slowly by suitable gearing. The water enters by the pipe B, and passes through the meshes ef the wire-cloth, the pulp gra- dually finding its way to the wider end, where it is discharged into the box C. It can then be returned to the beaters. The pulp-saver can also be used for recovering the fibre from other waste water, such as the wash water from the Fig. 51. Fm. 52. washing and beating-engines ; or it can be used for freeing bleached pulp from water in cases where drainage or hy- draulic pressure is not resorted to. If any pattern or name is required on the paper, it is pro- duced by means of a light skeleton roll, call a ' dandy-roll,' covered with raised wires in the form of the desired pattern, placed between the suction-boxes, and pressing lightly on the still moist paper. The paper is thinned where the wire pattern presses, and thus a mark (water-mark) is produced. The other side of the paper has a mark corresponding to the wire-cloth ; by using a dandy-roll covered with wire-cloth, the two sides can be obtained alike, such paper going by the name of * wove.* 222 PAPER-MAKING. Paper in which a series of parallel lines are produced is called a ' laid ' paper. Some water-marks are produced by means of a dandy-roll in which the pattern is formed by depressions in the surface. The paper is thus thicker where the pattern is formed than in the ground-work. De la Eue (Patent No. 8348, 1884) has patented the use of dandy-rolls so formed as to produce upon the paper the effects both of thickening and thinning. Imitation water-marks can be produced on the fioished paper by subjecting it to pressure in contact with plates on which the design has been produced in relief. In this way very beautiful results can be obtained. (See Patent No. 13,455, 1884.) It sometimes happens that the wire-cloth slips slightly to one side. This can be obviated by the machine-man shifting, by means of screws, one of the rolls provided for the purpose with a movable journal, until its axis is at a slight angle to that of the other rolls. An automatic apparatus has been in- vented for this purpose. Two brass plates are fixed, one on each side of the wire-cloth, to a long rod, connected by suit- able machinery with the screws working the movable journal, so that, as the wire presses against one or the other of these plates, the roll- is shifted so as to correct this. The paper, which, even after passing the suction-boxes, is still very wet, passes with the wire-cloth between the couch- rolls G G'. These are hollow copper or brass cylinders, covered with a tightly-fitting endless jacket of felt. They may also be made of wood (sj'-camore or mahogany) or of iron or iron and brass combined. The pressure of the upper couch-roll upon the lower can be regulated by means of screws or levers. They serve to press out water from the paper, and to detach the paper from the wire-cloth. By dexterous manipulation on the part of the machine-men, the paper is transferred to the endless felt, travelling over the rolls Ic in the direction of the arrows. It is known as the ' wet felt,' from the condition of the paper at this stage. In its passage along this felt, the paper passes between PAPER MACHINES; HAND-TIADE PAPER. 223 two iron rolls K, called the first jiress-rolls, with the object of having the M^ater squeezed or pressed out of it. These rolls are sometimes covered with a thin bi ass case, and the top one is provided with an arrangement called the ' doctor,' iu order to keep it clean, and free from pieces of paper that may have stuck to it. The lower press roll is some- times covered with an indiarubber jacket. The ' doctor ' is a kind of a knife placed along the whole length of the roll and pressing against it at every point. The pressure on the rolls can be regulated by means of levers, or, as in the illustration, by the screw h'. ]t will be readily seen that the under side of the paper that has been next to the felt will, in its still moist condition,' have taken to some ex- tent an impression from the felt, while the upper side Mall have been made comparatively smooth by the pressure against the top roll of the 1st press. In order to make both sides of the paper as nearly as possible alike, it is passed through another set of rolls L, called the 2nd press. This time it is reversed, and enters at the back of the rolls ; thus the other side of the paper is next the metal, being taken through by the felt (called the '2nd press felt') travelling on the small rolls I, the paper, after leaving the wet felt, and before being taken on to the 2nd press felt, travels over the rolls Z'. The 2nd press felt is necessary, because the paper is too tender to withstand, unsupported, the pressure of the rolls. The paper, after passing the 2nd press rolls, travels over the drying-cylinders M, the number of which varies some- what. In the machine shown, there are in all eight cylinders. Sometimes as many as twelve are employed. Between the 2nd press rolls and the cylinders, a passage S allows easy access to the other side of the machine. The paper generally passes alone over the first two, which are only slightly heated ; afterwards it is led over the others by means of felts, as shown. The arrangements shown at R are for the purpose of stretching the felts. The cylinders are heated by means of steam, and are generally divided into two sets, between 224 PAPEH-MAKING. which is a pair of chilled-iron, highly-polisliecl rolls N, called ' sraoothers,' the function of which is sufficiently explained by their name. They are also heated by means of steam. The cylinders are usually made of slightly decreasing diameter, in order to allow for the shrinking of the paper on drying. Messrs. G. and W. Bertram introduce into some of their machines one or two small drying cylinders, over which the 2nd press felt travels, the object being to drive off some of the moisture absorbed from the paper. This contrivance is said to give excellent results, and to ensure a considerable saving in felts. The arrangement is shown in Fig. 53. It Fig. 53. is especially useful in machines running at a high speed, or in those having limited drying power. After leaving the cylinders, the paper should be quite dry; it is then led through the calenders, of which there are in some machines as many as three sets, though only one is shown. These are similar to the smoothing-rolls, just described. Pressure is applied by the screws 0', or by levers and weights. The friction of the hot calenders on the dry paper develops a large quantity of electricity, which occasionally discharges itself in bright sparks. It is the practice in some mills to cool the paper before PAPER MACHINES; HAND-MADE PAPER. 225 passing it through the calenders. This is effected by passing it over a copper cylinder, through which a stream of cold water runs. The degree of smoothness or ' finish ' that can be given to the paper by the calenders, depends to a large extent upon the degree of moisture which it contains. As it leaves the last cylinder it is perfectly dry, that is to say, it contains only that amount of water which cellulose, from whatever source, always carries. This amount varies slightly with the nature of the cellulose, and with the plant from which it has been isolated, and also with the state of the atmosphere, it being greater on damp days. It would be impossible so to regulate the drying action of the cylinders that the re- quisite amount of water should always be left in the paper ; it is therefore better to dry it as thoroughly as possible, and then to add the water, by artificial moans, just before it passes through the calenders. This method, moreover, has the advantage of damping only the surface of the paper. Pig. 54 shows a damping arrangement as manufactured by Messrs. James Bertram and Son. The paper on leaving the last drying-cylinder passes on to two copper cylinders c, filled with cold water. Steam passes through the pipe a, and issues through a number of fine holes in a pipe running at right angles to the direction in which the paper is travelling, and near to it. The steam condenses on the paper and on the surface of the cylinders, from which the paper absorbs it. The supply can be regulated by the cocks shown in the drawing. The troughs d carry off any excess of condensed water. Amongst other methods proposed is one by Annandale ; it- consists in breaking up jets of water into a very fine spray by means of a blast of air. The finished paper, after passing through the calenders,, is wound on the reels P. The gearing by which the whole machine is driven is shown in Plate II. It sometimes happens that, owing to the increased tension due to the contraction of the paper in drying, the paper Q 226 PAPEE-MAKING. breaks. It is therefore necessary to alter the speed of tome of the cylinders to compensate for this. A rough expedient which is largely adopted, consists in attaching, by means of a mixture of rosin and tallow, a piece of felt to one or other of the pulleys (Plate II.), and thus altering its speed. Tub-sizing. — The foregoing description is of a machine for the manufacture of engine-sized papers ; some slight modifications are necessary in the case of papers that are tub-sized on the machine. In making the cheaper qualities Fig. 51 of tub-sized papers, the paper, after being partially dried over a few cylinders, is passed through a vessel containing a solution of gelatine (see p. 228). It then goes between a pair of rollers, which press out the excess of tize, and then again over drying cylinders. The vessel or trough contain- ing the size is filled to overflowing fi-om a tank placed at a higher level ; the overflow passes into a lower vessel, from which it is pumped up to the top tank. The size is kept warm by means of a coil through which steam passes placed 227 in one or other of tlie vessels containing it. The paper passes between a pair of rollers placed in a trough and dipping into the size. The size may be made to flow through a pipe pierced with a number of holes, the stream running directly on to the web of paper. In the other, and perhaps most gene- ral way, at least for the better qualities, the paper is wound off immediately after leaving the last drying-cylinder A (Fig. 55), and sized at some future time, or it may be passed directly into the vat B containing the size. After passing between the squeez- ing-rolls 0, it is generally wound off as at D ; and after having stood some time to allow the size to be evenly absorbed by the paper, it is wound off from F, and passed over the cylinders H, of which there may be a very great number, some machines having over 300 of them. These cylinders are made of light spars of wood ; inside them, and revolving rapidly in an opposite direction, are the fans G. The paper in travelling over these drums, is only slowly dried, and is Q 2 228 PAPEr-MAKING. supposed by this means to be more perfectly sized, and in- creased in strength. It is wound on to reels again at I. Only the first and last two drums are shown. This method was devised to imitate, as far as possible, the sizing process of hand-made paper. Even now, paper that has been made on the machine is sized by hand, alter having been cut into sheets, much in the same way as hand-made. Preparation of Size. — Very pure gelatine can now be obtained in the form of colourless sheets. It is, however, more economical for the paper-maker to prepare his own : this is therefore done in the majority of cases. A great many animal substances, such as clippings of hides, horns, bones, Ac, yield gelatine when heated with water. Any of these substances may be used, the first being the chief raw mate- rial. They are first softened by soaking in cold water for some days : they are then cleansed by washing in a stream of water. The next operation is that of heating with water. This is usually done in iron or copper vessels provided with a fake bottom or outer jacket, into which steam may be passed. The hide cuttings are covered with water, and the mass gradually raised to a temperature of about 85^ (185° F.). In from ten to fifteen hours nearly the whole of the cuttings will have passed into solution as gelatine. This solution is then drawn otf and any insoluble substance suspended in it is removed by subsidence or filtration. The hides contain a certain quantity of fat, which collects on the surface of the solution. The residue in the boiler is again heated with water, and the solution added to the bulk. It is of the greatest importance that too high a temperature be avoided, as gelatine is liable, when strongly heated with water, to lose its power of gelatinising. To the filtered solution of gelatine, which should be, if properly prepared, of a pale colour, a quantity of alum solu- tion is added. The effect of alum upon gelatine is very remarkable. If added slowly it will be found gradually to render it stiff until at a certain point the mass will become almost solid ; a further addition renders it fluid again. It is PAPER MACHINES ; HAND-MADE PAPER. 229 ia this state ready to be used for sizing purposes. The amount of alum necessary to produce this effect is about 20 p.ct. of the weight of the raw material. The alum is also useful in preventing the decomposition of the gelatine, but its chief (jharacteristic is to render the gelatine a more efficient sizing material. There is no doubt that, besides its action upon the gelatine itself, it has a considerable effect upon cellulose. It is of the greatest importance that the alum or the sulphate of alumina, which can be substituted, should not contain any free acid. This is especially necessary in the case of papers made from rags, in the bleaching of which an acid has been used ; one effect of which is to remove all basic substances derived from the boiling or bleaching processes. In the case of esparto and similar fibres, a considerable quantity of such substances are present in the pulp, the result being that the free acid of alum is to a large extent neutralised, and its injurious effects prevented. The effect of the free acid is seen in the weakening of the paper and the destruction of metal surfaces with which it comes in contact. The effect of acid upon cellulose will be found more fully described in p. 13. Many paper-makers add to the gelatine a certain quantity of soap, the effect of which is to render the paper capable of taking a high finish. The soaps employed should be white and firm, and should be free from rosin. Some soap-makers prepare a special soap for paper-making. They are usually made from tallow, or a mixture of this with a small quantity of cocoanut oil. If a solution of soap be added to one of alum, a double decomposition immediately occurs, the fatty acid being thrown down in combination with the alumina, and the soda combining with the sulphuric acid. If, however, the soap solution be previously -mixed with a solution of gelatine, no precipitation takes place, the mixture forming an emulsion having somewhat the appearance of thin milk. Yarious attempts have from time to time been made to size paper in the engine with gelatine, by precipitating it in 230 PAPER-MAKIlNJG. the pulp after the manner of rosin sizing, but as yet no successful method has been devised. Single-cylinder Machines. — Modified forms of the original Fourdrinier machine have been introduced to meet various lequirenients. One suitable for the manufacture of very thin paper, or of papers one side only of which is required to be glazed, called a single-cylinder or Yankee machine, is shown in Fig. 56. It resembles the Fourdrinier machine as far as the couch- rolls A and B. The paper is taken off the wire-cloth on to an endless felt running round the upper couch-roll A, and travelling in the direction of the arrows. It is taken from Fig. 56. the felt on to the large drying-cylinder C, of about 10 feet diameter, heated with steam. This is carefully turned and polished so as to impart a high gloss to the surface of the paper with which it is in contact. Calender-rolls are some- times supplied in addition. The arrangement shown at D is for the purpose of washing the felt. This is necessary to cool and open it out after passing between the cold press-roll E and the hot cylinder. The paper, after passing over the greater part of the surface of the cylinder, is sufficiently dried, and it is then wound off at F. A machine of a very different construction from the ordinary form is shown in Fig. 57. The pulp, after passing PAPER MACHINKS; HAND-MADE PAPER. 231 through the strainer A, enters the vat B, in the centre of which a large drum or cylinder C revolves. This cylinder is covered with fine wire-cloth, and on it the paper is made. As it revolves the fibres attach themselves to the wire, and the water passes through the meshes, being assisted by means of a pump. The sheet of paper thus formed is taken on to the endless felt passing round the couch-roll D, and travels along with it to the large drying-cylinder E, heated Fig. 57. by steam. It leaves the felt at F, and is then taken on to the cylinder, after travelling round which it is sufficiently dried, and is then wound off as at G. The felt on its return journey passes through the washer H, where it is cleaned and freed from adhering particles by the scraper I. It is squeezed free from excess of water by the rolls K. Paper made on such a machine is weaker than that made in the ordinary way, because it has not been found possible to give a shaking motion to the cylinder, and thus the fibres are only imperfectly felted. A modification of this machine is used for making mill- boards, the difi'erence being that it has no drying-cylinder. 232 PA.PER-MAKING. The felt carrj'ing the paper passes between a pair of press- rolls, which squeeze out the water. The sheet of paper is then allowed to wind round the top press-roll until of the required thickness. When this happens, it is cut off the roll by a knife. The thick sheets so produced are dried either in the open air or in a room heated for the purpose. 