FROM THR AUTHOR'S COLLECTION. COPYRIGHT 1895, BY HOWARD HART. ...THB... CHEMISTRY OF POTTERY BY SUPERINTENDENT OF THE MOSAIC TIL.E COMPANY, ZANESVH,I,E, O., FORMERLY SUPERINTENDENT OF ROOKWOOD POTTERY, CHEMIST OF THE AMERICAN ENCAUSTIC TILING COMPANY, ETC. KASTON, PA.: CHEMICAL PUBLISHING CO. 1895. DEDICATED TO MY ASSOCIATE IN CERAMIC WORK, HERMAN C. MUEUvER, SCULPTOR. CONTENTS. Page. Chapter I. Analysis of Pottery Materials and Products- i " II. Physical and Empirical Tests ^5 " III. Pyrometry 26 " IV. Classification of Ceramics 41 V. Pottery Glazes 48 VI. Red Ware 58 VII. Rockingham and Yellow Ware 66 " VIII. Stoneware 77 " IX. Raw Materials of White-Ware Bodies 93 " X. White Granite and Cream-Colored Ware 117 " XI. Majolica and Enameled Tile 127 " XII. White Enameled Brick 139 XIII. Floor-Tile and Terra-Cotta 149 " XIV. Refractory Materials 158 " XV. Burning the Ware 175 38S233 LIST OF ILLUSTRATIONS. Plate i. From the author's collection Frontispiece. " 2. Clay Bank", Muskingum Co., Ohio to face p. 97. Page. Initial i " - 15 Vicat's Needle 19 Gas Trial Kiln 23 Initial 26 A vSeger Cone in Natural Size 36 Initial 41 Tailpiece 47 Initial 48 Tailpiece 57 Initial 58 " 66 " 77 11 93 " JI 7 ;* 127 Tailpiece 138 Initial 139 " H9 " 158 11 175 Kiln Barometer or Draught Meter 178 PREFACE. The pottery industries of Kngland and America have afforded chemists little opportunity for systematic work, so that they have remained largely on an empirical basis and have supplied nothing of moment to chemical tech- nology. With the exception of the excellent treatises on porce- lain, but little work relating strictly to pottery has been published in France. The best of such work, accessible to chemists, is that of German ceramists and has appeared in that excellent technical periodical, "Die Thonindustrie-Zeitung," founded by the late Professor Seger, sometime chemist of the Royal Porcelain Factory at Charlottenburg, near Berlin. Most of the technical publications on pottery, in book form, dwell on the mechanical technology of the clay industries, the chemical parts consisting merely of re- ceipts, claiming a practical origin. Such receipts are almost altogether without value, as the materials specified are not characterized by accom- panying analyses and the temperatures to which the products are to be subjected are not given, or not determined in a way that they can be fixed with certainty. VI PREFACE. While to many of our chemists, engaged in develop- ing the natural resources of the country the clays found on every side have seemed to offer a fruitful field for investigation, the information at command concerning the chemical needs of the potter has been so meager, that their efforts have been practically abortive. The thousands of analyses published are mostly worthless, because they do not go far enough or because unaccom- panied by essential physical tests and practical trials. It is hoped that this little treatise will supply informa- tion that will turn future labors in this channel to good account. The writer is well aware that in confining the subject matter to the results of his personal experience the shortcomings of the treatise are numerous and manifest ; but he believes that as such it is a more direct expres- sion of the practical needs of the working potter, and therefore of more immediate value, than a compilation of the published work of European chemists on their ceramic industries. In a treatise of this kind, only the essentials, pottery bodies and glazes, could be considered. The colors and operations of decorating, though interesting objects of chemical study, would have led too far from the main purpose of the book. CHAPTER I. ANALYSIS OF POTTERY MATERIALS AND PRODUCTS. T IS most difficult to convey to the public the necessity of exercising the greatest care in taking the sample of a clay for exam- ination. Chemists and engineers will, of course, appre- ciate this, though they often do not realize that a carelessly taken sample is absolutely worth- less for even a preliminary test. The usefulness of a clay for a given purpose often hinges on a slight point of plasticity and shrinkage or color at a certain temperature, which may either have been imparted to it, or taken away from it, by a few years weathering at the exposed outcrop of a deposit, or by the infiltration, or washing out, at that point, of small amounts of iron, lime, or alkalis. The first rule in clay examinations, then, is that a clay deposit which does not warrant the expense of thorough sampling does not justify even a preliminary examination, for we have not, as in iron-ore, for exam- ple, a constituent which must be contained in at least a certain amount, and which any exposure is not likely to * CtlEMISTRY OF POTTERY. materially vary, nor are there constituents which must be practically absent. Every constituent of a clay is valuable. It is the work of the chemist to determine to what use it can be put ; and, considering its constituents, to what conditions of manufacture it must be subjected. His first question, then, is, whether there be enough like the sample to warrant difficult tests. If the clay be not too hard, and superimposed strata too difficult to work, it is best to take the samples over a larger area by systematic borings, using a one and one- half to two-inch pod-auger, welded to a section of gas- pipe for the purpose. If the overlying strata are sandy or otherwise liable, by dropping down into the bore-hole, to vitiate the char- acter of the sample, it becomes necessary to exclude them by driving down a section of tubing of suitable width to the top of the clay stratum which is to be sam- pled. If sampling by boring be not practicable, the deposit must be opened at as many exposures as is practicable and laid bare, if possible, through its entire depth, well beyond the frost line or other surface influence, and the sample there taken down the whole exposed face of the vein. The sampling down of the clay obtained by these various means is done in the manner familiar to all accus- tomed to handling ores, and should be done in the labor- atory, if the man in the field is not perfectly familiar with the process. ANALYSIS OF MATERIALS AND PRODUCTS. 3 In preparing a sample of clay for chemical analysis, it is important to bear in mind what constituents, separa- ble by mechanical means, the clay contains which are objectionable in the industry in which it is likely to be used, and to reject, by a suitable treatment, from the analytical sample, those constituents that will not enter into the ultimate product, though it is, of course, neces- sary to at least approximately determine their character and amount, as this determines the future mechanical purification process to which such a clay, if used, would have to be subjected. Thus "lime-dogs" and pebbles are objectionable in terra-cotta, brick and red ware clays, and are custom- arily removed, if present, by screening the dried and crushed clay through a sieve of ten to twenty meshes to the inch. Coarse sand and particles of iron-pyrites are objection- able in finer terra-cotta, floor tile and yellow ware, and are removed by disintegrating the clay in water, passing the " slip" through a sieve of sixty meshes to the inch, and drying the purified clay. In clays for finer ware, flakes of mica, in addition to the last mentioned impurities, are objectionable, and must be removed by passing the "slip" through a looto 1 20 mesh sieve. It may seem superfluous to say that the chemical anal- ysis of a clay should be as accurate as possible ; yet the very considerable number of slovenly analyses published yearly would seem to make it necessary to insist upon this point. 4 THK CHEMISTRY OF POTTKRY. By far the greater number of technical analysts are engaged in making single determinations, and not com- plete analyses of the substances with which they have to deal. Clays, in which all the constituents are of equal importance, seem, therefore, to many as difficult as they certainly are tedious, to analyze, and the determinations of their various constituents are treated as if they were a collection of individual determinations, the incorrect footing up of which is an "arbitrary impertinence," which may be ignored. While it is true that the chemical analysis alone, how- ever accurate, is insufficient without accompanying phys- ical tests, to give one a perfect characterization of a clay, its value is sufficiently great to warrant the most careful work. Although the scope of this book makes it necessary to refer to works on analytical chemistry for the details of clay analysis, it seems desirable to point to special con- siderations upon which text-books do not sufficiently insist. The character of a clay depends largely upon the mutual proportions of alumina and silica, so that, although these are generally the largest constituents of a clay, it is important that they be determined with par- ticular accuracy. In spite of the various and detailed descriptions of the treatment of the residue of the acidified fusion of a clay with alkali carbonates, it seems practically impossible to accurately separate silica from the alumina group, and to obtain both without mutual contamination. The wri- ter, therefore, thinks it indispensable to finally obtain the ANALYSIS OF MATERIALS AND PRODUCTS. 5 proportion of silica by difference, smoking off that obtained from the fusion of the clay with alkali carbon- ates, with hydrofluoric acid, taking cognizance of the residue obtained from this operation. The alumina and ferric oxid precipitate must simi- larly be obtained by difference, when the silica, which it almost invariably contains, has been separated by dis- solving the weighed precipitate in Mitscherlich's mix- ture, or, if very refractory, by fusion in acid potassium sulfate, the silica, after weighing and volatilizing as fluorid, being, of course, added to the main portion. Occasionally, the errors due to alumina in the silica and silica in the alumina group mutually compensate each other within the limits of ordinary analytical accu- racy. It is, however, a dangerous practice to depend upon such a balancing of errors, It must further be borne in mind that the common impurities of analytical reagents are the normal constit- uents of clays, and may frequently throw out an accu- rately manipulated analysis several per cent. Hydrochloric acid and ammonia frequently contain very appreciable traces of alumina and soluble silica. Precipitated calcium carbonate (if ^awrence Smith's alkali determination be used) cannot be bought free from alkali, and dry sodium carbonate contains, almost invariably, traces of silica, and has not infrequently been found by the writer to contain fine splinters of parian, from the " porcelain" linings, possibly, of the mills in which it was ground, whereby silica and alumina would both be introduced into the analysis. 6 THE CHEMISTRY OP POTTERY. Distilled water is also a very fruitful source of diffi- culty ; for in washing the bulky precipitates occurring in clay analyses, traces of impurity rapidly heap up to important factors. Trouble has been found with water which, it was supposed, had been distilled with great care and condensed in a block-tin worm that could not be sus- pected. Still the latter proved to be the source of the difficulty. The upper coil, into which the hot steam enters, had, in the course of two years, become highly crystalline, and was riddled with fine cracks. Into the latter, the cooling water had seeped and vitiated the dis- tillate. Such a defect in the condensing worm may readily be detected by drawing the cooling water from the tub, and now blowing the steam from the still into the coil. Pres- ently the mud and scale deposited on the latter are dried through the heating of the worm. If now little spots of the same remain persistently damp, it becomes apparent that underneath are cracks in the worm sufficient to let the steam through, and sufficient, also, when the tub is filled, to give the cooling water access to the interior of the worm. In the case of clays containing a larger amount of alkali, if potash and soda be not separated, but the titer of the chlorine found in the weighed chlorids, and the equivalent oxids calculated, it is important to give the combining weight of the alkali obtained, so that the chemical formula of the clay may be calculated, in case it should be desired to use it as or in a glaze. Beside the ultimate analysis, a proximate one is very ANALYSIS OF MATERIALS AND PRODUCTS. 7 important. Considering the formation of clays, it will be appreciated that its various ultimate constituents are grouped in a variety of mixed minerals of widely differ- ing physical properties, and that these, rather than the former, determine the character of the clay. An accurate separation of the clay into its various com- ponent minerals is, in the present state of analytical knowledge, out of the question; yet, the well-known separation of mineral and quartz sand, by digesting the clay in strong sulfuric acid, and after washing out the excess of acid and the sulf ates formed by the decomposi- tion of the clay substance proper and the micaceous min- erals, and removal of the separated soluble silica with a solution of alkali carbonate, described in text-books on analytical chemistry, gives data of great practical value. When the sand thus found in a clay is not pure quartz, but, as is generally the case, consists of the detrital mat- ter of a great variety of minerals, it becomes very diffi- cult to estimate its probable effect in its influence on the refractory qualities of a clay, and the resulting coefficient of expansion of the burned body. For minerals, more or less fusible, greatly reduce the refractoriness of a clay, while quartz, at moderate temperatures (though not at high ones) , increases it ; and with reference to the coeffi- cient of expansion, the effect of these substances is also diametrically opposite. As now, there is no way of determining in a mixture of quartz and mineral sand how much of the silica belongs to the minerals and how much to the quartz itself, 8 THE CHEMISTRY OF POTTERY. Seger and Aron 1 have suggested that, inasmuch as the mineral detritus of the sand is more or less fusible, and plays the same role in a pottery body as powdered feld- spar, a substance with the action of which, in a clay ,* the potter is perfectly familiar; and as, further, such min- eral sand is often mainly feldspathic, it is, for all practi- cal purposes, sufficient to consider it feldspar. Hence, the alumina obtained on analyzing the sand is used as the factor with which the silica to be assigned to the mineral sand is determined, its weight being multi- plied by the factor expressing the proportion of silica to alumina in feldspar, and this amount of silica being deducted from the total silica of the sand, the remaining silica is called quartz, while that deducted, together with the alumina and the alkalis and alkaline earths found in the sand, are added, and designated ^feldspathic" sand or detritus. The difference between the weights of the constituents found in the complete analysis of the clay and those of the sand, left by the sulfuric acid and alkali carbonate treatment, are calculated as constituting the " clay sub- stance." In order that the proportion of the latter may be more satisfactorily surveyed and compared with those of kaolinite, it is necessary to reduce them to a basis of 100. The ' ' rational analysis, ' ' then, looks upon every clay as consisting of a "clay substance" deviating more or less from kaolinite, in which the alumina is substituted, iNotizblatt des T6pfer-u. Ziegler-vereins, 1874, S. 226, and Zwick, Jahres- bericht, 1878, S. 21, and 1879, S. 41. ANALYSIS OF MATERIALS AND PRODUCTS. 9 to a greater or less degree, by ferric oxid, etc., and the combined, water by alkalis and alkaline earths ; quartz, in a finely divided state, and mineral sand, prac- tically feldspathic. Scientifically, it must be conceded Dr. C. Bischof that the result is ' ' ein kiinstlich theoretisches Bild, ' ' ! both on account of the assumption underlying the calculation of the mineral sand as feldspar, and because the method of separating the mineral and quartz sand from the" clay substance is by no means an accurate analytical process. Thus a pure feldspar (orthoclase) treated in the same manner as the clays are treated, 2 left in one instance a residue of 81.44 percent., and in another 83.95 per cent. In the hope that clays might be decomposed with dilu- ted sulfuric acid under pressure under conditions that would cause less action on feldspathic minerals, it was found that two grams of kaolin treated with twenty cubic centimeters of five times normal sulfuric acid in a sealed tube of hard glass, was only completely decomposed in two hours, at a temperature of 200 C., conditions that were insufficient for decomposing some plastic clays. Yet a feldspar subjected to the same treatment, left after sep- aration of the acid and washing out the soluble silica with solution of sodium carbonate, a residue of only 82.67 P er cent., practically the same as the regular treat- ment. Vfc It is hardly possible then, with existing methods, to 1 Bockmann : Untersuchungs methoden, S. 354, 1884. 2 Fresenius : Anleitung zur Quantitative!! Chemischen Analyse, Sechste Auflage, ii, 352, f. 10 THE CHEMISTRY OF POTTERY. dissolve the "clay substance" out of a clay without a loss of feldspar that may reach twenty per cent, of the amount present. The treatment with sulfuric acid even affects the quartz to some extent and leads to its removal by the subse- quent sodium carbonate treatment. Thus a pure finely powdered quartz was found to con- tain 3.75 per cent* silica soluble in sodium carbonate solu- tion. Subjected after removal of the same to the treat- ment used for solution of "clay substance," it lost 3.88 per cent, apparently rendered soluble by the action of the sulfuric acid. Practically, this division of the analytical data of clays is of the greatest service to the potter, who is familiar with "clay, flint, and spar," and knows how to vary their proportions for his various ends. It has resolved the composition of natural clays into components with which he knows how to deal, and made once unintelli- gible analyses serviceable to him. The following may serve as an example of the analy- sis of a clay, conducted on these lines. It is a kaolin from Nelson County, Virginia. The entire clay. The contained sand. Silica 50.02 12.62 Alumina 35- I 8 i-72 Ferric oxid 0.36 0.09 Lime 0.12 0.06 Magnesia 0.07 * 0.02 Alkalis 3-39 1 i.o8 2 Combined water 10.57 ' 99-71 15.59 1 Combining weight 43. 2 Combining weight 34.2. ANALYSIS OF MATERIALS AND PRODUCTS. II Rational Analysis. Clay substance ^. 84.12 Feldspathic detritus *. 9.04 Quartz 6.55 99.71 PERCENTAGE COMPOSITION OF Clay substance. Feldspathic detritus. Silica 44.47 67.15 Alumina 39-79 i9-3 Ferric oxid 0.32 i.oo Ivime 0.07 0.66 Magnesia 0.02 0.22 Alkalis 2.75 H-95 Combined water 12.58 o.oo 100.00 100.00 Should the clay contain limestone detritus and appre- ciable amounts of uncombined ferric oxid or alumina, it is proper to extract these with dilute hydrochloric acid, and place them in the rational analysis by themselves. Similarly, if the clay contain soluble silica, this should be extracted, and added to the quartz. When this is done, as before pointed out, the per- centage formula of the "clay substance," as a rule, closely approximates kaolinite^the variation found from the latter being in the partial substitution of alumina by ferric oxid and of the combined water by alkalis and alkaline earths. These variations, however, the potter can easily learn to allow for, by the now simple compar- ison with the percentage composition of kaolinite. It will occur to the chemist, that as the detritus of micas must be a constant constituent of clays, and as the dust of these minerals is practically all decomposable by sulfuric acid, and as their presence in the "clay sub- 12 THE CHEMISTRY OF POTTKRY. stance" will vary its percentage composition just as described above, it would seem both justifiable and ser- viceable to assume that the alkalis and alkaline earths of the ' ' clay substance' ' are contained in such combination, and using the average percentage of alkali in muscovite as a measure, deduct from each constituent a corres- ponding amount as going to make up so much mica. The "rational analysis" of the above clay would, under this consideration, be Clay substance 60.23 P er cent. Mica 23.89 " Feldspar 9.04 ' ' Quartz 6.55 " Feldspar. 67.15 19.03 1. 00 0.66 0.22 H-95 99.71 the percentage composition being Clay substance. Mica. Feldspar. Quartz. Silica 43-75 46.30 67.15 loo.oo Alumina 39.64 4o-*7 Ferric oxid 1.13 Lime 0.25 Magnesia 0.08 Alkalis 9.67 Combined water 16.61 2.40 100.00 100.00 100.00 100.00 This extension of the "rational analysis" introduces a second hypothetical factor, which may often be justified in a primary clay where mineralogical examination con- firms such a division, but beyond this, even the practical needs of the potter do not demand it, as the micas are not used as fluxes in ceramic industries, and their hypo- thetical presence would convey less meaning to him than the percentage composition of a " clay substance" easily comparable with kaolinite. ANALYSIS OF MATERIALS AND PRODUCTS. 