233 CHAPTER X. CALENDERING, CUTTING, Etc. The paper, as it leaves the machine, is for many purposes not sufficiently highly glazed, and it is therefore necessary for it to undergo a further process of calendering. This may be done in various ways. One method, called " web glazing," is to pass the paper between a number of rolls, alternately of polished iron and very highly compressed paper or cotton. The construction of such a calender will be understood by reference to Figs. 58 (end elevation) and 59 (front elevation). The reel of paper, as taken from the machine, is shown at A (Fig. 58), its course over the rolls being indicated by arrows. After passing over the bottom roll, it is wound off on a wooden or hollow iron cylinder B (Fig. 58), driven by the toothed wheel shown by the dotted line G, on the same shaft as the wheel D, which is driven by E, keyed upon the bottom roll. The whole ma- chinery is driven by the large toothed wheel F (Figs. 58 and 59), which is itself driven by the small wheel G on the main shaft H. The paper rolls are marked P, and the ii on rolls I. It will be seen that there are two paper rolls in the middle, for the purpose of, as it were, reversing the paper, and so making both sides alike. Pressure is applied to the rolls by means of the screws K, and by the weight L (Fig. 58) acting on the compound lever M. The brake, whioh consists of a strap of leather, pressing, by means of the weight and lever N, on the circumference of the wheel 0, connected by toothed wheels with the cylinder A on which the paper is wound, is used for the purpose of preventing the paper from leaving the cylinder too rapidly. But for this appliance, the paper 234. PAPEE-MAKING. would be apt to crease. The paper rolls liave an inner core of iron, the paper only extending to a depth of about 5 inches. Fig. 58. The iron rolls are hollow, and are connected with steam- pipes, "by which they can be heated. CALENDEKING, CUTTING, ETC. 235 only, is to pass it between a large paper roll and a smaller iron one, the latter revolving at a much greater speed than the 236 PAPER-MAKING. former. By this means a very smooth surface can be obtained. It is sometimes assisted by rubbing a small quantity of bees'- wax on the small iron roll. The above-mentioned methods apply to the glazing of paper in the web. Paper cut into sheets may also be treated in the same way. Various modi- fications of these calenders have been devised ; they do not, however, involve the application of any special principle. Plate-glazing, a method that is adapted to hand-made and ^e/ Fig. 60. the better qualities of paper, consists in applying heavy pressure to sheets placed between polished plates of copper or zinc. The metallic plates and the sheets of paper are made into bundles, and the whole is passed between two strong rolls, heavy pressure being communicated to them by means of screws or levers, and weights applied to the ends of the upper roll. A calender for this purpose is shown in Fig. 60. The bundle of plates and paper is passed along the table a ; after CALENDERING, CUTTING, ETC. 237 passing through the rolls c, it slides along the inclined table d, where it is received by a workman. The pressure on the rolls is regulated by the handle c and the weights b. By passing paper between rolls on which devices have been cut, the 'repped' and other similar papers are produced. With calender rolls of the ordinary construction, as the pressure is applied at the extreme ends, the roll is liable to assume a slightly curved shape, the effect of which is to pro- duce an uneven surface on the paper, the outer portion of the web being more highly finished than the centre. This defect is obviated to a very large extent by Schur- mann's Patent Anti-Deflection Eolls (Fig. 61). The roll proper consists of an. outer shell a, through the centre of which and securely wedged in at h is the centre core c, the Fig. 61. ends of which run in journals, and to which pressure is applied in the usual way. The pressure is communicated to the outer shell at the point of contact h, the result being that the parallelism of the surfaces of the rolls is maintained, and, in consequence, the paper tends to be equally finished in every direction. Cutting. — Except for special purposes, such, for example, as for use in a continuous printing machine, paper is usually sent from the mill in the form of sheets. The form of cutter, called a revolving cutter, generally used, is shown in Fig. 62. The paper from the webs A is drawn forward by the rolls B ; it is then ripped into widths of a convenient size by means of two circular knives, the upper one of which is shown at C. It again passes between a pair of rollers, after leaving which it meets a knife D fastened to the revolving drum E, and 238 PAPER-MAKING. pressing against a fixed knife not shown. The cut sheets then fall upon the endless travelling felt P. The action of the knives will be understood by reference to Fig. 63. The edges of the two knives are shown at A and B. The knife B Fig. 62. has a slot, in which the bolt D slides, and it is kept in position by means of a spring. This spring causes the knife to slide back slightly as it comes against the fixed knife A, The position of the paper is shown by the dotted line C. The knife B is set .on the drum not quite horizontally, so Fig. 63. that one end meets the stationary knife a little before the other, thus acting in every respect like a pair of scissors. Fig. 64 shows a pair of ripping-knives. The upper one A is kept in position against the lower one B by means of the CALENDERING, CUTTING, ETC. 239 spring C. The cutting surfaces are slightly hollowed out, so as to have a sharper edge. The paper is shown by the dotted line D. By altering the relative speed of the drum E and the rolls B, by means of the expanding pulley G, sheets of any desired size can be cut. The cutting-knives are sometimes placed inclined to the drawing-in rolls B, so that the sheet, instead of being cut into a rectangle, is cut into a rhomboid. Such paper is used chiefly for the manufacture of envelopes, this shape occasioning a smaller loss when the envelopes are cut out. Single-sheet Cutter. — It is sometimes necessary, as in the case of paper having a water-mark, that the sheet should be A ] J B Fig. 64. cut with great exactness, so that the device shall come exactly in the centre. The ordinary cutter^cannot be relied on for this purpose, and, in its place, a machine called a ' single sheet cutter ' is used (Fig. 65). It consists essentially of a large wooden drum A, fixed on a horizontal axis, over which the paper is led by a pair of drawing-in rolls B. The paper is held against the drum by a clamp worked by the arm C. The paper is cut by the knife E moving against the stationary knife D. After the cut, the drum describes part of a circle, the paper being still held, so that it cannot go back with the drum. As soon as it has gone far enough, the clamp is re- moved, and the drum returns, bringing the paper with it. The length of the arc through which the drum moves, and 240 PAPEK-MAKING. therefore the size of sheet, is regulated by the length of the crank-arm F, If from any cause the cut should not take place at the right time, the man in charge can, by pressing against the clamp, retard the motion of the paper, and thus bring back the cut to the right place. The small roller G is for the purpose of keeping the paper always tight. Quillotine Cutter. — It is sometimes necessary to trim the edges of sheets of paper or to divide them into smaller sheets. Fig. 65. This is done by means of a guillotine cutter, an improved form of which, as manufactured by Salmon of Manchester, and called the 'Victory' cutter, is shown in Fig. 66. The paper to be cut is placed on the table F, touching the back gauge G, which can be brought backwards or forwards by means of the handle H, thus regulating^the size of sheet. The machine is set in motion by 'means of the lever A, CALENDERING, CUTTING, ETC. 241 which acts upon a friction-clutch connected with the driving- gear. This causes the knife-bar B and the clamp C to descend. When the latter meets the paper its progress is arrested. The knife D, however, continues to descend, and passes through the sheets of paper. By means of slots placed at the back, the knife is made to take a diagonal course. The pres- sure of the clamp is maintained upon the paper during the cutting operation by means of a powerful spring contained in the tube E. Fig. 66. Sorting. — The sheets of paper are now ready to be examined before being finally sent away from the mill. This is done in the ' !Pinishing-house,' or 'Salle' as it is sometimes called. This sorting is usually performed by women, who reject the inferior or damaged sheets. These are sold at a reduced price under the name of ' re tree.' Paper is sold in sheets of definite sizes, and is made up into R 242 PAPER-MAKING. reams containing from 480 to 516 sheets. These sizes corre- spond to different trade names — such, for example, as crown, demy, royal, imperial, &c. The weight per ream is usually expressed in addition to the name, thus : 14— lb. demy, 18— lb. double crown, &c. In this way the consumer is enabled to make a rough comparison of the thickness of the different kinds of paper. In making paper it is the duty of the machineman to examine and weigh from time to time a sheet of the desired size, in order to ensure uniformity. Special lever balances can be obtained, showing at a glance the weight per ream of different numbers of sheets. The following are some of the sizes of sheets as generally specified for writing and printing papers : — Writings. Printings. Foolscap 17 X 13i 17 X 13J Post 18| X 15| — Demy — 22i x 17| Royal — 25 X 20 It is particularly in dealing with such quantities as the above sizes and weights of papers that we become aware of the arbitrary and cumbersome nature of our English weights and measures. It is not for us to argue the question of the superiority of the metric system, but to admit it. Apart from the simplicity of calculations based upon the decimal S3'stem of units, there is the more important simplification of ideas in a notation based upon related quantities. Thus the ton on the metric system is the weight of the cubic metre of water. The English ton on the other hand is related to the yard or unit of length, for all practical pur- poses, as the dimensions of a ship are to the names of the crew ! In the weight-measure specification of papers on the metric system the weight per square metre in grammes is a measure from which we pass by an instantaneous process to relative thickness of all papers having the same specific CALENDERING, CUTTING, ETC. 243 gravity : or by direct measurement of thickness, to the re- lative gravities of papers which vary in this respect. The paper maker in the mill should adopt some uniform standard — preferably the metric system, having reference solely to the physical constants of paper as paper. The arbitrary and conventional weight-sizes of the stationers' trade may be calculated in terms of the systematic standard which should give a direct and simple universal expression to thickness and specific gravity. R 2 244 PAPER-MAKING. CHAPTER XI. CAUSTIC SODA, EECOVEEED SODA, Etc. As we have already pointed out, lime and caustic soda are the only alkalis generally employed by the paper-maker for boiling fibrous raw materials ; the special cases in which carbonate and sulphide of soda is used have been mentioned in their proper place. Lime — the oxide of the metal calcium or CaO — is pre- pared from ' limestones ' of varying quality and is also therefore a somewhat variable product. The unavoidable impurities are water, carbonic acid, silicious matters and residues from the coal used in ' burning ' the limestone. The lime should slake readily and the resulting hydrate (CaCHjO) should form a voluminous impalpable powder. The preparation of milk of lime requires little or no de- scription, as the operation is a simple and tolerably familiar one. Before using, it should be passed through a fine wire sieve, to keep back sand, coal, and similar impurities which the lime invariably contains. Caustic soda is prepared by makers who work on the Leblanc system, in various forms, differing from each other in alkaline strength and colour, and of course in price. The lowest quality is what is known as caustic ' bottoms ' : it consists of that portion remaining at the bottom of the caustic-pot after the clear fused caustic soda has been ladled out, and it contains a considerable quantity of ferric oxide and other insoluble impurities. It is of a dark reddish brown colour, and contains 50-60 p.ct. of alkali (NagO). Its use in paper-making cannot be recommended, except for the pre- paration of the very lowest grade of pulp. The solution CAUSTIC SODA, RECOVEEED SODA, ETC. 245 should be allowed to remain at rest, in order tliat the insolu- ble matter may subside. The next in quality is known as cream caustic, so called from its slightly brown colour. It is usually sold containing 60 p.ct. of alkali, in the form of a tolerably friable mass, hav- ing a crystalline structure. It is a convenient form of alkali for the paper-maker. The next in order is what is called 60 p.ct. white. It consists of a hard white mass, requiring considerable force to break it. Though whiter in appearance than cream caustic, it is in reality less pure, as it contains a considerable quantity of salt. It is made by continuing the evaporation of the caustic liquor to a further point than is the case with cream caustic, the result being that the whole of the water is driven off. A small quantity of nitre is then added to oxidise the sulphides and other compounds which impart the colour to cream caustic. Common salt is then added to reduce its strength to 60 p.ct., in order to satisfy the trade, which con- vention insists upon an article of a given definite strength. On this account it is somewhat inferior to cream caustic, and is, moreover, more expensive. A higher quality is white 70 p.ct. This resembles white 60 p.ct. in appearance, but is much purer, and more expensive. The following analyses of different forms of caustic soda and their relative prices will possibly be of interest : — (Davis.) (Davis.) (Morrison). White (70 per White (60 per Cream (60 per cent. Na20). cent. Na20). cent. Na2U). Sodium hydrate 89-60 75-25 70-00 „ carbonate .. 2-48 2-53 5-00 „ chloride „ sulphate 3-92 17-40 7-00 3-42 4-40 2-00 „ sulphide 0-02 0-03 „ sihcate 0-30 0-30 „ aluminate .. trace trace none none 15*80 Insoluble matter none none 0-20 246 PAPER-MAKING. Their relative prices per ton f.o.b. Liverpool, in December 1886, were:— 70 p.ct. white, 8Z. 5s.; 60 p.ct. white, 71 5s.; 60 p.ct. cream, 71. The following are the results of more complete analyses of trade'sampleSj carried out in the authors' laboratory : Caustic from Spicers (Hunt & Co.). NaOH 72-39 = Na^O 56-11 Na^COs 6-25 = Na^O 3-65 Na^S — NajS^Os 2-90 59-76 NaaSOj 1-13 Na^SO^ 1-00 NaCl 9-44 Na2SiO, 1-08 Na2Al264 -99 Fe^Oj -21 CaSO^ -27 MaO '03 H^O 4-31 100-00 Caustic from N. W. Paper Co. {Muspratts), February 1889. NaOH 72-1.55* Na^COa 8 -2251 NaCl 8-470 Na^SO, 2-378 Na^SOj 3-340 SiO^ 0-200 AI2O3 2-780 Fe^Oj -180 Water, A;c. (diff.) .. 2 272 100-000 • From litmus test, this includes i NagO as NajSOs, also NajO as Silicate, t Diff. between litmus and pli. ph. " Bottoms" from McMurray, March 31, 1890. Litmus 63-8 I Al^O^ 5-0 Ph 63-4 I Na^Sba 0-3 All the above-mentioned forms of caustic soda occur in the form of more or less hard masses, and are contained in thin wrought-iron drums, which are filled with the caustic in a fused state. The soda is removed by breaking the drum CAUSTIC SODA, RECOVERED SODA, ETC. 247 with a chisel and hammer. The lumps of caustic may be put direct into the boiler, though it is better to dissolve them previously in water and allow any insoluble impurities to settle to the bottom of the solution. At the present time there is a considerable industry in concentrated solutions of caustic soda : Solutions containing 30-40 p.ct. NaOH. Thus ' pure liquor 90° Tw.' is quoted to- day (December 31, 1898) at 3Z. 15s. per ton. Also there is a considerable supply of much purer forms of solid caustic soda, such as are made by causticising ' ammonia soda.' The rela- tive prices current we may note as under : ^ £ s. d. 77 p.ct. Caustic soda 7 12 6 74 „ „. 7 7 6 Instead of buying caustic soda direct, it is the custom of some paper-makers to prepare it for themselves from car- bonate of soda, which can now be obtained in various forms, some of great purity. It occurs as caustic ash, containing about 40 p.ct. of alkali as carbonate, and about 10 p.ct. as caustic soda; soda-ash containing about 52 p.ct. total alkali, nearly all as carbonate ; and refined soda-ash containing 57-58 p.ct. of alkali as car- bonate. Some forms of carbonate of soda, such, for example, as those made by the " ammonia process," (Solvay) are of even greater purity. The methoi of converting sodium carbonate into caustic soda will be described subsequently.* Unless the arrangements for washing and draining the lime-mud are very perfect, it is probably cheaper for the paper-maker to buy his caustic direct from the maker. Ferric Oxide Causticising Process. The principle of this process is the expulsion of the carbonic acid from sodium carbonate by ferric oxide at high temperatures; this oxide playing the part of a weak acid, entering into a loose combination with the soda, which is overcome by water in the subsequent lixiviation process, * For a fuller discussion of these questions, see ' The Economy of pure Caustic Soda,' a paper by the Authors in J. Soc. Chem. Ind., 1889. 248 PAPER-MAKING. the ferric oxide being regenerated and a solution of caustic soda obtained. The ferric oxide is therefore continuously available. The main features of the process, disregarding the question of commercial economy, are (1) that it leaves no by-products to be disposed of, (2) tbat it enables the manufacturer to produce directly, without evaporation, a highly concentrated caustic lye. Soda Becovery.—hi former years, the liquors in which rags, esparto, and other paper material had been boiled, was run into the nearest watercourse; but now, owing partly to the fact that is insisted upon by the Eivers Pollution Act of 1876, and partly because it can be made remunerative, all these liquors are preserved, and the soda they contain utilised. 'J'he method adopted is to evaporate to dryness and ignite the residue. The soda during the process of boiling takes up a large amount of non-cellulose fibre constituents and dissolves them as resinous compounds. These on evaporation and ignition become converted into sodium carbonate. From the previous account of the boiling processes it will be seen that in many cases one-half the organic matter of the original passes into solution in the alkaline lye. The calorific value of this organic matter is considerable, and its combustion therefore under regulated conditions may be made to furnish the greater part of the heat required for concentration of the liquor to the firing point. Many raw materials, especially esparto and straw, contain a large amount of silica, a large proportion of which is dissolved by the soda in the form of sodium silicate, in which form also it is found in the re- covered soda. The apparatus for accomplishing the evaporation varies with almost every mill. In some, it is very primitive and crude, consisting perhaps of only a furnace for incinerating the residue, and over it a pan containing the liquor, the latter being heated and evaporated by the heat from the furnace. It is obvious that, with such an arrangement, a large quantity of heat must be wasted. To economise as much as possible CAUSTIC SODA, KECOVERED SODA, ETC. 249 of this waste heat, various plans have been suggested. That of Eoeckner, of Newcastle, appears to be to a great extent efficacious. It consists practically of a series of shallow trays B (Fig. 67) placed in a brick chamber, alternated so as to allow the heated air from the furnace below to play upon the 250 PAPER-MAKING. surface of each in succession, on its way to the chimney, with which the whole system is in connection. Above the chamber containing these trays, is a large tank C, containing a store of the liquor to be evaporated, placed there so as still further to economise the heat, and from which the liquor runs on to the trays. The furnace A is of the or- dinary reverberatory kind ; below it, and connected with it by a kind of damper, is a large chamber J, where the calcined residue from the furnace is put to cool, thus preventing any nuisance from the smell of the burn- ing mass. The chamber is provided with a pipe L, through which the vapours pass into the furnace.' Seve- ral pipes E from the furnace pass through the tank, to assist in warming the liquor. The residue, when cold, is drawn through doors from the chamber below the fur- nace. Eoeckner has devised an apparatus (Fig. 68), con- sisting of a small chamber containing a series of pipes A, through which a stream of cold water constantlj' runs, in connection with the flue from his evaporator, for the purpose of condensing volatile bodies, and thus preventing, to a cer- tain extent, contamination of the surrounding air. CAUSTIC SODA, RECOVERED SODA, ETC. 251 A very economical form of evaporator is that invented by Porion, a French Distiller, and named after him. It is shown in sectional elevation and plan in Figs. 69 and 70. It is largely used on the Continent, and also in England and Scotland. It consists of a large chamber h, the floor of which is slightly inclined from the chimney shaft, and through which the waste heat from the furnace a passes. The liquor to be evaporated is run in at the end nearest the chimney from the tank placed above the chamber c. A Figs. 69 and 70. number of cast-iron fanners i, dip into the liquor and revolve rapidly, usually at the rate of about 300 revolutions per minute, producing and filling the chamber with a very fine spray, thus presenting a very large evaporating surface. Between the furnace and the evaporator are placed the chambers c and /. In c a number of brick walls d are so placed that the flames from the furnace are intercepted and broken up. The object of this is to give time for all the products of combustion to be thoroughly burnt up, which would not be the case without the ' smell-consumer,' as 252 PAPER-MAKING. these chambers are called. This part is an addition to the original evaporator, and was devised by Messrs. Menzies and Davis. The liquor after having been concentrated in the chamber Tc runs into a trough placed alongside the doors h and flows into one or other of the furnace beds 6 where it is still further concentrated, and the residue ignited by the flames from the fires a. The draught can be regulated by the damper g, and also by one placed near the shaft y. The doors e, in the smell-consuming chamber, are for the purpose of cleaning out. The fanners i are worked by a small steam engine, not shown in the drawing. Under properly regu- lated conditions very excellent results can be obtained with this evaporator. The temperature of the gases near the chimney should not be higher than about 85°. By running the fanners at a very high speed the temperature of the gases may be even further reduced, thus showing the completeness of the evaporation. This form of evaporator is open to the objection that the whole of the sulphur in the coal employed for the furnaces, finds its way into the recovered soda. It combines with the alkali to form sulphite of soda, part of which is decomposed in the furnace with formation of sodium sulphate, sulphide, and other sulphur compounds. The same objection, of course, applies, though perhaps in a less degree, to all systems of evaporation in which the flame is in contact with the liquor to be evaporated. The Porion evaporator can be erected at very small cost and costs but little for maintenance. It is capable of pro- ducing ^ ton of recovered soda per ton of coal with liquors of the usual strength. It has proved itself to be perhaps the most economical evaporator existing. Some time ago there was erected in Lancashire an evapo- rator invented and patented by Mr. Alfred Chapman. It is shown in Figs. 71, 72, 73, and 74. The evaporation is effected at a low temperature in three vacuum pans E, and with the unusual result that the concentrated liquor gelatinises after leaving the third vacuum-pan, instead of taking the ordinary CAUSTIC SODA, RECOVERED SODA, ETC. 253 form of the concentrated products of other evaporators. It is said that this apparatus gives an excellent product, with great economy of labour and water, and with no drainage of foul liquor from the buildings. Observations extending over three months have proved that it evaporates 22 lb. of water from the liquor per lb. of coal used under the boiler. It is however very costly to erect. The waste liquor is discharged into the tank A, whence it Figs. 71 and 72. is pumped by the donkey-engine B, through the feed-heater C, into the boiler D, which receives heat from the incinerat- ing furnace H, and, in case of need, from an auxiliary furnace shown on the plan, under the feed-liquor-heater. The steam produced in D is taken to the first vacuum-pan at E, and having heated its contents, the products of evaporation pass over into the tubes of the second pan ; this, in its turn, gives up its products of evaporation to the third, whence they go to the condenser of the vacuum-engine F. Thus the heat from 254 PAPER-MAKING. the furnace H is used for incinerating the concentrated liquor on its bed, for heating the feed-liquor in the feed-heater pipes, and for making steam out of the liquor itself in the boiler ; Figs. 73 and 74. this steam finally drives the donkey-pump and vacuum- engine, and causes the evaporation in the three vacuum-pans E. One advantage of this evaporator is the fact that the liquor is evaporated out of contact with the furnace gases. CAUSTIC SODA, EECOVEICED SODA, ETC. 255 " Evaf oration hy Multiple Effect." In recent years there Las been a large development of methods of evaporation based upon the ideal of an exhaustive utilisation of the heat expended upon the liquors to be con- centrated. The principle of these methods is that of ' multiple effects,' which may be briefly explained as follows : — A liquid is converted into vapour under ordinary conditions of boiling, by overcoming the pressure of the atmosphere upon its surface. The quantity of heat required to vaporise, as also the temper- ature of the ebullition, will be less as the pressure to be over- come is less. Further, the vapour continuously driven oif carries with it a quantity of heat, which is its heat of condition or latent heat. This heat it imparts to any colder body (e. g. a further quantity of the same liquid) with which it comes in contact, direct or indirect ; if the quantity of the latter be relatively small, it will raise its temperature approximately to that of the ebullition of the first liquid. If now the pres- sure (atmospheric) on the surface of the latter be slightly reduced, by any means, it will boil. The vapour from this can be made to boil a third quantity of the liquid, under a further diminished pressure. The successive effects in economic evaporation consist, therefore, in utilising the latent heat of a vapour given off" from a liquid under a certain pressure (e.g. that of the atmosphere) to vaporise a further quantity of the liquid under a pressure maintained by mechanical means below that of the first. In certain methods the vapour does its work in the successive eifects by passage through systems of tubes, the liquid to be heated being in contact externally ; in the Yaryan system, on the other hand, the arrangement is reversed. The liquid to be evaporated traverses the system of tubes which are heated externally by the vapours. At the end of each effect, the liquid is caused to impinge, in a special chamber, upon a disc: in this way a complete separation of liquid and vapour is eff"ected, each then passing on to the next effect, the former through the tube-system. 256 PAPEK-MAKING. the latter to the chamber inclosing these. The flow of liquid is maintained by a force-pump, and the diminished pressure by a vacuum-pump suitably disposed. This system differs from that described on p. 252, in that the evaporation is continuous, the dilute liquors entering the apparatus and the highly concentrated liquors leaving it in an unbroken stream. The rate of flow is such that the evaporation of the caustic liquors from wood boiling from 8-10° to 80° Twaddell, in a quadruple effect, requires only a few minutes. At the latter concentration it is ready for the incineration process, which by means of a rotary furnace, such as that of Mr. J. W. Hammond, of the firm of S. D. Warren and Co., is also effected continuously. It is found, moreover, that the excess of heat available from this process is sufficient for the evaporation.* Whatever be the method of evaporating or concentrating the liquor, the final treatment in the furnaces is much the same in every case. The furnaces shown in Figs. 69 and 70 may be taken to represent the ordinary form. The concen- trated liquor is run on to either of the beds 6, where the last portions of water are driven off by the heat from the fire- places a, and the residual mass is ignited until all the organic matter contained in the liquors is carbonised and the soda is converted into carbonate of soda. This takes place in about 4 hours, according to the degree of concentration of the liquor as it is run into the furnaces. The running in of liquor should be done with great care, as explosions sometimes occur through the sudden liberation of steam on the liquor coming in contact with the hot beds. The charge should be allowed to remain in the furnace until it is thoroughly carbonised and all volatile matters have been driven ofi", otherwise a nuisance may be caused when the still burning mass is exposed to the air. Roeckner's evaporator is provided ♦ A full account of these methods is given by Griffin and Little, loc. cit. pp. 165-170. The student may also consult a treatise by J. Foster on " Evaporation by the Multiple System," London : Simpkin, Marshall and Co., 1890. CAUSTIC SODA, RECOVERED SODA, ETC. 257 ■vvitli a special chamber into wliicli the charge is drawn (J, Fig. 67). The composition of the recovered soda varies with the nature of the liquors from which it has been obtained, and, as has been already pointed out, with the form of evaporator employed. It consists essentially of carbonate of soda, to- gether with a certain amount of silicate of soda, if derived from liquors in which straw or esparto have been boiled, chloride of sodium, sulphate of soda, sulphite of soda, sulphide of sodium, and other sulphur compounds, the rest being made up of carbon and insoluble impurities. The amount of soda varies from 35 to 45 p.ct. (Na20). The following analysis will give some idea of the composition of Eecovered Soda : — NaiO ♦Sodium carbonate 72-33 = 42-306 Sodium hydrate 1-93 = 1-497 Sodium chloride 8-30 Sodium sulphate 3-95 43-80a. Sodium sulphite '63 Silica 7-09 Carbon 4-70 Oxide of iron and alumina -50 Other constituents (by difference) . . . . -57 100 00 The whole of the soda present as sulphur compounds is not lost, as a large proportion of it is present as sodium sulphite, most of which is converted into caustic soda by the causticis- ing process. A certain amount of soda is carried forward, partly me- chanically and partly volatilised, to the flue leading to the chimney. This accumulates, and may be from time to time removed in the form of fine dust. It contains, besides carbonate of soda, much sulphate and chloride. In two different samples examined by the authors, the amounts of soda (NaaO) present were 25-0 and 27*1 p.ct. * A certain quantity of potash derived from the mineral constituents of esparto and straw is always present in recovered soda. s ^58 PAPER-MAKING. Causticising. — The next process consists in converting the sodium carbonate in the recovered soda into caustic soda. This operation is known as ' causticising,' and consists in heating a solution of the soda with lime. The decom- position which takes place is shown in the following equation : — Sodium ,. w«tPT- Caustic Calcium Carbonate. water. Carbonate. Na^COs + CaO + H^O = 2NaOH + CaCOa It must be remembered that this is a reversible reaction and that caustic soda will attack calcium carbonate under certain conditions of concentration and temperature, to re- form sodium carbonate and lime. Hence the limits of decomposition, which have been determined by G. Lunge as follows : — r.ct. NasCOa Sp. gr. before P.ct. Soda in Liquor. Causticising. CaustlciseU. 2 1 022 99-4 5 1-052 99-0 10 1-107 97-2 j12 1-127 •,. 96-8 14 1-150 94-5 16 — 93-7 20 — 90-7 'The recovered soda should be dissolved in separate vessels. Perhaps the best form of apparatus is a series of lixiviating " tanks such as are used for dissolving the alkali in black ash. By this means a nearly perfect exhaustion of the mass can be effected with a minimum of labour. Special tanks are sometimes made for the purpose, provided with mechanical stirrers. It is essential in dissolving the recovered soda that a high temperature should be employed, as otherwise a portion of the soda piesent as silicate of soda will be lost, as it is only with difficulty soluble, and requires rather prolonged heating with -water. Whatever the f(>rm of apparatus employed it should be so arranged that, after running off the strong liquor, the insoluble residue may be further treated with water. In the CAUSTIC SODA, RECOVERED SODA, ETC. 259 case of the vats mentioned above, this process is made con- tinuous, pure water being run in at one end, and strong liquor flowing from the other. If other forms are used, the liquor after settling, may be run off by means of a pipe passing through the bottom or side of the vessel, and near the bottom and consisting of two parts, one long and one short. The short part is stationary, and is connected to the longer part by means of a movable knee joint, allowing it to be deflected. The liquor having settled sufficiently, the movable limb is lowered beneath the surface of the liquor which is then allowed to flow through. As the surface of the liquor falls, the pipe is gradually lowered. In this way the clear solution can be run off without disturbing the residue at the bottom. The open end of the pipe is usually covered with coarse wii'e gauze, to keep back insoluble impurities. With properly calcined recovered soda, the solution should be bright and almost colourless. If at all brown in colour, and if it has an empyreumatic odoui-, it indicates imperfect calcination. The residue in the dissolving tanks consists chiefly of carbonaceous matter, together with some soda, insoluble matter, &c. The liquor is now ready to be causticised. This should be done in a separate vessel, although it is the practice in many mills to perform this operation in the same vessel in which the solut'on of the soda has been conducted. A good form of causticiser can be made from an old egg-shaped boiler, by cutting it in two along its length. It should be provided with two or more vertical steam pipes, connected at the bottom of the boiler with a horizontal pipe perforated with numerous holes. The vertical steam pipes should be furnished with injectors, whereby air is drawn in, and forced with the steam through the holes in the horizontal pipe. The stream of air serves the double purpose of thoroughly agitating the liquor and of oxidising any sodium sulphide in the recovered soda. The liquor before causticising should be reduced in strength to about 20-25 degrees Twaddle, which may be done w^ith the washings of the residue from the recovered soda, or from the s 2 260 PAPEE-MAKING. washings obtained subsequently from the lirae-mud. This strength should never be exceeded, otherwise imperfect conversion into caustic soda is the result. This is due to the fact that concentrated solutions of caustic soda react upon calcium carbonate, forming sodium carbonate, and calcium hj'drate, the reaction being the reverse of that indicated in the above equation. If the liquors are very strong in carbonate of soda, and comparatively free from sulphate, they should not be causticised at much over 20"^ Twaddle, if they contain much sulphate, and therefore less carbonate, the higher strength can with safety be adopted. The causticising vessel should be provided with a stout iron cage or basket, into which the lime can be put. This should be securely fastened to the vessel, and should dip into the liquid. The liquor having been properly diluted, is now heated by means of the steam pipes, and the lime put into its cage. It should be put in in lumps. As the liquor reaches the boiling point, the reaction will proceed rapidly, and the lime will gradually disajipear ; fresh lumps should be added if necessary. If the liquor is sufficiently heated the caus- ticising will be complete iu from two to three hours. The liquor should be tested from time to time ; this is usually done by a workman. He withdraws a sample of the liquor^ and after allowing the calcium carbonate to subside, poui-s off a portion of the clear liquid into a glass vessel. He then adds an excess of either sulphuric or hydrochloric acid. If any effervescence takes place, due to the evolution of carbonic acid gas, he knows that the operation of caus- ticising is incomplete ; the heating must therefore be con- tinued. It is difficult, without an undue expenditure of time and steam, to convert the whole of the soda into- caustic : it should however be so perfect, that on testing only a very slight effervescence occurs. It is quite easy to convert as much as 95 p.ct. of the soda, or even more (see p. 258). The actual amount converted can only be ascer- tained by a careful analysis of the liquor. CAUSTIC SODA, EECOVEKED SODA, ETC. 261 The amount of lime used is generally somewhat in excess of the theoretical quantity; 106 parts of sodium cai'bonate (Na2C03) require 56 parts of lime (CaO) : it is necessary, how- ever, to add about GO parts. A very good plan is to conduct two or even three causticisings in the same vessel without cleaning out or removing the calcium carbonate, using in the first operation a largo excess of lime. The causticising being completed, the calcium carbonate and excess of lime are allowed to settle down, and the clear liquor run off by an arraugment such as that already described in the dis- solving process. Fresh solution is then run in and the whiole mass heated for some time, until the excess of lime is converted into carbonate. Fresh lime is then added if necessary until the conversion of the carbonate of soda is complete. The liquor is then allowed to settle, and is run off as before : this operation may again be repeated. The residual calcium carbonate, or ' lime-mud ' as it is called in alkali works, is then washed once or twice by running in water, boiling up, allowing to settle, and running off tlie clear liquor. If these liquors are too weak for use in boiling fibres, they may be used for diluting fresh recovered soda liquor before caustici^iing, or for dissolving the soda. Some ariangement should be provided for removing as much as possible of the liquor from the lime-mud before throwing it away or otherwise disposing of it. This is best done by throwing it on a filter made of layers of stones, ashes and sand, and covered at the top with perforated iron plates. The filter is connected with a vacuum pump. In this way very perfect di ainage is accomplished, and the mud forms a hard mass on the surface of the filter, from which it can be easily removed with spades. In this form it con- tains only 50-60 p.ct. of water. If properly washed it should contain in this state only about 2 p.ct. of alkali (Na^O). By careful manipulation, even this amount can be reduced. The importance of thoroughly washing the mud can hardly be too much insisted upon. Where proper means are not 262 PAPER-MAKING. employed for draining, the washing should be made more perfect. The lime-mud consisls chiefly of carbonate of lime, together with silicate, free lime, &c. The following analysis is of a mud obtained b}' causticising recovered soda derived from the liquors in which esparto and straw had been boiled : — Calcium carbonate 40 '02 Calcium hydrate 5 • 13 Wlica 4-01 Sodium hydrate .. .. 