13 Clay analyses should always be calculated to the basis of the sample, dried at 120 C., as analyses showing vary- ing amounts of moisture are not readily comparable. Upon the analyses of the other minerals, oxids, and salts used by the potter, it is not necessary to dwell. Especial impurities, for which it is alone necessary, often, to make examination, will be mentioned when the substances * themselves are discussed. Of ceramic products, the analysis of the pottery bodies is conducted similarly to that of a clay, care being taken in the preparation of the sample that every trace of glaze, etc., be chiseled or ground off. The calculation of the composition of a mass or paste for producing the same from known clays and minerals is readily done on well- known stoichiometric rules. A practical burning trial will then show how the mass obtained from the analysis will have to be modified to allow for the individual phys- ical characteristics of the clays used. The obtaining of samples of enamels and glazes from finished pieces of ware is often attended with considerable difficulty ; inasmuch as these are frequently very thin, or the body of the ware is so soft that it chips off with them, contaminating the sample. The most practical way of obtaining them is to bed the piece from which the sample is to be obtained in clay, on a convenient table, surrounding the piece with large sheets of glazed paper, to catch the splinters of the glass as they are struck off. The enamel or glaze is now dressed off with small chisels of hardened steel, driven with a light hammer. As the chisels soon dull, a sufficient number 14 THE CHEMISTRY OF POTTERY. of them must be kept on hand to carry on the work. The blows of the hammer must be so regulated that the chisel does not cut into the body of the ware. When sufficient of the sample has been dressed off, it is carefully swept together with a camel's hair pencil, and probed with a bright, clean magnet until all the iron introduced by the chisels is extracted. It is then ground fine in the agate mortar for analysis. From the usual components of potters' glazes, given in a later chapter, the chemist will understand what class of substances it is necessary to look for, and what pre- cautions must be taken in the analyses. The almost con- stant presence of reducible metals would seem to make these analyses difficult, on account of the unavoidable fusions in platinum ; but by unlocking the glasses with a liberal amount of alkali carbonate, and keeping the crucible well up in the flame, no reduction need be feared. The frequent presence of considerable amounts of bo- racic acid in these silicates formerly presented an almost unsurmountable difficulty to accurate analysis, but Gooch's method of determining this substance 1 has over- come this trouble ; yet we have still to deal with the most difficult phase of the determination, for the unlocking of boro-silicates containing frequently a very high per- centage of reducible metal, compels fusion with large amounts of alkali, and the following acidification fills the retort with a great mass of troublesome salts. 1 Bulletin of the U. S. Geological Survey, No. 43, p. 64. See also J. Anal. Appl. Chem., 2, 86. CHAPTER II. PHYSICAL AND EMPIRICAL TESTS. HE physical properties of clays play so im- portant a part in their use and in deter- mining the character of the products which can be made from them, that the chemical analysis, by itself, is altogether insufficient to tell what qualities may be expected from such materials. Kven in the matter of the color to be obtained on burning the clay, which would seem to hang most closely together with its chemical composition, the analysis leaves us, except in extreme cases and in a general way, in doubt. The coloring of clays in the fire is due, primarily, to the presence of iron oxid, the tint being modified by the amount of lime occurring and influenced, occasionally by the presence of manganese, as also by the chemical quality of the flame during burning. But the depth of color of a burned clay is not at all. in proportion to the contained iron oxid ; a secondary clay may contain considerably less iron than a kaolin and yet burn quite yellow, while the latter burns snowy white. All depends upon the combinations in which the iron is held, and as long as we are unable to accurately separate the minerals of a clay, and subject them individually to examination, the chemical determination of the gross amounts of the l6 THE CHEMISTRY OF POTTERY. coloring oxids will give us but a very rough idea of the tints we may expect. To formulate a systematic set of physical and empir- ical tests to characterize a clay is, on account of the widely differing needs of clay-workers, a very difficult thing. Yet obvious operations connected with its general use must be undertaken, and the conditions of the trials and their resulte must be described, as far as possible, in terms of physical and chemical measurement at com- mon command. If this be done, inferences can be drawn from the experiments sufficiently close to the behavior of the clay under working conditions, to reliably forecast its serviceability or worthlessness for this or that purpose. In this view, at least the following tests should be made: 1. The fineness of grain of the constituents of the clay, measured by passing it dry or suspended in water through sieves of different mesh, or washing it with water at dif- ferent measurable speeds of the wash-water, recording the proportion of each grade of fineness measured. 2. The plasticity, expressed numerically, by some measurable quality connected with it. 3. The binding property, given as tensile strength. 4. A firing under definite conditions of the chemical character of the flame, with record of the duration of the fire and the temperature reached, measured in a satisfac- tory pyrometric standard, to attain a specific hardness. 5. The porosity of the product burned as given, deter- mined by its water-absorption. 6. The shrinkage from the clay state in which the material is formed to its burned condition. PHYSICAL AND EMPIRICAL TESTS. 17 7. Its coefficient of expansion, when burned as described, measured empirically by melting on it, at a heat lower than that of its original baking, a glaze of definite chemical composition, noting if in time, the glaze alone crack, a phenomenon called "crazing," or if it shatter the body or fly off at the edges, tearing the clay along with it, a phenomenon opposite to t;hat of ' ' crazing' ' and called "shivering," which result from either a too great or too small coefficient of expansion of the clay as compared with that of the glaze. It is of course, quite important, that the sample for these physical and empirical tests be properly averaged, and in fact the same from which the portioia for the chemical analysis is taken. Much was hoped of the mechanical analysis of clays, 1 by clay- workers, in the direction of a serviceable separa- tion of the material into its constituent minerals by their hydraulic values, as a preparation of the sample for chemical analysis and many separations were made with the apparatus of Schone, 2 modified by Schiitze. 3 Disappointment at the results, in the light of the mis- applied and over-great expectations has reduced this branch of operations to mere sieve-analysis and prepara- tion of the sample for the analytical and physical tests of the clay, as the needs of the particular industry, in which it is likely to be used would indicate. 1 For an excellent discussion of the principles and methods see Wiley Principles and Practice of Agricultural Analysis, Vol. i, Part Fourth, p. 171. 2 Zeitschrift fiir analytische Chemie, 7, 29. SNotizblatt fiir Fabrikation von Ziegeln, etc., 1872, 88. 1 8 THE CHEMISTRY OF POTTERY. The appearance, geological origin or a preliminary burning trial, giving the general use to which a clay may be put, the sample is prepared by passing it dry or more commonly by washing it through such a sieve as the mechanical preparation of the clay for the industry would demand. The material remaining on the sieve used should be weighed and described. That passing through is dried and used for the various tests. There is not, as yet, an altogether satisfactory meas- ure of the plasticity of clays ; though the possibility of giving direct numerical expression to this subtle and most valuable property would be of great practical value. The best determination that can thus far be made is based on the observation that, in the main, the greater the plasticity of a clay the larger the amount of water required to bring it to a definite degree of softness, at which it can be worked. In order to determine this point, apparatus have been devised for pushing, with a fixed load, a wire, rod or thin-walled cylinder to a certain depth and within a cer- tain time, into the softened clay. The proportion of water required to soften one hun- dred parts of the dry clay to the requisite emollescence is taken as the direct measure of the plasticity. The apparatus necessary to determine, if the requisite softness of the clay is attained, which is most conve- nient, inasmuch as it is already in use for determining the time of setting of Portland cement, is Vicat's needle. 1 1 Transactions of the American Society of Civil Engineers, 1893. Max Gary : The Testing of Portland Cement. PHYSICAL AND EMPIRICAL TESTS. IQ It has a wire of circular cross-section, the end cut at right angles to its axis, with an area of one square mil- limeter. The wire is over four centimeters long and attached to the end of a rod, suitably guided, which weighs 300 grams. The softened clay is well ' ' wedged' ' to make it per- fectly homogeneous, and pressed into a ring four centi- meters deep, set on a glass plate under the needle and struck off level with a steel spatula. The needle is then VICAT'S NEEDLE. allowed to penetrate the clay, and if within five minutes it sinks to a depth of four centimeters into the same, the clay is of the proper consistence ; if not, it is either made stiff er or softened, as the case may be. 20 THE CHEMISTRY OF POTTERY. A weighed sample of the mass of proper consistence is dried and the proportion of water to 100 parts of dry clay is calculated, the figure being used as the direct index of the plasticity. The binding power of clays is determined by tearing well-dried specimens formed in cement-briquette molds, the results being expressed in the weight in grams per square centimeter sufficient to break the clay. Any of the standard machines for testing the tensile strength of cement 1 will answer the purpose. It is necessary, however, to form the test briquettes very carefully, as they are liable to contain flaws, par- ticularly with very plastic clays, which make it difficult to get concordant data. The clay must be well "wedged" and beaten out into a slab rather thicker than the mold. By means of a bent tin stamp a piece of such dimensions, that it will very nearly fit the mold, is cut out, into which, after well oiling the latter and setting it on a dry plaster of Paris slab, it is beaten, and the excess of clay cut off with a thin wire. The mold is then slipped off the bri- quette, care being taken not to disturb its shape. After the air-dried pieces have been broken in the machine, the surfaces of fracture must be measured before calculating the breaking load, that due allowance be given for the shrinkage of the clay in drying. The binding power of clays is not necessarily propor- tional to their plasticity, as one would naturally sup- i Journal of the American Chemical Society, 16, 161, 1894. Thos. B. Still- man : The Chemical and Physical Examination of Portland Cement. PHYSICAL AND KMPIRICAI, TESTS. 21 pose. The New Jersey ball clay describe4 in a subse- quent chapter is in striking illustration of this fact. Clays which require a large amount of water to render them workable, and therefore show considerable shrink- age in the clay state, on drying, but which when dry are of low tensile strength, are liable to be very troublesome, particularly in the making of heavy and of dust-pressed wares by ''checking," that is, showing surface cracks or "dunting" (cracking through). For drying, taking place from the surface of the piece, the material must be able to stand the strain of its shrinkage on a more slowly contracting center. Where clays are deficient in this particular, great pre- cautions have to be taken to prevent the loss of ware made with them in rapid drying, which often involves so much cost and care that the use of the material becomes out of the question. Although half a dozen disks of the properly sampled clay, of the diameter and twice the thickness of a silver quarter, would answer to determine all that it is desired to know of the firing of a clay, this test cannot be made over a Bunsen or blast-flame, with the pieces packed in a platinum crucible, even with the help of the Brdman furnace ; the difficulty being that with the sim- ple appliances of the ordinary laboratory, it is not possi- ble to get a sufficiently large zone in which the chemical character of the flame remains the same through a suffi- cient length of time, and in which the temperature is likewise the same and can be observed or measured with- out altering the conditions. 22 THE CHEMISTRY OF POTTERY. Hence it becomes necessary to make use of larger pieces of apparatus, which in the greater amount of time required in their heating, give a better opportunity, in the admission of fuel and regulation of the draft, to get the quality of flame desired, and further, by their slower and more regular advance in temperature, give the clay pieces, which are, in themselves, very poor conductors of heat, the opportunity of progressing in their entirety through the changes caused by advancing temperature, without a shrinkage of the periphery on the centers, causing distortion and unequal tension. The ordinary assayer's muffle will answer the purpose of firing very well, whether it be heated with coke, gas, or gasoline ; but the draught used should be the natural one of a good chimney, regulable by a damper and not the forced draught of a fan or bellows, as the quality of the flame is, in the latter case, too uncertain and diffi- cult to regulate. Where gas is available, the most convenient furnace is one made after designs of Seger, by Geith, in Coburg, Germany, consisting of a ring of eight Bunsen burners, the flames playing into a fire-clay furnace and forced by the connection with the chimney in the bottom, to play over a ring and down on the crucible or ' ' saggar' ' before passing into the flue. In this way, by the up and down passage of the flame, a large zone of uniform tempera- ture is well attained. The arrangement will be well understood from the sketch of a similar contrivance which can easily be extemporized with good fire-brick and tile by a practical brick -layer. PHYSICAL AND EMPIRICAL TESTS. For attaining high heats, it is, of course, imperative that the gas pressure be sufficient, a con- dition often only to' be had in the even- ing, as most of our gas companies are conducted. The temperature at which different clays should be fired and the melting of glazes on the fired pieces, to empirical- ly test their coeffi- cients of expansion, can best be explained^ in connection with the considerations applying to their use , and is deferred to the respective chapters on the different wares as the results of ex- perimental burning are only of value when looking to the conditions of definite products, and when the heats to which the trials have been subjected are measurable and can be reproduced and controlled, to a similar degree, in the kilns of the potter. The subject of pyrometry be- comes one of the greatest importance in this connection ; to it, the subsequent chapter will be devoted. Scale, fc to 1 GAS TRIAL KILN. 24 THE CHEMISTRY OF POTTERY. In the case of clays intended for making floor-tile and paving-brick, and for ware which is to be partly glazed and exposed to the influences of the weather, it is im- portant to know how porous it still is after receiving the proper fire. It is sufficient, for this purpose, to deter- mine the weight of water that the burned specimen will absorb. In making the test it is important not to immerse the en- tire piece in water, as this would seal the pores and make it very difficult to get rid of the air in the interior. One face of the specimen must be left dry, and the water al- lowed to rise by capillarity, the piece not being lowered under the surface of the water until all its air has been expelled. It is superficially wiped off before weighing. As all clays shrink in the fire, but show the greatest variation in this property, and as articles made from them are generally required to come to definite sizes, hollow- ware being made to certain capacities, sanitary- ware having to fit metallic fittings, and brick and tile being required of standard sizes, a description of the physical properties of a clay involves a statement of the amount of its shrinkage from the plastic condition in which it is worked to the condition reached at the temperature in which it is burned ; so that from it the artisan will be enabled to calculate the sizes of his molds or dies. If, for the burning trials, disks of about one and one- half inches in diameter and three-sixteenths of an inch thick, be beaten out from the plastic clay on clean bats of plaster of Paris, the surface of the disks being smoothed with a knife or steel spatula and two dots impressed or PHYSICAL AND EMPIRICAL TESTS. 25 lines drawn into the soft clay, just twenty-five millimeters apart, the shrinkage of the clay is found with sufficient accuracy after the trials have been burned, as in the sec- ond measurement, quarter millimeters are easily estima- ted so that the result may be given in per cents. Articles of which the greatest precision in size is re- quired, flooring and wall-tile and fine pressed brick, are not formed from clay in the plastic state, but from fine clay-dust containing from eight to twelve per cent, mois- ture, so that it just packs when squeezed between the fingers. The shrinkage of the dust-pressed and plastic-pressed clay will vary somewhat, and it is useful to make for the burning a few dust-pressed disks, beside the plastic ones. This is easily done in a large diamond mortar, twenty-five millimeters in diameter. It must be well cleaned and oiled, the ring filled one-third to one-half full of the slightly dampened clay-dust, the pestle insert- ed and pushed down, then struck twice gently with the hammer in order to pack the clay, but still let out the air, and finally struck two sharp blows with the hammer to compress and seal the disk. It is then carefully pushed out, and after being dried and fired to the established heat, is measured to quarter millimeters with a pair of calipers and the shrinkage recorded in per cents. CHAPTER III. PYROMETRY. HE subject of pyrometry is one of the most vital interest to an industry making use of the greatest range of high tem- peratures, and risking large quantities of ware to an operation which will make or mar it, unless the heat desired be attained within a very narrow range of variation. Potters use, for their guidance in determining if the requisite high temperatures have been reached, empiric trial-pieces made of materials used in their manufacture, the behavior of which in the fire they have become familiar with in the course of their practical experience. Such trial-pieces are usually rings of clay or shards of baked clay coated with glaze mixtures, fusible clay, or feldspar. Usually these empiric standards serve their purpose very w r ell in the hands of experienced men, though they fail at times, even under such use, through mechanical and chemical changes of the materials, which nothing but careful analyses would detect and which are usually not made. It will be appreciated that the personal factor is a very great one in the judging of such empiric pyrometric trials, PYROMKTRY. 