2 "13 Oxide of iron and alumina 0-30 Water 48-10 Other constituents 0'31 100-00 As already pointed out the liquors contain a certain amount of soda, as sodium sulphide and other sulphur com- pounds. The presence of the former, if in large quantities, is objectionable, as it is liable to discolour fibres boiled in liquors containing it. It is therefore best to remove it. This can be conveniently done by blowing air into the liquors during the process of causticising : this has the eft'ect of oxi- dising it to sulphate of soda, in which form it is harmless. The air can of course be blown into the liquor by means of a pump ; the most economical way is to connect with the steam pipe an injector constructed on the j principle of the injectors used for feeding boilers and for other purposes. By this means a strong current of air is drawn in, and being forced with the stream to the bottom of the liquor, passes through it in a number of fine streams. If the amount of sulphide present be very high it may be necessary to prolong the oxidising operation beyond the time necessary for complete causticising. In many paper-mills the causticising is conducted in cir- cular vessels furnished with mechanical agitators. These are more expensive than the simple form described above, and they possess no special advantages. The use of causticisers in which neither mechanical agitation nor agitation by means of air is provided for is exceedingly wasteful of labour, time, steam, and soda. The lime-mud settles at the bottom as a hard mas«, very difficult to manipulate. 263 CHAPTEE XII. PAPEE TESTING. There are two points of view from which a paper may be tested : first, of physical or mechanical properties ; secondly, of material composition. We shall consider the subject ac- cording to this division. (1) Quantitative measurements of such properties as re- sistances to breaking and tearing strains are seldom made by English paper-makers. In Germany, on the other hand, the matter has been very thoroughly investigated in connection with the work of the Konigl. Techn. Versuchsanstalt, Berlin, and through the agency and influence of Prof. Sell and C. Hofmann, a department has been organised exclusively for the -work of paper testing. The results of the tests are; becoming widely recognised by practical men and the trade in that country, as affording a true index of the quality of a paper. It is therefore of importance to give an outline of the method employed. The determination of the stiaih or weight which a paper is capable of supporting is a very obvious measure of the strength of the paper. Observations of the limiting strain or breaking weight are sometimes made by paper-makers^ but the apparatus and method employed are usually crude. The simplest means consist in clamping the paper— a strip of standard length and breadth, arbitrarily chosen — at one- end, the clamp being firmly held in a fixed support, and to- the other attaching by means of a similar clamp, an ordinary scale pan, the whole arrangement hanging vertically. Into- the pan, weights are added in due succession until fracture of the strip is determined. It is scarcely necessary to point 264 PAPER-MAKING. out that the errors of experiment with such a method are very great : indeed it has been found that even with the refined apparatus about to be described the errors are not inconsiderable. However, by exhaustive investigation, ac- cording to the well-known ' law of errors,' these have been quantified, and a careful operator can therefore obtain results which are trustworthy. The apparatus in question is the Hartig-Keusch machine.* It is shown in sectional elevation and plan in Figs. 75 and 76. The principle will be readily grasped by inspection of the diagrammatic representation of its essential parts — Fig. 77. The strip of paper is held horizontally by the clamps a and h, a being held by the fixed support A, h by the movable carriage B. B is connected by means of a swivel with the spiral spring F, and this again is similarly connected with the screw, which is made to rotate by the wheel D. By turning D, therefore, the spiral may be extended, and a corresponding strain communicated through B and b to the paper. The paper undergoes a certain elongation under the strain, and the carriage B moves from right to left in conse- quence. The rotation of the screw is continued, and the extension of the spiral pi'oceeds until the paper is fractured. At this point it is required to determine, (1) the elongation of the spiral which is the measure of the breaking strain, and (2) the distance through which the carriage has moved, i. e. the elongation of the paper. Both these effects are com- municated to the pencil G, the latter directly, since the pencil-holder is in rigid connection with B, the former th rough the rod I, from which by a special arrangement, the hoiizontal is converted into a vertical motion of the pencil. This, therefore, traces a curve, of which the ordi- nates represents the strains, and the abscissae the elongations of the paper produced by the strain. The scale shown in Fig. 76 indicators the exact position of the clamp A. * A complete description of this machine is sivcn in ' Civil Engineer,' 1879. 266 PAPEK-MAKING. The values for the spiral spring— i. e. extension for a given load — having been determined by previous observa- tions in a special apparafu", the curve obtained is at once a measure and a permanent record of these cardinal factors, breaking strain and elasticity. As with all other such in- struments, the recording apparatus introduces certain errors, which, however, by careful investigation and modification in accordance with the results, have been reduced to a minimum. Nevertheless, the director, Dr. Martens, has recently adopted a simpler instrument, altogether similar in principle, but based upon a direct reading of the two movements, in which of course these errors do not appear. For the student, however, the recording instrument is the more instructive, and we have given it preference for de- scription here, more especially as no difference in essential parts is involved. Those who wish to pursue the matter into the most in- teresting details of the investigations made upon the subject, are referred to the papers published by the Institute for 1885. A useful abstract of these publications so far as re- lates to paper, has been made by the Society of Arts Com- mittee on ' The Detei ioration of Paper,' and published in the report upon that subject.* In testing the strength of papers by this or similar machines, it is important to observe the hygrometric state of the atmosphere at the time the trials are made, as this has been found to exert a considerable influence on the results, a paper being weaker the moister the atmosphere. The results of the tests are expressed in the following terms : — The elongation is given directly in percentage of the original length. This is uniformly taken at 180 mm., a length arrived at after laborious investigation, as mini- mising the errors of experiment ; in other words, as giving mean value with tlie minimum of variation. For the break- ing strain an ingenious expression has been arrived at, * Journ. Soc. Arts, 1898. ins therefoTft a, nrnnmt.inTia+AKr o-T-oci+ot' fr^^^^ -j. The paper used for this publication is a normal printing paper, B quality, prepared according to the specification of the Society of Arts Committee, 1898, on Paper Standards. appear iK iiit! measure oi tne resistance. A sufficient uni- AT^,r^>.+-k^iiacc +.V1P1 rlirector. Dr. Martens, has macliines, It is importam lu uubcivc txxo PAPEK-TESTING. 267 viz. the length of the paper which suspended vertically, with one end hanging freely, the other fixed, would deter- mine fracture at the fixed end. As the breaking strain would vary with the thickness, the numbos obtained in units of force or weight for strips of constant breadth, would need correction in order to admit of strict comparison with one another. By substituting an expression in terms of the paper itself — since a paper of greater thickness and requir- ing therefore a proportionately greater force to fracture it, weighs more per unit of area, and in the same proportion — all the numbers for breaking strains are strictly comparative one with the other. In the same way also the question of width may be disregarded. A further mechanical test, forming a part of the scheme of investigation, is the resistance of the paper to rubbing. This test is an altogether empirical one, as the following brief description will show : — A piece of the paper, about 6 inches square, crumpled by successive folding in two directions at right angles, is grasped by the thumb and fore- finger of each hand, at a distance of 3 to 4 inches apart. It is then rubbed upon itself across the thumbs a given number of times (seven is the number chosen) and held up to the light. If no holes are visible, the rubbing is repeated. The number of times necessary to repeat the rubbing until holes appear is the measure of the resistance. A sufficient uni- formity in the results of this test has been observed to make it the basis of a classification of papers, in regard to their resistance to such disintegration ; they are divided into the following seven groups, beginning with the lowest : — 0. Extremely weak. 4. Moderately strong. 1. Very weak. 5. Strong. 2. Weak, 6. Very strong. 3. Medium. 7. Extremely strong. The classification of papers on the results of these tests cannot be more lucidly given than in the following scheme, under which the results are officially recorded:-- 268 TAPER-MAKING. Class a. Mean breaking length (metres)' 6000 5000 4000 3000 2000 1000 1 3 4 5 6 not less than b. Mean elongation (p.ct.) at frac- ture not less than c. Eesistance to rubbing j 4-5 6 4 6 3 5 2-5 4 2 8 1-5 1 This classification is based on the results of some hun- dreds of observations. It is interesting to note the differ- ences observed in the numbers for a and h according to the direction in which the paper is cut for the test, i. e. in the direction in which it was run on the paper machine, or at right angles, i.e. across the web. The mean ratio for the breaking lengths (strains) may be taken as 1 : 1 * 6, i. e. the paper is about 40 p.ct.* weaker across the web ; the elonga- tion under strain on the other hand is about double. It may be pointed out that resistance to strains applied as in the testing machine, is not a full measure of the strength of a paper as shown in actual use. Thus it does not give full value to the influence of length of fibre, which is an important factor of resistance to shearing or tearing strains. This quality is measured roughly by the paper expert, by tearing the sheet in the two directions, noting the kind of tear and the tearing force which he employs. It is also of interest to note the influence of the glazing process (p. 235) upon the quality of the paper as determined by these tests. First, we must notice the efi"ect of the treatment upon the substance of the paper itself. The mean reduction of thickness is 23 p.ct. On the other hand, the reduction of weight, calculated per unit of surface (square metre), is 6*7 p.ct., whence we may infer an increase of surface, flatten- ing out, in the process. These quantities, but more particu- larly the latter, will doubtless vary with the various methods of glazing and with the materials of which the paper is composed. * In the statement of results the mean of the numbers obtained in the two directions is given. PAPER-TESTING. 269 The breaking length (strain) shows a mean increase of ahont 8 p.ct, ; the elongation under strain, on the other hand, a di^iainution of 6 p.ct. For an interesting discussion of the question of the rela- tive strengths of machine and hand-made paper see ' Paper,' by Eichard Parkinson. The thickness of a paper may be determined by means of an ordinary micrometer, such as is shown in Pig. 78. The paper is placed in the jaws of the instrument, and the screw advanced until it touches the paper. The thickness is then read off on the scale. Other forms of apparatus are sold for the same purpose. In making a determination of the thick- ness of a paper it is necessary to take the mean of a series of observations at different points of the sheet, as the thickness may vary somewhat. The product obtained by multiplying area by thickness, determines the volume or cubic contents of the paper sub- stance : and the weight of the unit of volume being the specific gravity, this constant is readily deduced. ' Bulk ' as currently expressed, is the inverse of this, i. e. volume per unit of weight. In every mill it is important to keep records of the numbers expressing this quality. Special papers. — Such as are used for some special pro- perty, are tested with particular reference to that property. Thus blotting papers are tested by their relative capacities for absorbing and transmitting fluids. Strips of the paper are cut of constant and equal length, and suspended verti- cally. The lower extremities are then inserted in a vessel of water, and the time is noted during which the water rises to a constant height, also the extreme height to which the water rises by capillary transmission is also noted. D Fig. 78. 270 PAPER-MAKING. Determination of Composition of Papers. (II.) The analysis of a paper naturally divides itself into two parts :— (rt) The determination of the nature of the fibrous material of which it is composed ; and (6) the iden- tification of such adventitious substances as size and filling material. (a) This again is divided into two sets of observations — microscopical and chemical. A fragment of the paper is soaked for some time or boiled for a short time in dilute alkaline solution, and is then care- fully teased out with a pair of needles, and the fragments laid on a glass slip with a drop of diluted glycerine. A cover-glass is then laid on and lightly pressed down so as to spread the fibres in a thin layer. The microscopical features of the different fibres have been already described, and it is only necessary now to summarise the chief characteristics of the more important materials. Cotton. — Flat riband-like fibres, frequently twisted upon themselves. The ends generally appear laminated. The fibres are frequently covered with nximerous fine markings (see Frontispiece). Linen. — Cylindrical fibres, similar to the typical bast fibre (see Fig. 6). The ends are frequently drawn out into numerous fibrillae (see Frontispiece). Esparto. — Esparto pulp consists of a complex of bast fibres and epidermal cells. These serrated cells are, as has been already pointed out, characteristic of esparto, straw, and similar fibres. Certain differences exist between those of ■esparto and stiaw, and even between the different species of straw, which enable the microscopist to identity their source. The most characteristic feature of esparto pulp is the presence of a number of the fine hairs which line the inner surface of the leaf (e, Fig. 10), some of which invariably survive the •boiling and washing processes, although the greater portion passes away through the wire-cloth of the washing engines. PAPER-TESTING. 271 The presence of these hairs may be taken as conclusive evidence of the presence of esparto. Straw. — b^traw pulp very closely resembles esparto-pulp in its microscopical features. The hairs above alluded to are, however, absent. On the other hand, a number of flat oval cells are always present in paper made from straw (h, Fig. 13). It should be borne in mind, however, that bamboo and similar pulps also contain these cells. Wood (Chemical). — Flat riband-like fibres, showing un- broken ends (see photographs, Frontispiece), l"he presence of the pitted vessels (Frontispiece, and a, Fig. 16) is eminently characteristic of pulp prepared from pine-wood. The fibres of other woods are not sufficiently characteristic. They mucli resemble those of pine-wood, with this difference, that the pitted vessels are absent. Wood {Mechanical). — Mechanical wood-pulp may be recog- nised by the peculiar configuration of the torn ends of the fibres, and from the fact that the fibres are rarely separated, but are generally more or less agglomerated (see Frontis- piece). Pulp from pine of course shows the pitted vessels already referred to. They are usually more distinct than in chemical wood-pulp. Occasionally fragments are to be met with connected together with portions of the medullary rays. The identification of the ligne-cellulose is rendered very certain loy previously staining the specimen with a basic- aniline dye, or with the phloroglucol reagent (p. 85), or with a solution of a salt of aniline or dimethyl p. phenylene diamine. The microscopical examination of a paper is a matter of vevy great difficulty, and one requiring much practice. The student is recommended to study closely for himself the microscopical features of pulps obtained from authentic specimens. A fair measurement of the relative proportion of the various fibres present in a paper can be obtained from a careful microscopical examination. 272 PAPER-MAKING. Yetillart * was the first to maintain that a quantitative determination within a fair limit of accuracy is possible. The degree of accuracy attainable depends first upon the kind of mixture under examination, and secondly, upon the experience of the observer. In examining a paper under the microscope, it should be observed whether the fibres appear as fragments, or whether they consist of whole (bast) cells in which the tapered ends appear. Cotton and linen, owing to the great length of their ulti-> mate fibres, yield, when beaten, fragments showing where the fracture has taken place. From the appearance of this fracture it is possible to ascertain whether or not the beating operation has been properly conducted. If the beater-knives have been too sharp, or have been let down to the betl-plate too quickly, the fractures will appear as clean cuts, whereas when the operation has been properly conducted the fracture will appear ragged and drawn out. The bearing of this on the strength of the finished paper is considerable. Esparto, straw, and wood, whose ultimate fibres do not ex- ceed 1-2 mm., should, in the majority of cases, appear as whole (bast) fibres with two tapered ends ; the beating, when pro- perly conducted, being confined merely to the separation into these ultimate fibres. (6) For the chemical identification of the fibres in writing and printing papers, the most useful reactions are those with aniline sulphate solution. The fibres (celluloses) of the rag and wood groups give no reaction, but straw and esparto celluloses and mechanical wood-pulp can be identified by its means. The authors have found that when a paper contain- ing straw or esparto is treated for some time with a boiling 1 p.ct. solution of aniline sulphate, a red colour is produced. Esparto gives the reaction with greater intensity than straw. In this way the presence of a very small quantity of these pulps can be detected with certainty. Since esparto and * Etudes sur lea Fibres Vegetales Textiles, Paris, 1876 (Firmin-Didot). PAPER-TESTING. 273 titraw cellulose give large yields of furfural in boiling with hydrochloric acid of 1 -06 sp. gr., a determination of furfural gives a close approximation to the proportion of these cellu- loses in a paper containing only linen and cotton celluloses in addition. The percentage is obtained by multiplying the furfural number by 8.* Mechanical wood-pulp, when treated with a solution of aniline sulphate, develops, even in the cold, a deep yellow colour. If a paper containing mechanical wood-pulp so treated be examined under the microscope, the fragments of wood will be found to be deeply stained, whereas the other fibres remain colourless or nearly so. It must be borne in mind that cellulose obtained from lignified fibres, if the boil- ing and bleaching processes have not been carried sufficiently far, will give with aniline sulphate a more or less intense yellow coloration. Various other reagents have been sug- gested for the identification of mechanical wood-pulp, all based upon the production of a colour with lignose. Of these the most important is the solution of the coal tar base, dimethyl p. phenylene diamine, the reactions of which have been previously described. The reaction of lignose with chlorine and sodium sulphite t-olution already referred to (p. 54) may also be made a^' ail- able for the detection of mechanical wood-pulp in a paper. Imperfectly boiled or bleached pulps sometimes give this reaction faintly. Quantitative Estimation of Mechanical Wood-pulp. — The determination of the amount of mechanical wood-pulp present in a paper is sometimes a matter of some importance, and it is also a matter of some difficulty. A general idea of the amount present can be obtained by observing the depth of the yellow colour produced with aniline sulphate or the intensity of the magenta reaction with chlorine and sodium sulphite. Wurster's colour reactions (supra) are more pro- nounced and the quantitative findings based upon them are * For details of furfural estimation see ' Cellulose.' T 274 PAPEK-MAKING. generally accepted as affording a sufficiently close approxima- tion. Test papers containing the diamine base together with a scale of colours corresponding to varying percentages of the lignocellulose, are obtained from Schuchardt of Gorlitz (Germany). It is also possible to calculate approximately the percentage from the percentage of cellulose contained in the specimen. Mechanical wood-pulp (pine) may be taken to contain 60 p.ct. of cellulose. If, therefore, a paper ascer- tained to contain a pure cellulose in addition to this con- stituent, yield 75 p.ct. of cellulose on the ordinary test, it may be assumed that about 62*5 p.ct. of mechanical wood- pulp is present. The authors have proposed a method of estimating the amount of mechanical wood-pulp present in a paper, based upon the absorption of iodine in definite proportions by wood in a finely divided state, under strictly regulated conditions. The paper is carefully reduced to a fine pulp and is then left in contact with a standard solution of iodine in potas- sium iodide. At the end of twenty-four hours the amount of free iodine is determined by titration with sodium thio- sulphate and by deducting this from the amount originally taken, the amount absorbed is ascertained. As tlds amount, under strictly comparative conditions, always corresponds to a definite amount of mechanical wood-pulp the amount present can be readily calculated. 2. Loading, Sizing Materials, &c. — The determination of the amount of loading material in a paper has been already described (p. 201). The identification of the material can only be arrived at by a careful chemical analysis. The principal loading materials are china-clay and pearl-harden- ing (calcium sulphate), and in coated or ' art ' papers, barium sulphate. The ash from a paper containing cliina-clay is insoluble in boiling dilute hydrochloric acid. ; that from paper containing calcium sulphate is soluble : the solution on cooling deposits long needle-shaped crystals of CaSO^ 2H2O (20*93 p.ct.) and gives with barium chloride a copious precipitate of BaSO^ (barium sulphate) insoluble in acids, PAPEE-TESTING. 275 and with ammonia and oxalate of ammonia a precipitate of calcium oxalate. To examine for barium sulphate the ash is fused with a mixture of sodium and potassium carbonate. The mass is boiled out Avith water. The presence of starch in a paper can be readily ascer- tained by its behaviour with a solution of iodine. If starch bo present the well-known blue colour of the compound of iodine and starch will be produced. The determination of the amount of starch present is a matter of some difficulty, the details of which are somewhat beyond the scope of the present work. It is based upon the conversion of the starch into sugar, and the estimation of the latter with Fehling's solution. The authors have investigated the removal of starch by the action of saliva. The amylolytic action of this secretion is favoured by a slightly alkaline reaction, and a temperature of 30-40° C. The paper to be tieated is treated with boiling water and after cooling to 40° a little sodium bicarbonate is added to slightly akaline reaction, then a small quantity of saliva. With this it is digested 4-5 hours at 30-40° and then washed. The starch is estimated by the loss of weight sustained. The nature of the material with which a paper is sized may be ascertained in the following way : — The sample, cut up into small fragments is warmed for a few minutes with absolute alcohol containing a few drops of Hcetio acid. The alcohol is allowed to cool, and is then l)oured into four or five times its bulk of distilled water. If any precipitate or cloudiness is produced, it indicates that the paper has been sized with rosin. The alcohol dissolves the rosin, which, being insoluble in water, is thrown down on dilution. The alcohol used may be obtained from com- mercial spirit by digesting it with lumps of quick lime and distilling from a water bath. The paper after treatment with alcohol should now be boiled for some minutes with water: the solution allowed to cool, and then filtered. To the filtrate a few drops of a T 2 276 PAPER-MAKING. solution of tannic acid are added, when, if the paper has been sized with gelatine, a white curdy precipitate will be formed. The estimation of the amount of sizing material in a paper is a very complicated process and one which demands con- siderable chemical experience for its proper conduct. The amount of gelatine present is best ascertained by determining the amount of nitrogen present by combustion with soda-lime, and from this, calculating the amount of gelatine. Pure gelatine contains 18-16 p.ct. of nitrogen (Muntz). The comparison of one paper with another with a view to ascertain the relative degree of sizing, is usually performed in a more or less rough and ready way by moisten- ing the samples with the tongue for a certain time, and noticing the degree of tiansparency produced, which is of course inversely to the degree of " hardness." A more accurate method consists in placing a drop of a mixture of alcohol and water containing some colouring matter in solution and ■determining the time necessary for the colour to make its appearance on the other side. In this way more trustworthy comparisons can be made. Colouring Matters. — The chemical characteristics of the chief colouring matters have been already described. 'Mineral' pigments will be obtained in the ash, in some cases without •chemical change (smalts, iron oxides, &c.), in other cases in the form of the characteristic constituent of the original pigments. Thus the prussian blues leave a residue of iron oxide; the chromates may be more or less 'reduced' to chromic oxide. The coal tar colours and soluble dyes are recognised by characteristic reactions, for which the special test books dealing with the subject must be consulted. 277 CHAPTER XIII. GENEKAL CHEMICAL ANALYSIS FOE PAPEE- MAKEES. A KNOWLEDGE of the nietliods usually employed for the qualitative and quantitative analysis of the various che- mical substances met with in paper manufacture is of con- siderable importance. The scope of this work will not allow us to enter fully into the necessary details of manipu- lation ; for these the reader is referred to standard works on analysis : we will merely indicate the methods. Lime. — In its commercial forms contains in addition to the active calcium oxide, water and carbonic acid, in combina- tion with lime, and substances insoluble in acids, derived partly from the original limestone, partly from the coal used in ' burning ' it. Magnesian limestones yield a lime containing of course magnesia. These facts indicate the methods to be pursued by the analyst in valuing a sample of lime. Water and carbonic acid may be estimated jointly by the loss of weight on igniting to a white heat in a platinum vessel till constant. Earthy impurities such as sand by treating the lime with a dilute acid in slight excess, filtering off and weighing the insoluble matter. For an exact determination of silica, alumina, iron oxide and magnesia, the usual systematic separation must be followed. In actual practice it is usual to estimate the lime by a direct volumetric method ; a standard solution of oxalic acid being added to the lime previously slaked and brought to 278 PAPEE-MAKING. the condiliun of ' milk of lime.' As an indicator, phenol phthalein is used. The reaction is carried out in a stoppered bottle, which is vigorously sbaken from time to time. The oxalic acid solution is added until the pink colour of the phthalein is discharged. Caustic Soda, Soda Ash, Recovered Soda, dtc. — These sub- stances are always valued by the amount of real alkali (NagO) that they contain, which is determined by titration with a standard acid. In testing recovered soda, it is necessary to boil the finely powdered ash for some time with water; unless this be done, the amount of alkali will probably be under-estimated, as part of the soda is present as a difficultly soluble silicate. It is sometimes desirable to determine the amount both of sodium hydrate and carbonate in a sample of alkali. For this purpose an excess of a solution of barium chloride is added to a solution containing a known quantity of the substance under examination, and the mixture made up to a definite volume. It is then fi.ltered, and an aliquot portion •of the filtrate titrated with standard allcali. The filtrate con- tains only the alkali that was originally present as hydrate, the carbonate of soda having formed with the barium chloride, insoluble b.rium carbonate, and sodium chloride. Bleaching P< toder. — The value of bleaching powder de- pends upon the amount of 'available chlorine' that it contains. Thi s is determined by means of a standard solution of arsc- nious acid. The standard solution is prepared by dissolving pure arsenious oxide (As^Oa) in todium carbonate solution. To test the powder, a quantity, say about 5 grms., is taken, ■and is carefully ground up in a mortar with a small quantity of water, more water is then added, and the coarser particles allowed to settle for a few seconds. The milky liquid is then run off into a graduated flask ; the residue is again ground up with water, and the operations repeated until the whole of the powder is transferred to the flask. The flask is now filled up to the proper mark, carefully shaken, and an aliquot portion withdrawn as quickly as possible. GENEEAL CHEMICAL ANALYSIS FOR PAPER-MAKERS. 279 It is necessary to take both tlie soluble and insoluble portions in order to obtain uniform results. Standard arsenious acid solution is now run in until tbe solution ceases to produce a blue colour with iodide of potassium and starch papers. A very rapid and satisfactory method is that of causing the bleaching powder to liberate its equivalent of iodine from potassium iodide in presence of hydrochloric acid, using sufficient excess of the iodide to dissolve the titrated iodine completely. The iodine is then titrated in the usual way with a standard solution of sodium thiosulphate. The method of estimating the amount of available chlorine in bleaching powder by determining the quantity of ferrous sulphate it is capable of oxidising is fallacious, as by this means calcium chlorate is included in the result. In bleaching powders of low strength this test becomes, however, of value as a measure of the spontaneous decompo- sition of the powder causing loss of ' available chlorine.' The 'chlorate' strength may be taken as the diflFerence between the numbers for total oxidising chlorine and avail- able or hypochlorite chlorine. Bleaching powder should contain about 35 p.ct. of avail- able chlorine, but the percentage frequently falls below this amount, especially in warm weather. Great care should be exercised in the sampling of bleach- ing powder, as indeed of all chemicals, in order to insure an average result. Small portions should be taken from different parts of the bulk ; the whole should then be care- fully mixed, and, if necessary, reduced to powder. Portions of the mixture should be taken, pounded, and again thoroughly mixed. If this process be repeated once or twice, a perfectly uniform sample will be obtained. Alum, Sulphate of Alumina, Alum Cake, t&c— The points to be considered in an examination of such of these products 280 PAPER-MAKIN( as are perfectly soluble in water are (1) the percentage of alumina contained in them ; (2) the percentage of iron, and (3) the amount of free sulphuric acid present. (1) The alumina is estimated as such by precipitation with ammonia. Ferric oxide is also thrown down with alumina, and must therefore be deducted from the amount of the mixed oxides obtained. (2) The percentage of iron oxide present is determined in the following way. A considerable quantity of the sample under examination is dissolved in water, and to the boiling solution a large excess of caustic potash solution is added"! The caustic potash precipitates both the alumina and ferric hydrates, the former, however, redissolves. The solution is filtered, and the precipitate of ferric hydrate washed, dis- solved in dilute hydrochloric acid, and reprecipitated with excess of caustic potash. It is again thrown on to a filter, washed, redissolved in hydrochloric acid, and again repre- cipitated with ammonia. In this way it is obtained free from alumina.* An alternative method, to be recommended when the iron is present in larger proportion, is the volumetric process with permanganate : the ferric being reduced to ferrous oxide by reduction with zinc in presence of acid, the latter being then quantitatively re-oxidised by the titrated permanganate solution. (3) Free Acid.—T^hm can be estimated indirectly, i.e. by estimating the alumina and ferric oxide, calculating the amount of sulphuric acid necessary to combine with them, and deducting this from the total amount of sulphuric acid estimated. In the case of alum it is of course necessary to make due allowance for the sulphate of potash or ammonia present. The sulphuric acid is estimated as barium sulphate in the usual way. An approximate method of estimating free acid consists * For a metliod of estimatino; minute quantities of iron in alums see • Journ. Soc. Chem. Ind.,' April 1887, p. 276. GENEEAL CHEMICAL ANALYSIS FOR PArER-MAKERS. 281. in digesting a weighed quantity of tlie finely powdered sample in strong alcohol. The alcohol dissolves away the free acid, which can be estimated in the solution by means of a standard solution of alkali. Free acid can be detected by means of a solution of Congo- red, which is turned blue with free acid, but not with pure sulphate of alumina. Insoluble Matter is present in greater or le.'-s proportion in many of the * alum-cakes ' of commerce. Such constituents are of course destitute of ' alum value.' They are estimated by boiling out the sample with water; collecting, washing, drying and weighing the undissolved residue. Sizing Test. — GrifiBn and Little recommend a direct measure of the sulphate of alumina in terms of rosin size precipitated. The following are the practical details of the test as given in their 'Chemistry of Paper-making ' pp. 831, 382. A standard size solution is prepared by dissolving about 25 grammes of good ordinary rosin in about 250 c.c. of strong alcohol. The solution is then filtered from insoluble matter, and diluted with a mixture of 500 c.c. of strong alcohol and 300 c.c. of water to nearly 1000 c.c. A little phenolphtalein in solution i« then added, and standard soda solution added drop by drop, shaking after each addition until a faint pink tinge is observed in the solution. This shows that all the rosin acids are combined with soda, and that the solution is one of neutral resinate of soda or neutral rosin size. The solution is now to be made to 1000 c.c. with the diluted alcohol mentioned above, and if not entirely clear, filtered again or allowed to stand until it settles clear. The clear alcoholic solution constitutes the standard size solution. The value of this solution is next to be determined, best by means of a solution of pure crystallised ammonia alum, one part of which alum we have found to precipitate 2*46 parts of neutral rosin size. For this purpose the clear, colorless crystals should be coarsely crushed in a mortar, and the resulting powder pressed 282 PAPER-MAKING. between two sbeets of filtering paper to remove any accidental moisture. Five grammes are then carefully weighed and dis- solved to 500 c.c. Each cubic centimetre of this solution will then contain O'Ol gramme of alum. Two burettes are next filled, one with the size solution, and one with the alum solution. A flask of 160 to 200 c.c. capacity is filled about two- thirds full of water, and 20 c.c. of the size solution is run into it from the burette. The alum solution is next run in, a few drops at a time the mouth of the flask being closed with the thumb and the flask vigorously shaken after each addition of alum, and allowed to rest until the flocculent precipitate formed has risen clear, which takes but a few moments. The addition of the alum solution should be continued until the precipitate on rising leaves the solution entirely clear, with- out the slightest trace of milkiness or opalescence. The number of cubic centimetres of alum X 0-01 equals the amount of ammonia alum required to precipitate the size in the 20 c.c. of standard size employed. This multiplied by the factor for ammonia alum, as above, equals the quantity in grammes of neutral size in 20 c.c. of the standard solution. The actual test of an alum is performed in exactly the same way ; a solution of 5 grammes of the alum to 500 c.c. being employed, and, if necessary, filtered through a dry filter before titrating. 