27 and it is this that the burner of better classes of ware, par- ticularly, turns to his account, in making himself indis- pensable to a manufacturer, who is loth to entrust the dangerous operation of burning to one who has first to make his experience with the trials. But even in the hands of the experienced burner, they have the disadvantage that they do not enable him, ex- cept in a very imperfect way, to strike other tempera- tures with which he has not previously worked. He is therefore helpless under all but the limited conditions of his immediate experience. As the trials of different establishments bear no rela- tion to each other, thejr experiences in firing are not comparable, and they cannot be mutually helpful, even when there is an entirely cordial disposition to be so. As the first principles of systematic chemical investi- gation involve the use of standards that are comparable and easily attained, it would be useless to go into the present practical kiln trials of our potters beyond men- tioning them in connection with the different classes of manufactures, which will be done in the appropriate places. It is necessary, however, to carefully consider the con- ditions imposed upon pyrometric observations in pottery kilns, in order to select the best means, of those pro- posed, for making them. The first and most obvious condition is that the pyrom- eter should be of such positive action that there enter no personal lactor into its reading. That while it be sufficiently wide in range to cover all temperatures of 28 THE CHKMISTRY OF POTTKRY. ceramic operation, the divisions must be so close that the progress of physical change of clays and glazes made in the fire be not beyond the observable progress of the pyrometer. These obvious conditions rule out all optical pyrome- ters which have thus far been made. Furthermore, kiln-men, even if they be teachable, have to do rough work, and cannot be expected to manipulate delicate instruments with hands roughened and sense of touch blunted by coal-shoveling. Pottery kilns are large affairs, having commonly an internal diameter of sixteen feet and a height of fourteen feet. The heat, proceeding from a number of fires, can only be made to progress uniformly, by observation at a number of points, which are seldom less than four, and generally about eight. These points must be average ones, as far as temperature conditions are concerned, and can hardly be nearer than from four to six feet to the outer circumference of the kiln. Pyrometers whose in- dications niUvSt be transmitted through this distance of widely varying temperatures in order to be read on the outside are difficult to place, and their readings become complicated with correcting factors that are hard to fix. If the expense of the appliance be such that one can- not be left in each place of observation, but one or two instruments must be used to make all the measurements of perhaps several kilns; the length of time consumed in the work will be such that at the critical time all neces- sary observations and regulation of the fire by them can- not be made with sufficient expedition ; for upon open- PYROMETRY. 29 ing the kiln for the introduction of the instrument, the invariable suction of a strong draught of cold air to the point where it is to be placed results and so far cools that locality that some time must elapse, after all has again been made tight, before a reading can be taken. Another consideration, inherent in almost all condi- tions where pyrometric determinations become desirable, is very frequently lost sight of ; namely, that heat is re- quired to do work in a body which, through its mass or poor conductivity, responds but slowly to the impact of the flashes of the former. Through this comparatively slow response, the body averages within itself the effect of heat waves of widely differing temperatures which it is constantly receiving. Now, a pyrometer which is not similarly capable of averaging the heat impressions received will be too sen- sitive ; many readings will have to be made in order to get the average, or these will have to be recorded as a curve, by a suitable instrument. The former occupies too much time ; the latter involves the exposure of a delicate mechanism to places and conditions that are very destructive to it. One fact has done much to retard real progress in pyrometry ; namely, the fundamental association, in most minds, of the thermometric degrees of the mercury thermometer as something inseparable from all tempera- ture measurements. The great utility of using the expansion of liquids, where the range of temperature makes it admissible, accounts for this very natural error. But even in cer- 3O THE CHEMISTRY OF POTTERY. amics, where, perhaps, the greatest range of heat requir- ing exact control is used, not self-regulable as is, to a certain extent, the case in many chemical and metallur- gical processes, there is no occasion to pass from the thermometric scale to higher heats by a regularly graduated system, and the attempts to extend, by calcu- lation and interpolation by over 2,000, a system which gives data for less than 400, have awakened false and unattainable aims. The connecting of pjTometric with the common ther- mometric system can, and, in any case, must be done by the calorimeter ; but the former need not, therefore, and should not be divided by thermometric degrees. The difficulties of pyrometry and the varying condi- tions under which it must be executed may make instru- ments based on a variety of manifestations of physical change, varying positively under increasing heat neces- sary. The expansion of gases, the decreasing electrical conductivity of metallic circuits, the melting down of pyro- scopes of progressive variation in chemical composition, may all be successfully employed for the purpose in dif- ferent places, but only the latter have maintained the promise of permanent success under the difficult condi- tions of ceramic pyrometry, as enumerated. The first of such melting pyroscopes that gave good service and can still be employed under proper precau- tions and in certain range, are the metals silver and gold and their alloys, and the alloys of gold with platinum, as suggested by Prinsep. 1 1 Philosophical Transactions, 1828, p 79. PYROMETRY. 31 These are made by weighing off accurately the rela- tive proportions of the pure metals, in the form of wire, and then combining the portions by thorough fusion before the blowpipe. The alloy beads are hammered to flat disks, the melting of which is easily observed by their collapsing and drawing into the bead-form. The alloys usually used are the following : their melting- points were determined by Erhard and Schertel. 1 Percentages. Melting- Percentages. Melting- Gold. Silver. point. Gold. Platinum. point, oo loo 954C. 95 5 nooC. 20 80 975 90 10 1130 4o 60 995 85 15 1160 60 40 1020 80 20 1190 80 20 1045 100 oo 1075 On account of the difficulty, however, of obtaining amounts of alloys of absolutely uniform composition, these can only be used in small plates one or two deci- grams in weight, which are not observable at any dis- tance, particularly not in luminous surroundings, and require their extraction from the places of exposure for observation Furthermore, on account of their becom- ing crystalline in a very slowly rising heat, and the leaching out from the crystalline mass of drops virtually different in composition and melting-point from those of the entire alloy, it is not possible to leave a number of them in the kiln and withdraw them one by one with the rise in heat, for purposes of observation, but they must be introduced when the heat is suspected to be near the 1 Jahrbuch fiir das Berg-und Huttenwesen in Sachsen, 32 THE CHEMISTRY OF POTTERY. point wished, remain until they have the full heat of the surroundings, and then be withdrawn for observation. This makes too much manipulation for practical work, and the small alloy plates are frequently lost, entailing, on account of their cost, considerable expense. The difficulty of the changing physical character of the gold platinum alloys is particularly troublesome when the latter metal amounts to twenty per cent, and over, so that considerable uncertainty attaches to the tests at higher temperatures. The advantages of the Prinsep alloys as bodies prac- tically uninfluenced by oxidizing or reducing atmos- pheres, melting down at regular sufficiently close inter- vals, suggested to Dr. Heintz 1 the adoption of glass mix- tures on the type of porcelain glazes, for the same pur- pose. Their cheapness, the possibility of forming them into masses easily observed without withdrawal from the place of exposure, rendered them suitable for the pur- pose. The chief problem became the making of as great a range of them as possible without the introduction of reducible elements, and to make the composition such as to obviate all danger of their undergoing, on long expos- ure, a crystalline change or devitrification which would alter or disturb their certain melting down, the condi- tion in which the ultimate value of a large part of the series of Prinsep's alloys failed. This problem was undertaken by Seger, 2 and solved 1 Thonindustrie-Zeitung, 7886, p. 135. 2 Thonindustrie-Zeitung, 1886, pp. 135, 145, 168. PYROMETRY. 33 so successfully that his " Normal Pyrometric Cones" an- swer all conditions required in ceramic pyrometry. He established that the most fusible mixture of the porcelain glaze type was one of the chemical compo- sition : From this the more infusible glasses were obtained by the systematic increase of silica, with a proportional in- crease of alumina to correct the well-known tendency toward devitrification of highly siliceous glasses. By the partial substitution of alumina with ferric oxid, the series was brought down, in its member, o. 3 K 2 0}o. 3 Al 2 3 } d Si0 o.yCaO jo. 2 Fe 2 3 I 4 " to equal the melting point of the alloy ten per cent, plati- num, ninety per cent, gold, and thus became continuous with the most useful members of Prinsep's series. The use of the cones proved to be so practical for the purpose that the need to substitute the Prinsep alloys by equivalent cones was soon felt, and -E. Cramer, 2 making use of the well-known fact that if the acidity of a glass be maintained, but the silica substituted by boracic acid, the melting-point is lowered, increased the series to be- gin with the melting-point of silver. These pyrometers are, in the entire range demanded by ceramic firings, represented in the following list. 1 Thouindustrie-Zeitung, 1886, p. 168. 2 Thonindustrie-Zeitung, 1892, p. 155. 34 THK CHEMISTRY OF POTTERY. For the benefit of those who find it difficult to free themselves from the association of thermometric degrees with advances in heat, such degrees have been interpo- lated, from calorimetric determinations, as follows : Cone oio, melting down with pure silver, 960 Celsius; cone i, melting with the alloy ninety pet cent, gold, ten per cent, platinum, 1 150 Celsius ; cone 20, 1700 Celsius. Cone Estimated temperature in degrees number. Chemical Composition. Celsius. OIO ] I 0.7CaO o.2Fe 2 O 3 o.3A! 2 O 3 3-5oSi0 2 o.5oB 2 O 3 } 960 09 \ r o. 3 K 2 o L o.7Cao o.2Fe 2 O 3 o.3A! 2 O 3 3-55810, o.4 5 B 2 3 } 979 08 < f o. 3 K 2 L o.7CaO o.2FeO, o. 3 Al 2 6 3 3.6oSiO 2 o.4oB 2 O 3 } 998 j f o.3K 2 O o.2Fe 2 O s 3 .6 5 Si0 2 V T(~lT *7 07 < L o.;CaO o.3A! 2 3 Q.35BA i ALII y 06 < f o. 3 K 2 L o.7CaO . o.2Fe. 2 O 3 o.3A! 2 O 3 3 .7oSi0 2 o.3oB 2 O 3 | 1036 f o.3K 2 O o.2Fe 2 3 3-75Si0 2 I Tf~\C C 5 < t o.7CaO o.3A! 2 O 3 0.258,0, | IO 55 04 < r o. 3 K 2 t o.7CaO o.2Fe 2 O 3 o. 3 Al 2 3 3 .8oSi0 2 o.2oB 2 O 3 } I0 74 03 - f o. 3 K 2 t o.7CaO o.2Fe 2 O, o.3A! 2 O 3 o.isB 2 O 3 } 1093 02 - f o. 3 K 2 I. o.7CaO o.2Fe. 2 O 3 o.3A! 2 3 o.ioB 2 O 3 I 1112 01 < r o. 3 K 2 o i. o.7CaO o. 2 Fe 2 3 o.3A! 2 3 o.o5B 2 O 3 } "SI I < r o. 3 K 2 o t o.7CaO o.2Fe 2 O 3 o.3Al. 2 O 3 | 4SiO 2 | 1150 2 - f o. 3 K 2 ( o.7CaO o.iFe 2 O, o.4A! 2 3 } 4SiO 2 } U79 3 < f o. 3 K 2 ( o.7CaO o.o5Fe 2 O 3 o.45A! 2 O 3 } 4SiO 2 | 1208 4 n ^PaO { o.5A! 2 O 3 } 4Si0 2 } 1237 PYROMKTRY. 35 Estimated temperature Cone in degrees number. Chemical composition. Celsius. 5 { ayCaO { '5 A1 2 3 } 5SiO 2 } 1266 6 { C:7CaO { - 6A1 *3 } 6Si 2 } I2 95 ' 1323 8 '35* 9 { C^CaO { -9 A1 2 03 } 9Si0 2 } 1381 10 { C.'yCaO { I ' A1 '3 } IoSiO ^ } HIO 11 { ayCaS { X - 2A1 2 3 } I2 iO 2 } 1439 12 { ayCaO { r ^ A1 203 } i4Si0 2 } 1468 * I 49 7 15 2 ' IA1 *3 2rSi 1555 16 18 { al&O { 3' IA1 2 3 } 3iSi0 2 } 1642 19 { ayCaO { 3.5A1 2 3 } 3 5SiO 2 } 1671 20 { ayCaO { 3-9Al 2 O 3 } 3 9SiO 2 } 1700^ In order to secure for these pyrometric mixtures a standard character, and justify the name, " Normal 1 Thonindustrie-Zeitung, No. 49, ffyj, p. 1252-3. 36 THE CHEMISTRY OF POTTERY. Pyrometric Cones," they are manufactured by the Prussian Government in their ceramic experiment sta- tion at the Royal Porcelain Works, in Charlottenburg, near Berlin. It is, however, not at all difficult to make them of suf- ficient accuracy, from the excellent native materials that are on the market in the United States. Thus the original series of Dr. Seger are made as fol- lows : The potash is taken in the form of orthoclase ; the remaining alumina that is requisite is introduced in the form of kaolinite ; and the remaining silica, not supplied by these, is added as quartz. Calcium carbon- ate gives the calcium oxid, and those needing ferric oxid have it directly added. The necessary amounts of these in- gredients are weighed off for each num- ber, introduced with water into a small porcelain jar mill, thoroughly ground for half a day, settled, the water drawn off, and the mixture dried. It is then worked into a mass with dextrin muci- lage, and formed into tetrahedrons six NATURAL SIZE. centimeters high and one and one-half centimeters on the sides of the triangular base. 1 The materials selected by Dr. Seger as sufficiently pure for the purpose analyzed as follows : 1 Thonindustrie-Zeitung, 1886, p. 136. A SEGER CONE IN PYROMETRY. 37 Rorstrand feldspar. Per cent. Silica 64.32 Alnmina 19-41 Ferric oxid.- 0.14 Lime trace Magnesia 0.35 Potash 12.90! Soda 2.10 / Lossonglow'g 0.57 Carbon dioxid 99-74 Zettlitz kaolin. Per cent. 46.87 38.56 0.83 trace trace .06 12.73 100.05 Norwegian quartz. Per cent. 98.52 1.04-1 0.04 j 0.40 Carrara marble. Per cent. 1.00 0.12 54-93 0.20 43-76 100.02 One can make with the raw materials produced and put on the market in this country quite as near an ap- proach to the theoretical figures, and need, therefore, have no hesitancy in using them for such a purpose. The following analyses will verify this fact : Kaolin, from Western North Carolina, of which the writer has made over fifty analyses, not varying mate- rially from the following extremes : Analyzed November, January, 1894. 1890. Silica 46.47 Alumina 38.82 Ferric oxid 0.89 Lime 0.28 Magnesia 0.25 Potash 0.63 -j Soda 0.12 } Combined water 3-34 100.53 46.67 38.14 0.36 0.50 0.09 0.64 13.61 100. I I Feldspar from New York : 38 THE CHEMISTRY OF POTTERY. Per cent. Silica 65.85 Alumina 19-32 Ferric oxid 0.24 Lime 0.56 Magnesia 0.08 Alkalies 1 14.10 100.15 Quartz from Illinois : Alumina o. 155 Ferric oxid 0.069 Lime 0.026 Magnesia 0.013 Alkalies 0.112 Total impurities -375 Silica (by difference) 99-625 100.000 Commercial Whiting : Silica trace Alumina Ferric oxid / o> * 3 Magnesia trace In order to make cone No. i, having the formula o.3K 2 O)o.2Fe 2 O 3 ) USi0 2 o.7CaO)o. 3 Al 2 3 ) there would be ground together and formed into tetra- hedrons, as described, 0.3 equivalents feldspar 83.55 parts. 0.7 " calcium carbonate .. 35.00 " 2.2 " quartz 66.00 " 0.2 " ferric oxid 16.00 " 1 Combining weight, 45.9. PYROMETRY. 39 Xo.5Al,0 3 .4Si0. 2 . For cone No. 4, having the formula o. 3 K 2 o.yCaO there would be used 0.3 equivalents feldspar 83.55 0.7 calcium carbonate- 35.00 0.2 ' kaolinite 2 5-9 1.8 " quartz 54.00 For cone No. 10, having the formula o. 3 K 2 CM ViAl 2 O 3 .ioSiO 2 . o.yCaO J there would be used 0.3 equivalents feldspar 83.55 0.7 " calcium carbonate - 35.00 0.7 kaolinite 90.65 6.8 " quartz 204.00 For the introduction of boracic acid into the lower scale, according to E. Cramer this element is first fixed as a lime-soda glass of the composition o.5Na 2 O) (2SiO 2 . to.iAl.Qj o.sCaO J liB 2 3 . by melting together 191 parts crystallized borax. 50 " calcium carbonate. 52 " kaolinite. 96 " quartz. Then equivalent amounts of this glass are ground with equivalent amounts of feldspar, kaolinite, quarz, calcium carbonate, and ferric oxid, to make bodies of the corres- 40 THK CHEMISTRY OF POTTERY. ponding formulas, and the masses are formed into tetra- hedrons, as described before. The use of the Seger cones is very simple. One of the number, representing the heat which it is wished to reach, is set upright in each of the proper places of the kiln opposite a spy-hole, which latter is closed with a plate of mica. When, by the advancing heat, the cone bends until, finally, its apex touches the base on which it is set, the temperature which it is intended to indicate is reached, and the firing stopped. The cones should be set about three feet back from the kiln-wall, in a piece of fire-clay piping or muffle ex- tending to the kiln-wall and running a foot or more back of the pyrometer, so that this is seen, when the fire is well advanced, as a luminous body in a darker field. The pyrometer is best stuck with a little soft mud on a fire-clay shard or tile, to prevent its falling over. The determination of the proper pyrometer to be used as the index of the heat to w T hich a clay under examina- tion should be burned is best explained in connection with the examination of clays for specific industries, and is deferred to these chapters. CHAPTER IV. CLASSIFICATION OF CERAMICS. GENERAL classification of clay manufac- tures will be of service in this connection, particularly to the chemist, who may not, heretofore, have given much considera- tion to ceramics, in order to emphasize characteristics that are of moment to the potter. The varieties of ware manufactured in the United States are not very great, and can be quite sharply char- acterized. It will therefore be most practical to take as simple a classification as possible, especially as it is mainly made for the benefit of the technical chemist, and not for the connoisseur, and need merely enumerate the chief technical characteristics without classifying all their possible combinations. The old classification of Brogniart is accordingly taken as a basis for the purpose. He divides all ceramic products into three main classes and nine subdivisions, as follows : FIRST Civ ASS. Bodies sufficiently soft to be scratched with a knife, of a sandy-argillaceous character or con- taining lime and fusible in the heat of the porcelain kiln. First Division. Bodies soft burned, with a dull un- glazed surface. Examples : Brick, building terra-cotta, drain tile, etc. 42 THE CHEMISTRY OF POTTERY. Second Division. With the faint luster of an alkali- earthy-silicate, the gloss produced by polishing or by incipient fusion. Example : Antique vessels. Third Division A body similar to that of the first or second, but with a transparent lead glaze. Example : Common pottery ware. Fourth Division. Ware enameled with nontransparent glazes containing tin oxid. Examples : Tile for porce- lain stoves, and common dishes. SECOND CLASS. Bodies that are hard, nontransparent, of a siliceous-argillaceous mass that cannot be scratched with a knife, and are infusible. Fifth Division. White body with a transparent lead glaze. Examples: Fine faience and dish- ware. Sixth Division. A colored body with an earthy alka- line silicate glaze, or sufficiently dense to require no glaze. Example : Stoneware. THIRD CLASS. Bodies hard, translucent, high* in alkali, siliceous-argillaceous, softening in the hardest fire. Seventh Division. Kaolinitic body with a glaze mainly feldspathic. Example: True hard porcelain. Eighth Division. Body of kaolin, plastic clay, and bone ash, with a lead-boracic-acid glaze. Example : English soft porcelain. Ninth Division. Body of a glass-frit with addition of clay and a lead glaze. Example : French soft porce- lain. In addition to the manufactures innumerated by Brog- niart in his first division, flooring tile would belong in CLASSIFICATION OF CERAMICS. 43 this category. It is now, however, customary to burn these, as well as the better grades of building terra-cot- tas, so hard that they are not to be scratched with a knife ; enabling them to better resist wear and the dis- integrating action of the elements. Our common " red ware" corresponds exactly to that of the third division. In the fourth division, the only representatives of American ware are white enameled brick. Occasionally dishes are brought to the New England coast towns from Fayal, in the Azores, corresponding exactly to those described by Brogniart ; but clays rich in lime, such as the tin enamels melt best on, are only used with us in the manufacture of cream-colored brick not for dishes ; and tin-enamels, apart from their employment on brick, have only been employed on metal vessels. Our common dishes, better in grade than red ware, are made in yellow and Rockingham ware, which should come under the second class, in having a hard siliceous- argillaceous body that cannot be cut with a knife, and is porous and opaque. There is no division, however, into which it will exactly fit, for the body is yellow and the glaze transparent and plumbiferous, and, in the case of Rockingham ware, manganiferous. All enumerated varieties of ware of the fifth division are common with us, and are variously termed cream- colored ware ( " C. C. " ) , white granite ware ( ' ' W. G. " ) ' ivory white ware, ironstone china, etc. Similarly, all classes of the sixth division are common in the United States, the stoneware being commonly 44 THE CHEMISTRY OF POTTERY. divided into "salt-glazed" and "slip-glazed" products, the meaning of which terms will be explained in the chapter devoted to these wares. The paving-brick now extensively manufactured in the United States come under the unglazed products of this category. The wares thus far enumerated are called in English, collectively, "pottery," and it is with the chemical character of these, particularly with the glazed varie- ties, that this book, in the main, concerns itself. Those coming under the third class have, within recent years, begun to be manufactured in quantities assuming proportions of commercial importance, with the exception of the French pate tendre, which is not made. We have, however, a kind of ware standing between that of the fifth division and porcelain, for which there is in this classification no exact place. It is commonly known as "hotel china," and resembles porcelain in having a vitreous body, rich in alkali, that in thin places would be translucent, but which is covered, like the ware of the fifth division, with a clear lead glaze. There is one misnomer in popular use that may cause the chemist looking up the proper significance of the term some misunderstanding; namely, the word " Ma- jolica. ' ' In the United States, it is commercially a cheap ornamental ware, with a white, rather soft body, deco- rated with designs in relief, and these designs tinted with soft, transparent, colored glazes. Originally, the term was applied by the Italians of CLASSIFICATION OF CERAMICS. 