20 c.c. of the standard size solution are always employed, and the actual amount of neutral size it contains having been determined as above, it is easy to calculate from the data given by the titration the amount of size which one part of the alum tested will precipitate. This test, as is evident, gives the absolute precipitating power of the alum, and does not discriminate between sul- phates of alumina iron, or other metallic oxides which may be present, or free acid, all of which have the power of precipitating size. Antiddor, Sodium Thiosulphate, Sodium Sulphite, &c.— Sodium thiosulphate can be estimated by means of a GENERAL CHEMICAL ANALYSIS FOE PAPER-MAKERS. 283 standard solution of iodine in potassium iodide. The operation should be performed in dilute solution (see p 182). The same solution serves also for the estimation of sul- phurous acid in sodium and other sulphites. Starches. — After having made a careful microscopical ex- amination of the sample, it should be examined for water by drying at 100° C, and for mineral matter by igniting in a platinum crucible and weighing the residue. The water should not exceed 18 p.ct., and the ash '50 p.ct. The samples should be carefully examined for insoluble matter by dissolving in water and filtering. It is also useful in comparing different samples of starch to convert the speci- mens into pastes, under exactly similar conditions. When perfectly cold, the 'stiffness' of the pastes should be compared. This may be done by noting the relative resistance to weights placed upon their surface. Information on this subject is contained in a paper by W. Thomson in the ' Journ. Soc. Chem. Ind.,' March 1886. The identification of starches by means of their micro- scopical appearances is tolerably simple, as each particular kind possesses characteristic differences. Gelatine. — Samples of gelatine should be examined for water and ash. The water should not exceed 16 p.ct. and the ash 2 ■ 7 p.ct. The relative strengths of the jellies formed with water should also be compared. This may be done in the way indicated above for starches. Eaw materials intended for the preparation of gelatine may be similarly examined. They yield the whole of their gelatine on heating with water. The effect of prolonged heating on solution of gelatine should be boime in mind. The amount of residue left after treatment with water should be carefully determined. SoGjas. — The soaps most suitable for use in paper-making are those known as curd and mottled, the former being chiefly used for the fine qualities. They should be carefully examined for free alkali and unsaponified fat. Full descrip- 284 PAPER-MAKING. tions are contained in Lant Carpenter's ' Treatise on Soap- making.' Dyes, Pigments, Loading Materials, &c. — These substances are best examined from a physical and tinctorial point of view, by comparison with specimens of undoubted purity and excellence. The chemical examination is both difficult and somewhat misleading. A microscopical examination will often throw much light on their composition. For comparing the tinctorial power of pigments, weighed quantities should be ground up with starch or some other white powder. In this way the comparison of the intensity of colour is much facilitated. 285 CHAPTER XIV. SITE FOE PAPEE-MILL, WATEE-SUPPLY, WATEE PUEIFICATION, ETC. In choosing a spot on which to build a paper-mill, the manu- facturer has to take into consideration several very important circumstances. Chief of these is the necessity for having a large supply of water at command. Not only is a large quantity needful, but it should be free from impurities, such as suspended matter and soluble iron. The former, it is true, can be removed by settling and filtration ; the latter cannot, and is liable to injuriously affect the colour of the paper. Again, as a question of economy in working, it is advantageous to have convenient water-power ; therefore for this, as well as for the former reason, paper-mills are usually situated on the bank of a stream. In choosing such a site paper-makers are probably also influenced by the fact that it affords a ready means for the removal of impurities. In properly conducted mills, where suitable apparatus is em- ployed for evaporating the liquors in which the raw material has been boiled, the stream should not be polluted to any very great extent. Generally speaking, the greater the pollution, the more are valuable materials being lost to the manufacturer. It is obvious that the site for a mill should also be chosen with reference to its proximity to means of transit for the raw and manufactured materials. The disposition of the different divisions of a paper-mill depends of course upon the nature of the paper it is intended to make and to a large extent upon local circumstances. Fig. 79 will give some idea of the general arrangement of a mill for manufacturing esparto paper. 286 PAPEE-MAKING. Water Purification. — Arnot in his Cantor lectures on the technology of the paper trade states that from 30,000 to 40,000 gallons of water are required for the manufacture of each ton of paper. The importance of a pure and copious supply is therefore very evident. The impurities of water consist of two classes, insoluble and soluble. The former can be readily removed by pro- cesses of settling and filtration. For this purpose most paper mills are provided with large ponds capable of holding seve- ral days' supply. These are sometimes supplemented by filtering beds of sand. In- soluble impurities can also be removed by passing the water through filter-pres.^es. Of late years a number of filters have been devised which, while occupying much less ground space than that occu- pied by settling ponds and filters, effectually remove the greater part if not the whole, of the insoluble impurities. Of such we will briefly describe a few typical in- stances. Many of these are so arranged that they can be SITE FOR PAPER-MILL, WATER-SUPPLY, ETC. 287 readily cleansed and the collected insoluble impurities removed. Wilson's Patent Automatic Self-cleansing Filter as manu- factured by Messrs. Masson, Scott and Bertram is an example. Fig. 80 shows a side sectional elevation of the filter with its automatic mechanis:n. The filter proper is an open top Fig. 80. cast iron cistern containing the filtering medium. Above it is a flushing tank containing the water which cleanses the filtering medium of the impurities arrested and accumulated therein, at intervals varying, according to the state of the water being filtered. Entering through the valve regulating box A, the water passes on to the distributing board B, which runs the whole length of the filter. This serves to distribute the enter- ing water uniformly over the length of the filter. After 288 PAPER-MAKING. percolating through the medium it is discharged into the filter main C. This filtering process continues until the medium becomes silted up with impurities, thereby obstruct- ing the free percolation of the water under filtration, which being thus obstructed, accumulates until it rises to a certain level, when it overflows a syphon which draws off a quantity of the water into the float chamber D, i-ufficient to raise the float E. The float thus raised closes the inlet and outlet valves, and operates a valve in the flushing tank, by means of which the contents are discharged down the large syphon pipe P, and upwards through the filtering medium by means of a series of perforated pipes. The air occupying these pipes and the spaces in the cellular bottom underneath the filter bed being forcibly ejected by the downrush of water, performs a most important function in thoroughly aerating and agitat- ing the filtering medium, thereby separating it from the impurities, which, being lighter than the medium, are floated upward and drawn off by the large syphon H, into the float box, from which it escapes to a drain. "When all the impurities have been drawn off and have escaped from the float chamber, the float falls, causing the various valves to resume their normal position, and the ordinary operation of the filter is resumed. This cleansing of the filter, automatic from first to last, occupies from two and a half to three minutes. The quantity of water used for the cleansing operations is, when filtering water containing an average amount of matter in suspension, about 5 p.ct. of the total amount purified, but when filtering waters that are comparatively pure this amount seldom exceeds 3 p.ct. As the interval of time between the washings de- pends upon the amount of impurities in the water, these filters are very suitable for dealing with waters obtained from sources which are subject to sudden and frequent pol- lution, and also for the treatment of effluent discharges from works which have to meet the demands of the Elvers Pollu- tion Act. SITE FOR PAPER-MILL, WATEIt-SUPl'LY, ETC. 289 Greig's Patent Filtering Apparatus (No. 29,070, 1897) consists of a series of closed i?ettling tanks, connected together at the top by pipes and furnished with hopper-like bottom which serves to collect the insoluble impurities, which can be reihoved from time to tinie by sludge cocks. The last series of settling tanks is connected with a filtering tank provided with a number of compartments filled with ashes' which serves as a filtering medium. These compartments can be removed and replaced by others containing fresh ashes. This filter is also adapted to the treatment of effiuent waters. The Eeeves Patent Filters, manufactured by Messrs. Mather and Piatt, Limited, of Salfurd, is another example of the class of self-cleansing filters. Fig. 81 is an illustration of Roeckners Patent Clarifier. The cylinders C dip below the surface of the water to be purified, which is contained in the reservoir A. Tliey are open at their lower ends, but are closed at the top by the domes D. These are connected with the pump I by means of the pipes H. On starting the pump the water rises slowly in the cylinders. As soon as it reaches the level of the top, the action of the pump is stopped and the cocks K opened, when the water commences to flow down the pipes G, which together with the cylinders form a kind of syphon. If the reservoir A be kept full, the syphons can be made to act continuously. The flow of water being slow, insoluble im- purities have .time to subside. The impurities may from time to time be removed b}' means of the small pump L connected with the bottom of the reservoir. The clarifier can also be used for the purification of the effluent water from paper-mills. The soluble impurities of water consist mainly of carbon- ate and sulphate of calcium and magnesia, together with traces of iron and organic matter. It is a moot point among' paper-makers whether or not the presence of the two former u 290 PAPER-MAKING. in a water is objectionable. For boiling and bleaching purposes it undoubtedly is, as when mixed with caustic soda or bleaching liquor they form precipitates which line the insides of boilers, breakers, and potchers as a hard scale, which is always liable to become detached, and to find its way into the finished paper. Moreover, carbonates of calcium and magnesia are precipitated in the fibre, and carry with Fig. 81. them a certain quantity of colouring matter, the subsequent removal of which is difficult. Carbonate of calcium, though practically insoluble in pure water, is soluble in water containing carbonic acid. When this carbonic acid is neutralised by lime or soda, the carbon- ate is precipitated. The sulphate of calcium is unacted upon by iime, but by the action of caustic soda is converted into SITE FOU PAPER-MILL, WATER-SUPPLY, ETC. 291 free lime and podium sulphate. The lime then neutralises the free carbonic acid in the water, and forms carbonate of calcium, which is of course precipitated. These reactions may be represented by the following equations : — Calcium bicarbonate. Calcium Carbonic r • Calcium carbonate. acid. carbonate. 1. CaCOs + CO, + CaO = 2 CaCOj. Calcium Caustic Calcium Sodium sulphate. soda, hydrate. sulphate. 2. CaSO^ + 2NaOH = CaH^O, + Na^SO, Calcium Carbonic Calcium Water bydrate. acid. carbonate. CaH.,0„ + CO. = CaCOa + H^O. If sodium carbonate be used instead of caustic soda, the decomposition will take place thus : — (Calcium Sodium Calcium Sodium sulphate. carbonate. carlwnate, sulphate. CaSO^ + NajCOj = CaCOj + Nh^SO,. It will be seen from the above equation that'in the boiling processes the lime salts are removed from the water at the ex- pense of an equivalent quantity of caustic soda. The same thing applies also to the salts of magnesia. The amount thus decomposed is not perhaps sufficiently large to make it advisable on that account to purify the water. It is in- deed so small that processes of purification, based upon the use of lime and caustic soda, are now largely used, the cost for chemicals rarely exceeding Id. per 1000 gallons. But for the reasons we have stated above, and also from the fact that it serves to remove dissolved iron and organic matter fi-om a water, such a purification process is in certain cases advisable. The processes now in use consist, as we have indicated, in the addition to the water of lime and caustic soda or carbon- ate of soda, the quantities being regulated according to the hardness of the water and the relative proportions of the carbonates and sulphates of calcium and magnesia. The method of removing carbonates from water by the u 2 292 PAPER-MAKING. addition of lime is due to the late Dr. Clark, of Aberdeen, and the process is still called after him. It is in use in its original form in many places. The plant necessary consists simply of a tank for mixing the water and the lime, and of large settling tanks in which the carbonates of calcium and magnesia subside. Various modifications of this process have been proposed, chiefly in the direction of improved plant. In the Porter-Clark process the precipitate is removed by passing the water after the addition of lime (in the form of lime-water) through a filter-press. In Fig. 82 is shown the Stanhope purifier, which is largely used for the purpose of softening water. The following details will render its action clear : A is a store tank con- taining caustic soda solution. B is a tank into which the wa'er to be treated flows and which is maintained at a con- stant level by means of a valve and float. C and C are two tanks which are used alternately, and in which lima water is prepared. A definite quantity of caustic soda solution is added from B ; this mixture forms the reagent by which the soften- ing of the water is accomplished. D is a small tank for the purpose of maintaining a constant head of reagent. The re- agent and the water are mixed together in the pipe F, the quantities being regulated by the cocks shown at H. The pipe F leads into the vessel E, which is fitted with a number of V-shaped shelves, placed at an angle of 45^, and riveted alternately to opposite sides of the vessel. This arrangement causes the water to take a serpentine course. The position of the shelves is indicated by the dotted lines. As soon as the mixture of water and l eagent flows down the pipe F, and enters the vessel E, the calcium and magne- sium carbonates commence to precipitate and settle upon the V-shaped shelves. At the top of the tank a layer of wood shavings, inclosed in wire netting, is placed as shown by the dotted lines. This acts as a filter, and intercepts any particles that may not have subsided. The clear purified water passes away by the pipe I, 294 PAPER-MAKING. The precipitate that collects on the V-shapod plates is from time to time drawn off by the cocks shown at G ; this may be done without interfering with tlie working of the apparatus. While there can be no doubt that the use of soft water is advantageous in the boiling, washing, and bleaching processes, and also of course fur supplying steam boilers, it is probably an advantage to use a hard water for diluting the pulp before running it into the machine. Continuous softening processes such as that just described are open to the serious objection that unequal results are frequently obtained owing to changes in the composition of the solution of chemicals or of the water itself. Such an objection is obviated by the process of Messrs. Archbutt and Deeley, the plant for which is manufactured by Messrs. Mather and Pktt, Ld. of Salford. The plant consists of two large tanks, one of which is being softened, while at the same time the other is being run off. The precipitate formed by the action of the lime and alkali used in softening is allowed to accumulate up to a certain point, and by the ingenious device of blowing up the precipitate by means of a current of steam and air it has been found that the freshly formed precipitate is carried down with such rapidity that practi- cally clear water can be drawn off in about half an hour. By this means considerable economy in tank space is effected. Each tank-full of water can be tested and the proper amount of chemicals adjusted so that uniformity in composition of the softened water results. In the case of papers which are loaded with any of the forms of calcium sulphate (pearl-hardening, crystal-hardening, &c.) the use of very ?oft water is objectionable from the fact that a certain quantity of calcium sulphate is dissolved. This would not take place to the same extent with water which is already charged with sulphate and carbonate of calcium. The removal of soluble iron fiom a water is effected by the softening processes described. 295 CHAPTER XV. ACTION OF CUPBAMMONIUM ON CELLULOSE. PEEPAEATION OF WILLESDEN PAPER. The action of a solution of copper oxide in ammonia upon cellulose has been already referred to (see p. 10). Celluloses, and also certain lignocelluloses, when treated with such a solution, gradually gelatinise and finally dissolve. On evaporating the solution to dryness, a gummy amorphous- mass is obtained, containing the cellulose intermixed with copper oxide. If the cellulose be in excess, e. g. when the solution is evaporated on the surface of paper, calico, &c., merely dipped in the solution, the copper oxide is often not formed at all. but a green varnish-like mass of cellulose combined with copper oxide, which coats the surface of each filament, weld- ing and cementing them together. This cement-like cupro- cellulose, as it may be termed, being insoluble in water com- municates water-resisting properties to the material so treated : moreover, the presence of copper renders the fabric less prone than before to be attacked by insects and mould, so that animal and vegetable life of a parasitic nature and fungoid growths are rarely, if ever, to be observed, even when the material is kept under conditions where boring worms, ants, rot and mould, would be likely to attack it. As has been already indicated (p. 9), the solution of cuprammonium hydroxide is preferable to one containing cuprammonium salts; not only is the action on cellulose more energetic, but various other advantages are obtained. Preparation of the Copper Solution.