45 the fifteenth and sixteenth century to white enameled ware decorated in lusters that is, in colors having a metallic reflex which was, at first, of Moorish make, and brought to Italy, largely from the island of Majorca and afterward it was applied to all ornamented white enameled ware ; ware belonging, however, in all cases, to the fourth division of Brogniart. The word faience is also one commonly misunder- stood. Originally, it meant majolica from the potteries of Faenza, but finally became synonymous with majolica in general, meaning an ornamented white ware. But the name was most commonly used in France, where white ware, made with white clays and a transparent plumbiferous glaze, largely took the place, both in orna- mental and useful articles, of the white tin-enameled pottery ; so that the term lost its specific meaning, and came to mean pottery of a little better grade in general. Ceramists are inclined now to use the term " faience" in distinction from the application of the word ' ' majolica, ' ' using it to designate potteries with transparent glazes, and majolica, as a generic term for potteries with intrans- parent glazes. In this sense, our potters are already using the word faience quite correctly for ornamented wares belonging both to the third and fifth division of Brogniart. It is to be hoped that our incorrect application of the word majolica, which is only used as a trade name for an inferior grade of ornamental ware, will sink with the latter into obloquy. The word "enamel" has attained a specific signifi- 46 THE CHEMISTRY OF POTTERY. cance which the public frequently misapplies, using it indiscriminately for all glazes or glassy coverings. Technically, the word enamel means an nontransparent glass, completely covering the body upon which it is melted, the nontransparency being due to the presence of tin oxid, arsenic, bone ash, or cryolite. It is important that the technical significance be ad- hered to, and that the indiscriminate popular use of the word be realized, to prevent misunderstandings. Ceramic wares pass into each other by almost imper- ceptible gradations. The great variety of possible bodies, glazes, decorations, means of shaping, applica- tions, opens an almost infinite field to the fancy of the worker and the greatest number of combinations to meet specific uses. It is an error to suppose that the so-called finer grades of ceramics, though made of pure kaolin, feldspar, and quartz, and white-burning secondary clays, alone offer artistic possibilities or problems worthy of the chemist's consideration. On the contrary, the wares that have brought us the interest of foreign countries belong to the first, third, and sixth division of Brogniart's classification; they repre- sent the largest commercial factor in our clay industries, and in their midst the greatest artistic success has been achieved. While we are better supplied with pure pottery mater- ials than any other country, our common clays will, and properly should, always claim the greatest share of interest, not only because they must be the raw mate- CLASSIFICATION OF CERAMICS. 47 rials of the bulkiest manufactures, and those in which the greatest economies will have to be exercised, but be- cause they will always offer the greatest technical and artistic possibilities. CHAPTER V. POTTERY GLAZES. S EXPLAINED in the previous chapter, pot- tery bodies have, in the main, an earthy fracture, and even when sufficiently hard so as not to be scratched with a steel point, may absorb fifteen to twenty per cent, of their vol- ume of liquid, with avidity. But even vitre- ous bodies that were very brittle, and had a conchoidal glassy fracture, were found by the writer to absorb up to one and eight-tenths per cent, of their weight, or nearly four per cent, of their bulk, of distilled water. From this it will be seen that all pottery and porcelain articles, if they are to serve as containers or be subject to conditions making frequent cleaning necessary, should be covered with an impervious surface coating. This is practically always a glass. The glass, or "glaze," as it is designated, must neces- sarily answer the following conditions : 1 i ) It must have practically the same coefficient of ex- pansion as the body of the ware, in order not to " craze" ; that is, tear with transverse cracks ; or not to push off on high places and sharp angles, tearing the body along with it, called " shivering." (2) It must bear exposure in thin layers to gradually POTTERY GLAZES. 49 increasing temperatures, lasting hours, or even days, during baking, without suffering devitrification. (3) It must be sufficiently tenaceous when melted so as not, in the necessarily long time in which it is in this state, to run off of upright objects, nor yet so stiff as not to flow perfectly smooth on flat ones. (4) It must hold dissolved, without unsightly separa- tions, metallic oxids that have been added for imparting color. (5) It must not exert too strong a solvent action upon oxids used in painting and printing upon the body of the ware, over which the glass is afterward melted like a covering of indestructible varnish. The difficulties attending the fulfilling of these general conditions, besides many special ones that occur in con- nection with every kind of ware, have naturally led pot- ters to guard the composition of those that answered their particular purposes with great jealousy. The formulas that have been published are, in the main, misleading, or, where this is not the case, they are, as a rule, equally useless, inasmuch as the inter- dependence of glaze, clay, and the condition of firing is so intimate that even a correct description of one, with- out the others in conjunction, is of no practical service. It is furthermore impossible to judge a priori of what may be expected of a glaze in melting-point, coefficient of expansion, and tendency toward devitrification from its gross composition by weight. This must be resolved into the chemical formula, and in order to do this, the chemical character of the compound and the function of its elements must first be understood. 50 THE CHEMISTRY OF POTTERY. The practical experience of the potter teaches him that he needs as a glaze a body resembling in hardness, color, and refracting power ordinary flint glass ; but that ground flint glass, painted on his ware and exposed in the kiln, will not, under any circumstance, give him a glaze having the qualities the glass before, in itself, had, but a puck- ered, de vitrified mass, barely sintered together. To whatever circumstance this may be due, he knows still that what he needs is a glass or a mixture composed of the elements of a glass, which, to meet his special pur- poses, can be modified in the following ways. To reduce the melting point, the alkalies, alkaline earths, and certain oxides of the heavy metals, particu- larly and almost indispensably lead, must be increased ; that with this increase and resulting greater fusibility, the glaze becomes more brilliant ; that is, attains a higher refractive index ; but will inevitably craze if the increase of these bases exceeds a certain amount. Conversely, to counteract the crazing of the more fusi- ble glazes on certain clays, the addition of quartz to the glaze corrects this defect, with attendantly growing diffi- cult fusibility, until, finally, the coefficient of expansion is so far changed that the glaze tends to push off from the body at elevated places and angles, and this increase of quartz may be pushed to the extent that the glaze will shatter the entire piece of ware which, as already said, is called ' ' shivering. ' ' These are the extremes between which the proportion of base and acid must lie. The action of boracic acid is similar to that of the quartz or silica in changing the coefficient of expansion, POTTBRY GLAZES. 5^ in opposition to the effect of the bases, but the melting- point does not rise in the .same proportion. On the con- trary, by the substitution of silica with boracic acid, the melting point falls very materially. The glaze also in- creases in brilliancy with this addition, but is softer, being more easily scratched, and exerts a far greater sol- vent action on under-glaze colors. Glazes composed of the above constituents, while they seem to answer every requirement when melted in the experimental muffle, fail utterly when used in the large ware kiln. The potter has found empirically that no glaze can be perfect without the addition in certain pro- portion, of either clay, feldspar, or Cornwall stone, and the chemist will readily recognize that the new element introduced by these materials is merely alumina, and that it is this and other chemically similar elements which keep the glaze from devitrification in the protracted 'glost-fire" of the potter's oven. In calculating the chemical formula of glazes, the ba- ses, with the exception of those of the alumina group, are reduced to their equivalent proportions, and multi- plied by such a factor that their sum equals unity. The corresponding equivalent weights of the members of the alumina group are then recorded together, as they oc- cupy an intermediary position, in fixing the relations of the bases to the acids, under protracted firing, and then the equivalent weights of the acids are noted. A glaze would then have the formula icROx. R 2 O 3 y. SiO 2 . A common flint bottle-glass was found to have the composition : 52 THE CHEMISTRY OF POTTERY. Soda .............................. 17.80 per cent. Lime ............................. 6.37 " " Magnesia ......................... 2.83 " " Alumina .......................... 0.53 " " Silica (by difference) .............. 