— The cuprammonium 296 PAPER-MAKLNG. solution is prepared according to the patent of Dr. C. E. Alder Wright (No. 737, 1883). A series of cast-iron towers, two to three feet in diameter and ten to twelve feet high, is so arranged that a current of air can be blown by a powerful engine successively through the whole series. The towers are then nearly filled with fragments of metallic copper (crumbled up sheet, out- tings, &c.), and solution of ammonia ; the air-current being turned on, oxidation of the copper and solution of the oxide so produced are rapidly effected. The ammonia solution employed is previously impregnated with a considerable quantity of copper by passing a stream of water over scrap copper in similar towers, a current of ammonia gas mixed with a suitable proportion of air being at the same time forced in. The liquor passing from one tower is used over again to supply a second instead of water, and so on through the series : finally the liquor is brought up to full copper strength in the series of towers described above. The spent air issuing from the towers carries with it a notable quantity of ammonia ; this is intercepted by means of an "exhaust" scrubber containing copper, and well sup- plied with water, whereby a comparatively weak solution of ammonia and copper is obtained, which can be used instead of water in the first series of towers. In order to produce the maximum effect on the cellulose, the solution should contain from 100 to 150 lbs. of ammonia, and from 20 to 25 lbs. of copper per 100 gallons. By decomposing a cuprainmonium solution by means of metallic zinc, a corresponding solution of zinc-ammonium hydroxide can be obtained. This solution is also capable of gelatinising cellulose, but not to the same extent as the capper solution. It may, however, be advantageously em- ployed in certain cases in conjunction with a copper solution. Such a solution can be conveniently prepared by substituting brass for copper in the dissolving-towers. It is worthy of note that although, as we have seen, zinc has the power of ACTION OF CUPKAMMONIUM ON CELLULOSE. 297 replacing copper in a cuprammonium solution, iron is without any action, although it readily replaces copper in a solution of copper sulphate. This fact is of very great industrial importance, as it enables the manufacturer to employ vessels and machinery of iron. The ' Willesden ' goods may be divided into two classes. Goods of the first class, such as rope, cordage, netting, &c., are prepared by simply dipping the made-up materials in a bath of cuprammonium solution, using certain precautions as to the mode of immersion and its duration, and the strength of the solution. On subsequently drying the dipped fabrics, they are obtained coated and impregnated with cupro-cellulose, which thus not only forms a kind of varnish- like surface dressing, but furthur adds strength to the fibres by more or less intimately cementing them together. Goods of the second class constitute a much more impor- tant group to which at present the Willesden Company more especially devotes its attention. These fabrics are essentially of three kinds, viz. Willesden Canvas, Willesden Scrim, and Willesden Paper. The two former of these classes are pre- pared in much the same way as the goods just described, saving that the fabric to be treated is usually unwound from one roller and rewound upon another, after passing through the bath and over a succession of drying cylinders. Willesden Paper. — This may be divided into two depart- ments, viz. (1) Willesden unwelded ; (2) Willesden welded, the first class being prepared from a single web of paper by passing it through the bath, rolling and drying. Certain coarse varieties furnish a waterproof material excellently adapted for lining packages, &c. Finer qualities furnish envelopes and stationery, possessing the valuable property of not being aifected by water. Letters written on such paper would remain perfectly legible, even after prolonged im- mersion. It may be interesting to point out here that the cuprammonium solution offers a very simple means of fasten- ing envelopes in such a way as to be proof against any attempts at tampering. The method consists in using a 298 PAPER-MAKING. concentrated solution as the fastening material ; the envelope is then closed and ironed with a warm flat-iron, whereby the gelatinised cellulose is converted into an insoluble cupro- cellulose, and the cover is fastened so securely that the only possible mode of opening is by tearing the paper. Willesden Welded Goods aie prepared by simultaneously dipping two or more sheets of paper and pressing Ihem into one compact homogeneous sheet whilst the surfaces are still in a gelatinous state. In this way a continuous length of fabric of extraordinary durability is produced that is scarcely affected by water, even when heated in it for some weeks at a pressure of 60 lbs. per square inch. As, on drying fabrics treated with cuprammonium solution, the whole of the ammonia in the solution absorbed by them is volatilised, it is necessary, in order to make the process economical, to collect and re-use this ammonia. This is effected by conducting the drying process in closed chambers, from which the ammonia gas is conducted by suitable ap- pliances and recovered. In addition to the advantages already mentioned, it may be stated that Willesden paper is much less inflammable than ordinary paper. Among the many uses to which Willesden goods can be applied, the following may be cited :-Eoof« and sheds ; huts and tents; partitions; tanks and pipes; damp-proof founda- tions ; underlining slates ; sails, awnings, &c. &c. 299 CHAPTER XVI. STATISTICS. The statistical study of paper-making may be extended by the specialist to very wide limits, covering all numerical returns concerning raw materials, as well as the production, distribution and consumption of paper. Of these, certain only may be regarded as indispensable to the paper-maker, viz. (1) general, comprising such figures as the Board of Trade returns for the trade of the United Kingdom in raw materials and finished papers ; (2) special, which are the numerical returns for each mill showing the expenditure in quantities and values for the aggregate or unit of output of finished paper. Under the first head of general statistics we retain those prepared for the first edition of this work, as they have not lost their general significance and have a permanent value as records of the trade of fifteen years ago. By comparison with corresponding figures for the present time, which are added, some useful conclusions may be drawn as to the general tendencies of the manufacture and consumption of paper. Maw Materials. The statistics concerning the trade in raw materials for paper-making are not published in a detailed form, all the various kinds of fibre, being grouped under two heads, viz. rags (linen and cotton), and esparto (which embraces other fibres, and probably wood pulps) The figures for recent years are shown in the accompanying tables : — 300 PAPEK-MAKING. (a) Imports of cotton and linen rags : 1880. 1881. 1882. 29 , 642 tons. 26 , 773 tons. 21,200 tons. 451,782/. 396,274i. 303,349L 1883. 1884. 1885.1 28,543 tons. 36,233 tons. 35,470 tons. 401,922i. 487,866L 466,928/. (h) Esparto and other material : 1880. 1881. 1882. 228,580 tons. 238,043 tons. 251,594 tons. 1,642,903/. 1,626,800/. 1,784,078/. 1883. 1884. 288 , 549 tons. 260 , 544 tons. 1,943,732/. 1,638,564/. But a portion of these imports was exported again, the figures being as follows : — (a) Exports of rags and other paper-making material produced in the United Kingdom : 1880. 1881. 1882. 55,792 tons. 50,488 tons. 49,352 tons. 673,523/. 563,460/. 526,554/. 1883. 1884. 51,293 tons. 60,924 tons. 502,851/. 562,903/. (&) Exports of ditto of foreign and colonial produce : 1880. 1881. 1882. 6,965 tons. 10,183 tons. 7,004 tons. 102,499/. 146,101/. 84,515/. 1883. 1884. 11,561 tons. 26,498 tons. 121,992/. 274,664/. The relative proportions of raw material furnished by the various countries in 1884 (the latest available return) are shown in the subjoined tables : — STATISTICS. 301 (a) Imports of linen and cotton rags in 1884 : from Tons. £ ^ermany 21,280 294,883 §0"?°^ 4,406 56.537 ^elgium 2,867 34,713 ^"^^ey 2,036 16,228 1,875 28,133 nu^^'^ Tt ," , 1'784 33,293 Cnannel Islands 28'' 3 972 ^oiway 261 3^377 3,020 British South Africa 13 1,615 Australasia 113 2' 245 Canary Islands 96 1 ' 043 Spain •• - '.. 93 i;o09 Argentine Kepublic 64 1 280 Other Countries 74 5 6,518 36,233 787,866 (6) Imports of esparto and other fibres in 1884: From Tons. £ ^^S^ria 88,357 515,232 ^Pi^i°. 40,159 314,927 Tripoli 33,930 172,282 J^iyi's ••. 20,526 117,374 Other countries 1^033 5,733 184,005 1,125,553 (c) Imports of other materials, including rag and wood pulps, in 1884 : From Tons. £ ^oivf&y 47,923 276,204 gegium 7,169 58 383 Holland 6,484 46,104 Sweden 5,178 48,523 ^^rmaxiy 3,732 36,532 France.. 2,128 14,667 ^'205 13,332 British India 1,046 7 735 ^Spt .... 870 4 ,'933 Other countries 804 6,544 76,539 513,011 302 PAPEE-MAKING. A considerable export of raw material for paper-making also takes place from Britisli shores, the figures, embracing rags and other materials, for 1884 being as below :— (a) Exports of home produced rags and other paper-making materials in 1884 : To Tons. £ United States 59,222 550,924 Holland ^49 British America 390 4,255 Germany 262 1,673 Other countries " 401 60,924 562,903 (6) Exports of foreign produced ditto in 1884: (i.) Linen and cotton rags. To Tons. £ United States U.m 167,801 Other countries 514 14,705 173,320 (ii.) Esparto and other fibres. To Tons. £■ All countries HO 701 (iii.) Pulps and other materials. To Tons. £ United States 11,290 97,620 Other countries ^'^^ 6,^)^0 11,653 100,643 Our imports of ' rags and other paper-making materials ' from Norway increased from 23,483 tons, value 138,098Z., in 1880, to 48,199 tons, 279,679Z., in 1884, mainly owing to the development of the wood pulp industry. Later figures will probably show a similar increment. The same articles from Denmark grew from 359 tons, 3406/., in 1880, to 1408 tons, STATISTICS. 303 16,352/., from the same cause. German statistics have even a stronger iipward tendency : 11,587 tons, 196,051/., in 1880, against 196,051 tons, 331,591/., in 1884. Our receipts of esparto from Algeria show an advance, but not a marked one, being 60,6 12 tons, 421,343/.,in 1880, and 88,357 tons, 515,232/., in 1884. But shipments of the same fibre from Spain show a decline from 51,413 tuns, 454,713/., in 1880, to 40,159 tons, 314,927/., in 1884; and from Morocco they have fallen away from 2879 tons, 18,231/., in 1880, to 260 tons, 1290/., in 1884; while the figures relating to Tunis and Tripoli also show a disposition to recede, though not at such a rapid rate. The wood pulp industry of Norway for the year 1886 shows a very large increase upon the figures of a few years back, albeit prices have ruled very low. This latter circum- stance is attributed, not so much to over-production, as to excessive competition among the sellers of this article. The quantity exported during the year 1886 is about 120,000 tons ; in the year 1885 it was 107,651 tons; 1884, 88,220 tons'; 1883, 70,464 tons; 1882, 68,884 tons; 1881, 42,194 tons • 1880, 26,055 tons. Several of the old works have extended their production during the past year, and several new establishments are in the course of erection, so the produc- tion this year may probably be put at 150,000 tons; wood pulp with 50 p.ct. water. The greatest part of the Nor- wegian wood pulp is exported to England, P'rance and Belgium; in Kussia, the increase in the duty has stopped business, and the same can almost be said of Germany. America, too, has drawn part of her supply from Norway, but this trade is not 'expected to continue. 304 PAPER-MAKING. a 00 a; o oc -T- a; 00 J'l ■*! O t-- ■M 00 C5 X !M OO " O a, t>co CO ~ J; l> O O O O O I— C0V0t>Cl0OTfi{>4OC2CCOtN 5lf5COr-JOOCDO:~l>OC;COCC L^^CMCOiOwi — l>03C H l>!NiM>C-HCOI:~^l> O-H— if-i— i^OOCOODCOOC; r> CO coooeoao 2aop^(jJ"^cq'co'o o o ^ m o COrfi-^iCOCOC^fM^CNfMC^CN t--ocaiO'-i ^ 5c? — DO 5 c.~ ^ 1 •= 59 g § » So a-n-^ P p 8-2 ® •tj =<- G, c: 5f g g -5 .So 03-1 M ^ * CJ a ^ 5 oT o or JO ^ MM S a. STATISTICS. 305 Paper Trade of the United Kingdom. The following classification of production will be of interest. It represents the results of very careful computa- tion from statistics for the 3- ear 1895 : Brown 20 Grey (wrappings) 23 Printings (news) 16 (fine) 21 Writing 20 100 It is interesting to note that these five main classes of papers are produced in approximately equal tonnage. Manufactured Matet ial. Our import trade in manufactured paper has been grow- ing of late years, as the following figures will show : — Imports of paper and pasteboard, of all kinds except hangings : 1880. 1881. 1882. 1,021,952 cwt. 1,065,912 cwt. 1 ,098, 118 cwt. 1,159,646/. 1,138,943/. 1,202,905/. 1883. 1884. 1,160,104 cwt. 1,447,335 cwt. 1,245,861/. 1,403,446Z. At the same time, our exports of manufactured paper, though amounting to only about one-third the weight, have very nearly as high an aggregate value as the imports, and have increased in much the same proportion in corresponding years. Thus : — Exports of paper of all kinds but hangings : 1880. 1881. 1882. 472,168 cwt. 555,219 cwt. 584,947 cwt. 1,106,996/. 1,242,962/. ,1,305,025/. 1883. 1884. 597,923 cwt. 670,760 cwt. 1,284,862/. 1,374,392/. X 806 PAPER-MAKING. The above figures relate to paper manufactured in the United Kingdom. In addition, there were re-exports of paper of foreign and colonial manufacture to the following amounts : — Ee-exports of foreign-made paper : 1880. 1881. 1882. 68,843 cwt. 68,861 cwt. 50,403 cwt. 10S,426Z. 101,197^. 82,464i. 1883. 1884. 49,526 cwt. 51,467 cwt. 76,620/. 78,785Z. Taking the latest year for which detailed statistics are available, viz. 1884, we find that the importations of paper are derived from foreign states in the following propor- tions : — Imports of paper in 1884 — (a) Writing or printing paper : From Uwt. £ Germany 92,681 129,112 Belgium 42,139 68,370 Sweden 27,747 36,682 Holland 26,507 34,884 France 6,463 28,92.=) Austria 5,399 9,314 Norway 4,300 4,771 Other countries 3,174 5,655 208,410 317,713 (h) Unenumerated sorts : From Cwt. £ Germany 139,156 209.916 Sweden 116,488 120,793 Belgium 82,263 124,476 Holland 54,594 84,699 Norway .. 32,294 26,765 France' 19,328 69,763 United States 10,^80 31,201 Austria 5,927 12,012 Spain 73 J 4,059 Japan 421 4,736 Other countries 3,547 5,995 465,734 694,415 STATISTICS. 307 (c) Pasteboard and millboard : From C*t. £ Holland 439,749 201,894 Germany 235,658 116,185 Belgium 51,235 25,350 Sweden 32,331 20,241 Norway 8,260 5,726 France 3,262 19,259 Other countries 2,696 2,663 773,191 391,318 The exports of home-made paper in the same year were distributed as follows: — Exports of home-made paper in 1884 — (a) Writing, printing and envelopes : To Bombay and Scinde British South Africa British North America Madras Argentine Eepublic United States .. Sweden and Norwuy Belgium , Germany British West Indies Ceylon Cwt. £ 280,024 545,830 44,427 101,833 33,963 71,834 29,767 64,227 16,900 33,041 14,871 44,511 10,532 25,651 8,683 30,954 6,985 31,722 6,101 11,870 5,981 13,926 5,969 14,188 4,706 13,411 4,668 10,919 31,397 93,454 504,973 1,107,371 (b) Pasteboard and millboard : To Cwt. £ Australasia 14,542 20 771 Bengal and Burma 5,147 4^589 Bombay and Scinde 4,572 4 I3i France 987 2, '309 Other countries 5,571 13,367 30,819 45,170 X 2 308 PArEU-MAKlNG. (o) Unenumerated, and articles made as paper : ij To Cwt. £ ! Australasia 6>,731 81.385 1 British South Africa 9,341 11,809 '. Bombay anrl Scinde 7,146 10,389 Argentine Republic 6,283 8,113 ii Germany 5,147 11,394 I Belgium 5,122 11,300 i| France 5,052 16,238 H Bengal and Biii-ma 5,006 6,678 United States 4,841 14,668 ^ British America 3,737 7,801 I Holland 2,410 5,529 I Oth«r countries 18,152 36,547 134,968 221,851 The re-exports of foreign and colonial-made papers during the same period went to the following destinations : — Ee-exports of foreign paper in 1884 — (a) Printing and writing : To Cwt. £ British India 10,488 14,643 Australasia 4,908 7,695 Other countries 3,2:'.6 7,554 18,632 29,892 (b) Unenumerated: To Cwt. £ British India 5,852 7,354 Egypt 5,167 7,751 Australasia 2,131 4,398 Russia 1,560 7,829 Otlier countries 8,465 15,825 23,175 43,157 (c) Pasteboard and millboard All countries Cwt. 9,660 £ 5,736 1 STATISTICS. 309 (2) Special. — Classified returns of the working of a paper mill and its several departments, are just as necessary for the full control of the manufacture as are the ' account books' with the periodical balance sheet, for the proper working of the ' business ' of the mill. Snch returns fall obviously under the main heads of raw materials, labour, chemicals and steam. The further subdivision of the returns will take into account the various sections or departments of the mill; thus, the 'boiling' or half-stuff plant; bleaching, beating and machines, will be separately dealt with. In an esparto mill there will be special returns for the soda recovery plant, and in a sulphite mill for the liquor-making plant. The returns for boilers and steam raising will be controlled by periodical tests of evaporation in the boilers, and the combustion of coal in the furnaces by analysis of the flue gases. The purpose of these returns is obviously to secure ' efficiency and economy ' in all departments of the mill, to indicate all sources of waste or ' leakage,' and what is of course very important, to provide a statement of cost of QXQYy kind and quality of paper turned out. Such considera- tions are generally recognised and need not be further insisted on. There is, however, an aspect of statistical 1 eturns which is too often lost sight of, and that is that they provide permanent records of manufacturing experience. Such records have a practical usefulness in two directions, they provide against bad work, and for progressive improvement based upon continual observation of effects. In regard to the planning and carrying out of an adequate system of statistics and records, only general instructions can be given. From the point of view of the husiness, the one purpose is to arrive at costs per pound or ton ; from that of the manufacture the purpose is to give an account of every pound of material taken into work, in terms of final manufactured products, and as regards quality to refer the various items of quality of papers to the conditions obtaining in their production. 310 PAPE?.-MAKING. The latter order of statistics are those with which we are mainly concerned : they are the more comprehensive and include such returns from which actual money costs are calculated. We venture to think that the careful study of the preceding chapters of this work will supply the main outlines of a scheme of returns required, in any particular case. The knowledge which we have endeavoured to convey is based upon quantities and measurements ; and that know- ledge in practice implies the continued quantitative study of all the processes of the mill, as well as of the raw materials and finished papers. To take as an example the treatment of esparto : the following quantitative observations may be made. Yield of half stuff — including yield of fibre from boilers after washing, and from potchers or press-pate after bleach- ing ; proportion of caustic used in boiling ; amount of dilution in boilers by condensation of steam ; proportion of t^oda recovered ; proportion of bleaching liquor used in potchers, and the amount of residual bleaching ' chlorine.' At the beaters, every item of the ' furnish ' is carefully controlled, together with the agents used in antichloring, sizing and colouring ; on the machine, in addition to a return of its turn- off in finished paper, observations are made upon the back water to determine fibre, loading, sizing and colouring con- stituents passing away unfixed by the wet web of paper: lastly the physical constants of the finished paper, which, in connection with records of all the previous treatments, gives a complete history of all that makes for quality. It is unnecessary to prosecute this subject into minute detail. We repeat in conclusion that the paper-maker is only a ' technologist ' in his art when his experience issues in the practice of precise records, and his control of a niill is based upon these essential statistics of production. 311 CHAPTEE XVIL BIBLIOG K APH Y. J. Murray. Practical Kemarks on Modem Paper. Edinhurgli: 1829.- L. S. Le Normand. Manuel du Fabricant des Papiers. Paris : 1834. G. Planclie. L'Industrie de la Papeterie. Paris : 1853. L. Miiller. Die Fabrikation des Papiers. Berlin : 1855. Prcteaux. Manufacture of Paper and Boards. Philadelphia: 1866 C. Hofmann. Manufacture of Paper. Philadelphia : 1873. Hugo Muller. Pflanzenfaser. Leipzig : 1873. T. Routledge. Bamboo considered as a Paper-making Material. London and New York : 1875. Vetillart. Etudes sur les Fibres Vegetales. Paris: 1876. Arnot. Technology of the Paper Trade (Cantor Lecture, Society of Arts). London: 1877. J. Dunbar. The Practical Paper-maker. London : 1881. 312 PAPEE-MAKING. Forestry and Forest Products hibition). C. F. Cross and E. J. Bevan. Cellulose.* E. Parkinson. A Treatise on Paper. C. T. Davis. Manufacture of Paper. C. F. Cross and E. J. Bevan. Cellulose, Griffin and Little. Chemistry of Paper-making. Tomlinson. Manufacture of Paper. Paper Eecord. Paper-makers' Circular. Paper Trade Keview. Paper-makers' Monthly Journal. Paper Trade Journal. Papier Zeitung. (Edinburgh Forestry Ex- Edinburgh: 1884. London : 1885. Preston: 1886. Philadelphia: 1887. Longmans, London : 1895. London : weekly. London. London. London : since 1872. 'New York : since 1872. Berlin, since 1876. * Gives a list of works on Cellulose and its derivatives. 3i3 INDEX. Abietic acid, 185 Acetates, 28, 32 Acetic acid, in bleaching, 162 formed in boiling esparto, 131 Acid, action on paper, 229 on paper, evils of, 229 — and cellulose, 13, 18 — and lignocelluloses, 58 — ether, 18 — in alum, estimating, 280 — oleocutic, 75 — purification of fibres, 93 — radicals and cellulose, 28 — size. Beadle's formula for, 186 diluting with starch, 186 quantity required per 100 lbs. fiaper, 167 burster's formula for diluting, 186 — stearocutic, 75 — subtric, 76 — succinic, 76 Acme beater, 174 Adansonia, application of, 144 — l ast of, 88 — digifata, 102, 146 — fibre, 103 — soda treatment, 146 — treated with lime, 146 Adipo-celluloses, 53, 73 Adulteration of paper, 201 Agalite, analysis of, 200 — loading, 200 Agents, bleaching, 153 Alkali boil for rags, 115 — cellulose, 17 soluble, 84 xantbate, 22 Alkali solutions, technical applica- tion, 17 — treatment, China grass, 146 flax wastes, 144 hemp wastes, 144 jute, 144 Kamie, 146 Rhea, 146 Alkaline hydrates and cellulose nitrates, 32 extreme action of, 44 — hydrolysis, 92 — processes, advantage of. 111 — solutions, 57 Alkalis and cellulose, 11, 14, 16 Alizarin, 206 Aloe, 55 — fibre, 81 — fibres, 71 dimensions of, 88 — fibro-vascular bundles of, 89 Alum, 187 — analysing, 279 — actioTwof, on gelatine, 228 on cellulose, 229 — cake, analysing, 279 — cakes, 195 — crystal, 195 solutions, table of strengths, 195 — proportion required, 196 of, required by gelatine size, 229 — sulphate, analysing, 279 Alumina as a sizing agent, 190 — sulphate, 195 advantage of, 196 proportion required, 196 Aliiminate of soda as a sizing agent, 192 314 TAPE U-M AKIN G. Aluminic sulphate solutions, 185 solutions, specific gravity of, 195 Ammonia soda, caustic, 247 Ammoniacal cupric oxide, 9 Am\liii, 15 Amyloid, 20 Analyses of caustic soda, 245, 246 Analysing alum, 279 cake, 279 — alumina sulphate, 279 — antichlor, 282 — bleaching powder, 278 — caustic soda, 278 — dyes, 284 — gelatine, 283 — lime, 277 — loading materials, 274, 284 — paper, 270 — pigments, 284 — recovered soda, 278 — size, 275 — soap, 283 — soda ash, 278 — starches, 283 — soilium sidphite, 282 tiiiosulphate, 282 — wood-pulp, 272 Analysis of plant hubstances, 90 — proximate, for raw mati rials, scheme of, 92 Aniline blues, 203 — colours, solutions of, 80 — sulphate solution, 85 Annandale's damping methoil, 225 Antichlor, analysing, 282 — hydrogen peroxide, 183 — proportion requireil to neutralise bleaching powder, 182 — removing bleaching liquor by, 181 — sodium hyposulphite, 181 sulphite, 182, 183 thio.-sulphate, 181, 183 — testing, 183 Apparatus for bisulphite processes, 139 •— for determination of the thick- ness of paper, 2G9 Arabic acid, 68 Arabinose, 51, 68 Archbutt-Deeley process for soften- ing water, 294 Artificial silk, 32 Abh in fibre, 92 — of paper, estim;.ting, 201 A-pen, for mechanical wood pulp, 151 Available chlorine in bleaching powder, 156 Baoterium aceti, 36 Bamboo fibre, 108 — pulp, yield, 145 — stems, 72 Bambusa, 145 Banknote paper, 211 Baobab, 102 — fibre, 103 Barley straw, 132 Barium sulphate, estimating, 274 Bast fibres, 82 dicotyledonous. 95 dimensions of, 88 — filaments, n2 — tissues, 102 Beadle's formula for acid size, 186 B jater, acme, 174 — bed plate, 171 — UuUanJer, 172 — knce-|.lat( s, 171 • — knives, 170 cutting in situ, 171 — lead-lined, 170 — Marshall's en;;ine, 176 — refining engine, 176 — Umpherston's patent engine, 179 Beators, 166 — driving of, 180 rope pulleys for, ISO — saving of driving powi r, 174 Beating, esparto, 166 — importance of, 170 — proportion of water to" pulp, 168 — pulp, 166 — straw, 166 — time required by, 168 — water, quality of, 180 IKDEX. 315 Beating wood pulp, 1G6 Bed-plate of beater, 170 Beet juice, 36 Benzoates, 28 Bertram's damping cylinders, 225 — rag-boiler, 115 — smoothers, 224 — strainer, 214 Bibliography, 31 1 Bisulphite liquor. Griffin and Little's standard, composition, 140 preparation of, 139 — processes, apparatus for, 139 — process, boiling, 141 digesters, 140 pressure for boiling, 141 preparation of wood, 140 products of, 141 time for boiling, 141 waste liquors, application of, 142 — wood industry, 138 Bleached pulp, freeing from bleach, _ 162 improving colour, 161 Bleaching, 153 — acetic acid in, 1 62 — action should be confined to non- et llulose, 161 — agents, 153 — by chlorine, 162 — carbonic acid in, 163 — causes of damages in, 42 — tleclrolytic, 163 estimating cost of, 104 — Heruiite process, disadYantag( s of, 164 economy of, 163 — hydrochloric acid in, 158 — jute, 158 • — liqunr, dear, 156 storing, 156 — — strength, 156 — proportion of aiitichlor required to neutralise, 182 ~- po\\der, 153 action of, accelerated by bicar- bonate of soda, 159 amount needed, 159 analysing, 278 chkrine available in, 156 Bleaching powd( r, electrolytic solu- tions, economy of, compartd with, 164 Griffin and Little's composi- tion of, 154 neutralising by hydrogen p^ r- oxide, 183 by sodinm hyposulphite, 181 by sodium snlpLite, 182 by sodium thiosulj hate, 181 preparing, 155 sulphite for neutralising, 183 — — sodii:m thiosulphite fur neu- tralising. 183 — • — storing, 155 — l)ulp in potchers, 158 water required, 158 — rags in the tumbler, 159 Thompson's process, 163 refractory j ulps, 161 — • removing by antichlor, 181 from half-stuff, 181 — time required by, 160 — washing out from half siufl', 181 — with sodium hypochhirite, 160 Blotting papers, 183 starcliing, 184 testing, 269 Boihmeria cellulose, 45 — species, 71 Boiler for esparto, ] 27 Boilers for rags, 115 Boiling flax, 143 — hemp, 143 — jute, 103 — Manila, 143, 145 — rags, 115 — straw, 1 34 — Wdod, 137 Bombax heptaphyllum fibre, 88 Bottoms caustic, 244 Broke paper, 146 grinding-mill, 147 using up, 147 Bromine and ligno cellulose, 55 Bromoform, 43 Broussonetia ksempferia, 103 — papyrifera fibre, 103 Brown pigments, 206 Burning recovered soda, 256 316 PAPER-MAKING. Calcium carbonate mud, 261 — liy|)oclil(>rite, means of accelerat- ing action of, 158 — sulphate, estimating, 27-1 in paper asli, 202 loading, 197 Calendering, 233 Calenders, 224 — deflection, 237 Calotropis, 46 Cannabis sativa, 45 Capillary transmission of liquids, 12 Carbohydrates of low niolei-ular weight, oxidation of, 40 Carbon, percentage in fibres, 93 — tetra- bromide, 43 Carbonate of soda, 247 Carbonic acid bleaching, 163 Carbonising furnaces, 256. — recovered soda, 256 Hammond's furnace, 256 Cartridges, 145 Casein as a sizing agent, 193 Catalogue, Bertram's illustrated, of paper-mill machinery, 176 Causticisers, 259 Causticising, 2o8 — by ferric oxide, 247 — lime used, 261 — testing the liquor, 260 Causlie soda, analyses of, 245, 246 analysing, 278 and outo-celluloses, 93 and ligno-celluloses, 93^ and pecto-celluloses, 93 boiling with esparto, 123 rags, 116 commercial forms, 244 concentrated solutions, 247 drums, emptying, 246 prices of, 245 substances removed from es- parto by, 130 Cellular tissue in straw, 136 Cellulith, 169 Celluloid, 32 Cellulose, 5 — acetates, 32 technical applications, 33 — action of alum on, 229 — alkali-cellulose-xanthate, 22 — alkalis and, 14 Cellulose and acid radicfils, 28 — and acids, 13, 18 • — technical applications, 20 — and alkalis, 11, 16 — and colouring matters, 12 — and enzymes, 15 — and hydrolytic agents, 10 — and sulphuric acids, 34 — and water, 7 — atmospheric oxidation of, 41 — henzoates, 34 mercerised cellulose, 31 soluble alkali cellulose, 34 — calotropis, 46 — classification, 45 — compound, 52 types of, 73 — con^ititution of, 35 — conversion of, 16 of, into hydro-cellulose, 38 — cotton, 45, 46 — decomposition of, 37 — destructive resolution by fermen- tation, 44 — dioxythiocarbonnte, 22 — empirical composition of, 6 — esparto and zinc-sodium hypo- sulphate, 50 — essential identity of some, 46 — esters, 28 — estimating in fibres, 90 — fl IX isolating, 69 — from esparto, 49 and hydrofluoric acid, 49 lignified tissues, 47, 48 woods, 47, 48 — group, the, 45 — liydrates, 7 as a sizing agent, 190 — in fibres, 92 — in raw material, assay for, 90 — isolating from cork, 73 wood, industrial processes, 63, sulphate process, 64 with alkalis, 64 dilute nitric acid, 63 halogens, 63 — and nitric acid, 63 neutral sulphites, 66 sulphurous acids, 64 INDEX. 317 Cellulose, isolating from wood with water and dilute acids, 64 cellulose with bisulphite, 65 — Marsdeiiia tenac issima, 46 — nitrates, 28 hexa-nitrate, 30 penta-nitrates, 31 technical applications, 32 tetra and tri-nitiates, 31 — nitrogen free, 32 — non-hydrolysable, 68 — of China grass, 45 — of flax, 45 — of hemp, 45 — oxidation, 38 extreme action of alkaline hydrates, 44 — — in acid solutions, 38 in alkaline solutions, 41 by chromic acid, 39 by hypochlorites, 41 by nitric acid, 38 by other acids, 40 — — permanganates, 43 — oxycellulose, 25, 47 — pine-wood, 48 — presse-pate system for purifying, 162 — proportion in esparto, 129 of, ill well-boiled pulp, 160 — purification of, 91 in the laboratory, 46 — quantitative re?:enoration of, from solution as thio-carbon- ate, 24 — reacting unit of, 26 — redistribution, 44 — regenerated, empirical composi- tion, 25 general properties, 25 hygroscopic moisture, 25 — Schultze's method, for isolating from wood, 63 — solutions of, 7 — — in ammonia cal cupric- oxide, 9 of hydrolytic agents, concen- trated forms, 16 of technical applications, 8, 10 — Sunn liemp, 46 — sulpho-carbonates of, 21 I Cellulose, sulpho-carbonates, tech- i uical applications, 27 — syntiiesis, 36 — yield iiom esparto, 129 straw, 132 sulphite pulp, 142 — Xiinthates, coagulations by heat, 24 spontaneous decomposition, 23 Chapman's evaporator, 252 Chemical analysis of plant sub- stances, 90 — characteristics of principal raw mat( rials, 95 — wood pulp, 37 China clay, 27, 197 analysis of, 200 estimating, 274 — grass, 71, 95, 100, 146 cellulose, 45 composition, 100 fibres, dimensions of, 80 forms in which employed, 101 general chemical 'character- istics, 100 micro-chemical reaction, 100 microscopic characteristics, 100 Chloride of linie, preparation of, 153 Ciilorinating boiled jute, 144 — flax w astes, 144 — hemp wastes, 144 — straw pulp, 136 Chlorination of ligno - celluloses, 158 Chlorine, 154 i — as a bleaching figent, 162 I — available in bleaching powder I 156 I — liquefied, 162 ! — residual, 91 I — water solution, 85 Cleaning esparto, 123 — straw, 133 Coal-tar dyes, 42, 204, 206, 207 Cochineal liquor, 204 Cocos nucifera, 52 Coffea ai abica, 52 Collodion pyroxvline, 31 Colloids, 27 318 PAPER-MAKING. Colopliony, 185 Colour, improving 161 Colouring matters and cellulose, 12 proportions of, 204 — paper, 2 )2 Colours, of wood pulps, 153 Conifer pulp, 109 Cooling paper, 225 Corchorus, 53 Cork, 73 — isolating cellulose from, 73 — results of elementary analysis of, 73 Cotton, 95 — beating, 167 — cellulose, 45, 46 — classification, 95 — composition of raw fibres, 96 — fibres, dimensions of, 88 — forms ill which employed, 97 — general chemical ciiaracteristics, 96 — in paper, detecting, 270 — micro-chemical reaction, 96 — microscopic features, 95 — pulp, dimensions of fibre, 169 — rags statistics, 299, 300, 301 treatment for paper, 112 — sections, 96 Coucher,210_ Cream, caustic, 244 Crotalaria juncea, 146 cellulose, 46 Crystal monoliydrates, 194 Cupr.immonium, 295 — solution, 9 preparing, 295 Cuticuliirisation, 77 Cutlu, 74 Cuto-cellulosos, 53, 72 — and caustic soda, 93 CutOiC, 74, 75 Cuttfir guillotine, 240 — Salmop's, 240 — single-sheet, 239 Cutting, beater knives, 171 — hemp, 103 — iute, 143 — knives, 237 — Manila, 143 — paper, 237 Cutting straw, 133 — wood for pulp, 147 DAMPiNG-cylinders, 225 Dandy-roll, 221 DeckL', 144, 217 De la Eue's dandy-roll, 222 Dextrin, 15 Dextrose, 15, 35 Diagnosis of plant subotances, 90 Diastase, 15 Digesters, 140 — cement linings, 141 — for wood pulp, 137 — lead linings, 140 Dimensions of fibres and filaments, 87 Dimethyl paraphenylene diamine, 85 Disintegration, chemical processes for, oF wood, 67 Doctor, 223 Draining bleached pulp, 162 — boiled rags, 117 Drawing paper, 211 Driving of beaters, 180 rope pulleys, 180 Drum washer for boiled rags, 118 Drying cylinders, 223 Dye stuffs, soluble, 203 Dveing pulps, 207 Dyes, 203 — analysing, 284 — coal tar, 42, 204, 206, 207 Economy of eleclrolytic bleaching, 163 EdjROYED i>xjivri>s (Double and Single Acting) For Pumping Out of Deep Boreholes. Sole Licensees under the following Patents: HARD WATER SOFTENING APPARATUS (" Archbutt-Deeley " System) For Town Supplies, Industrial Purposes^ etc. REEVES' PATENT FILTERS. PURIFICATION OF TRADE EFFLUENTS. (" Archbtjtt-Deeley " System). . THE "MATHER-EEYNOLDS" HIGH-LIFT TURBINE PUMP. ~ I Z [WiZZ<;M> Index, 5 ADVERTISEMENT^. ESTABLISHED 1808. JOSEPH PORRITT & SONS, HELMSHORE, Near Manchester. FELTS for Paper makers* SIEVES, BLANKETS AND LAPPINGS For Paper Stainers. BLANKETING for Letterpres Printers. And all kinds of WOOLLEN, LINEN AND COT'ON CLOTHS for Mechanical Purposes. FELTS ADVERTISEMENTS. 6 ANDREWS & CO. 91 QUEEH VICTORIA STREET, LONDON, B.C. WOOD PULP IMPORTERS Telegraphic Address : Telephone No. : « ABBORABIUS, LONDON." « 835, BANX.'> A. B C OOr>E XTSEID. 7 ADVEKTISEMENTS. HENDERSON, CRAIB &C0. XaXTSj!LX'T:EJD, 97 Queen Victoria Street, LONDON, E.G. AND AT 2 St. Andrew Square, EDINBURGH ; 52 Corporation Street, MANCHESTER. Supply at Lowest Market Prices All the Principal Articles used in the Manufactu re of Paper. SPECIALITY : WOOD PULPS CHEMICAL AND MECHANICAL Codes used i Telegrams: "CRAIG, LONDON." a B c, LIEBERS, a 1, - - ANGLO-AMERICAN. 8 SPONS' ENCYCLOPAEDIA OF THE Industrial Arts, Manufactures and Comniercial Products. EDITBD BY 0. G. WARNFORD LOCK, F.LS., Etc. Etc. In Super royal 8vo, containing 2100 Pages, and Illnstrated by nearly 1500 Engravings. Can be had in the following Bindings: In 2 Vols., cloth ^3 10 0 In 5 Divisions, cloth 3 11 6 In 2 Vols., half-morocco, top edge gilt, bound in a superior manner 4 10 O In 33 Monthly Parts, at 2s. each. Any part can be had separate, price 2s. ; postage 2d. Complete List of all the Subjects. TART PAIiT P4RT 14 Narcotics . . . .21 22 Elefair Manufactures . 18 27 Hats IS 28 Ice, Aitifitial . . . 18 Silk 28 Iiidiarubber Manufac- 28 tures 18, 19 So.ip, Railway Grease, and 19 Glycerine . , . .28, 29 Cailx>lic Acid . . . . 11 Jute Manufactures . . 19 29 Coal-tar Products . . . 11 Knitted Fabrics (Ho- 29 19 31 Cofl'ee . . . . . 11, 12 19 32 Cork . . 12 fjealher 19, 20 32 Cotton Manufactures . 12, K) Linen Manufactures . 20 32 20 32 Dyeing and Calico Prlnt- 21 Wool and Woollen Manu- ing . 13, U 21 facturts . . . .32, 33 London: E. & F. N. SPON, Ltd., 125 Strand. 2 A 9 ADVERTISEMENTS. WM. MAKIN & SONS, ATTERCLIFFE STEEL WORKS, SHEFFIELD. ». E c T B ' Established A.D. 1736 for the Manufacture of PAPER MAKERS' TOOLS. ROLL BARS, PLATES, & REFINER BLADES In Shear, Bar, Crucible. Siemens and Bessemer Steels, Bronze, and Yellow Composition Metals. Cutter Cl)opper=macl)ine Kniues. DOCTOR 5LADES. STEEL. FILES. HAMMERS. Illustrateu and Price Lists on application, London Office : 83 UPPER THAMES STREET, E.G. Telegrams— Nat. Telephone "MAKIN, SHEFFIELD." No. 782, ADVERTISEMENTS. 10 PAPER-MAKIWC MACHINERY, STEAM ENGINES, GEARING, ETC. ANGLE PAPER CUTTER. Suitable for all angles up to i5°, Exceiptionally massive design. Helical Bevel Wheels machine-moulded. Other Gears machine-cut. YERYf BEST RESULTS IN ACCURACY M DURABILITY GUARANTEED. JAiMES MILNE & SON, Ltd., MILTON HOUSE WORKS, EDINBURGH. WRIGHT & CO. «immm eng 157 8 LIMITED, ^ ww\ : ■ ■ 1 - ^ - « ^1.5^2-.*-. Date Due MA] ROl S THE Bl ET. COL Write 1 ■kc. ENGINl TY. Roller a ' . _ efficiently repaired, erected and maintained, at Reasonable Charges, by experienced Mechanics. HYDRAULIC AND OTHER LIFTS AND CRANES. 11 Hemmer Bros., Ltd NEIDENFELS (Pfalz), Paper Makers Engineers and Millwrights. R MILLS. MILLS, ical AS Patent I- Flat NGINES. lERS. PAPER L — ««...•. k-woMS and WIRE STRETCHING MACHINES. STEAM ENGINES. TURBINES. sni E AGENTS FOR THE UNITED KINGDOM: GETTY CENTER LIBRARY iry Payement, LONDON, E.C. 3 3125 00140 7234