72.47 " " 100.00 the chemical formula being 0.61 Na. 2 O | 0.24 CaO f 2.56S1O,. ^ 0.15 i.o RO. Finely ground and applied to a burnt shard of clay, it was scarcely possible to get with it a bright transparent glaze in the experimental muffle. By taking a half equivalent of the glass and one-half equivalent of a lead compound, and making up the silica by addition of ground flint, as follows : r 0.305 Na 2 O i 0.5 equivalent of the glass ! 0.120 CaO [ 1.28 SiO 2 1 0.075 MgO J t 0.5 PbO 1.28 SiO 2 i.o 2.56 made by thoroughly mixing in a mortar with water 52.71 flint glass, 55.75 litharge, and 38.4 flint ; and painting the mixture on a potsherd, it flowed to a brilliant glaze at a temperature above that of the melting point of an alloy, fifty per cent, silver and fifty per cent. gold. In the potter's glost-kiln, under a thirty hours' fire and two days' cooling, the glaze was puckered and dull. POTTERY GLAZKS. 53 The introduction of two-tenths equivalents of alumina, in the form of china clay, with allowance for the silica necessarily introduced with it, corrected it for these con- ditions so that the glass, composed of 0.5 equivalent flint glass, 52.71 parts, 0.5 " litharge, 55.75 " 0.2 " china clay, 25.9 " 0.88 " flint, 26.4 " answered on a shard of the proper coefficient of expan- sion, and came brilliant and clear from the long fire. This example, illustrating the functions of the differ- ent elements composing a glaze, sufficiently explains the structure of the chemical formula and the insight it gives one into a complicated glaze formula. It must not be inferred from the above, however, that a flint glass should be the basis of every pottery glaze, nor that every glaze must contain lead. Potters are accustomed to distinguish between "raw" and "fritted" glazes. The former have merely their constituents ground together and are melted to a glass for the first time when exposed on the ware to the glost- fire. Fritted glazes are those that contain a glass or, rarely, are such as have been entirely melted to a glass before grinding and putting on the ware. The purpose of fritting all or a part of the constituents of a glaze is primarily to render insoluble such constitu- ents as would, by their solution in water, be carried away by the customary wet-grinding and the application of the glaze, suspended in water, to the ware. This is the 54 THE CHEMISTRY OF POTTERY. case when soda-ash, pearl-ash, borax, niter, boracic acid, and similar constituents are to be incorporated in the glaze . At times, too, when a glaze contains large amounts of alkaline earths, with little or no lead oxid, fritting is necessary, as the alkaline earths, not being in themselves fusible, enter with difficulty into combination, and would require, for the incipient union, heats far too excessive for the glass when actually formed. In order to illustrate the calculation of the formula of a glaze from its analysis, the following example will serve. The sample was taken from a plain-glazed tile of Amer- ican manufacture, as explained in the first chapter, and gave the following percentage composition : Lead oxid 24.21 per cent. Alumina 11.58 " " Silica 49.72 " " Lime 0.86 " " Magnesia 0.73 " " Potash 0.42 " Soda 4-68 " " Sulfuric anhydrid 1.46 " " Boracic acid (by difference) 6.34 " " 100.00 The chemical formula of the glaze, then, is 0.4952 PbO % r3.78oSiO a 0.1296 CaO [0.5728 A1 2 O 3 J 0.3752 Na 2 O ) 1 0.413 BO 3 i.o RO In order to make this glaze, one would melt together to form a fritt POTTERY GLAZES. 55 Parts give Melted. 0.207 equivalent borax 39.5 borax glass 20.91 0.168 " soda ash... 8.9 Na.X) 5.21 0.130 " whiting 6.5 CaO 3.64 o.ioo " china clay . 13.0 Al. 2 O 3 2SiO 2 11.15 0.800 " flint 24.0 SiO 2 24.00 The charge 91.9 melts to. . . 64.91 For the glaze, grind together 0.4952 equivalent white lead 64.23 0.4128 " china clay 53-46 2.78 equivalents flint 83.4 fritt 64.91 266.00 It will be noted that the analysis contains an apprecia- ble amount of sulfuric anhydrid, which was ignored in the calculation of the analysis. The anatytical sample was taken from the edges of the tile, where the glaze was thickest ; but here, although not over the surface of the glaze generally, there was a crystalline scum. This was " glass-gall," or a crystal- line separation of sulfates. The proportion of sulfuric anhydrid was therefore larger than if the sample had been averaged over the entire face. Sulfates are highly objectionable in glazes. They may form either unsightly crystalline separations on the edges of the ware, or cover its entire face with a very thin greasy scum, and when present in large amounts, may float in drops on the glaze, from which, on cooling, they may easily be pinched out, leaving pits in its surface. The chemist must make certain that the materials are 56 THE CHEMISTRY OF POTTERY. free from sulfates, and even look, at times, to the water in which the glazes are ground. Sulfates are frequently formed by the absorption of sulfuric anhydrid generated in the kiln from the burning of a very sulfurous coal. Obviating difficulties that may arise from this will be discussed under the subject of firing. It may merely be noted here that the phenome- non, commonly called "sulfuring" by potters, is the re- duction of the lead of the glaze, and is often the catch- word with which a careless fireman tries to throw on the fuel, the responsibility for inattention to clinkering or other manipulations affecting the draught in firing. The phenomenon of a reduction of the glaze frequently hangs together, in so far, with the presence of a large amount of sulfur, in that the coal contains the latter, mainly in the form of iron pyrites, which, leaving on burning the iron oxids, give a more or less fusible ash, or clinker, which is very liable to choke off a part of the necessary air supply. The range of formulas of pottery and porcelain glazes is from i RO o.i R 2 O 3 1.5 SiO 2 to i RO 1.2 R 2 O 3 12 SiO 2> and the corresponding range of temperatures required in their burning is from the melting point of silver, or pyr- ometric cone oio of Cramer's scale, to pyrometric cone 1 8 of Seger's scale, probably equivalent to the melting point of an alloy eighty per cent, platinum, and twenty per cent. gold. In this wide range of possible compounds, the one suit- POTTERY GRAZES. 57 ing the particular conditions demanded by the clay used, and the process of manufacture, and the uses to which the ware is to be put, must be sought. For the potter to do this empirically is certainly a very difficult and un- certain undertaking ; for the chemist, beginning with a mixture based on a definite chemical formula within the given limits, and systematically introducing the various possible elements in proportions varying by definite equivalents, it is but a question of time to get the desired formula, which, with experience and skill in experi- menting, need not take long. CHAPTER VI. RED WARE. I HE simplest and cheapest of glazed pottery is called from its color, " red ware." It is or- dinarily formed from the same materials used for making red brick; alluvial mud found in the river valleys and weathered ferruginous shales. It is important that the material contain but little lime, which if present in any considerable amount, destroys the bright red color imparted by the iron oxid, giving unsightly ware. At a heat sufficient to melt a wire of pure silver to a bead, the clay should bake so hard that it can barely be cut with a knife. The specimen baked at that heat should adhere when touched to the tongue, and be of a bright red color. The clay must be of such composition that with the hardness attained at the given temperature (from the melting point of silver to not exceeding that of an alloy of seventy-five per cent, silver, twenty-five per cent, gold) , it will have practically the same coefficient of expansion, and hence bear without fracture, a glass fusible at that heat. A glass or ' ' glaze ' ' of this character would be one having a chemical formula lying between iPbO, o.i A1 2 O 3 , iSiO 2 and iPbO, o.i5Al 2 O 3 , i.5SiO 2 made by RED WARE. 59 grinding together 129.7 parts white lead, 13.0 to 19.5 parts china clay, and 24 to 21 parts quartz. In order to absolutely resist the action of acids on cook- ing utensils of "red ware," it would be desirable to use more acid glazes, but such is not the practice in this in- dustry ; although there is a practice that should be abso- lutely condemned and against which chemists should throw their influence, namely, that of using litharge or galena alone as the glazing substance and depending on its taking up sufficient alumina and silica from the body of the ware, during fire, to form the glass. A glaze so formed is certain to be basic on the surface, and is sure to be attacked by the weakest acids used in cookery. The red ware potter has no china clay at his disposal, and usually makes his glaze by grinding the lead prepara- tion with a loamy sand. The glaze then has frequently some such composition as this (taken from practice). This glaze would be of a yel- lowish color, which is not objectionable. The coefficient of expansion of the clay depends on the amount and fineness of the uncombined silica and feldspathic detritus it contains, which constituents are determined by the "rational analysis." A practical "red ware" clay, burning at the indicated temperature to the required hardness and color, and bearing the glaze without suffering fracture ( ' 'shivering' ' ) or the glaze itself cracking ("crazing"), being of suffi- 60 THE CHEMISTRY OF POTTERY. cient plasticity and having from the clay to the baked condition a linear shrinkage of six per cent., is a weath- ered shale of the following analysis : TOTAI, ANALYSIS. In sol. in Per cent. H 2 SO 4 and Na 2 CO 3 . Silica 74.75 57.20 Alumina 12.55 0-62 Ferric oxid 5.28 0.70 Lime 1.28 0.77 Magnesia 0.85 o.oo Alkalies 2.27 1.80 Combined water 3.23 100.21 61.09 RATIONAL ANALYSIS. Per cent. Clay substance 39.12 Quartz 52.54 Feldspathic detritus 8.55 PERCENTAGE COMPOSITION OF THE " CI* Quartz 29.72 Calcium sulfate 1.87 Calcium carbonate 7.09 Magnesium carbonate 7.13 100.47 PERCENTAGE COMPOSITION OF THE CI,AY SUBSTANCE. Per cent. Silica 5LQ9 Alumina 33.12 i Combining weight 45.8. 84 THE CHEMISTRY OF POTTERY. Per cent. Ferric oxid 7-7 Lime 0.35 Alkalies 4-55 Combined water 3.11 99.92 As this clay is used upon the ware as a glaze, it is im- portant to calculate from its gross analysis the chemical formula that the glass resulting from melting it will have, in order to find the type of glaze required in this industry; to be able to reproduce it from other materials ; or to sys- tematically modify it for the purpose of producing glazes of lower or higher melting-points, to accommodate clays that should be burned at temperatures other than those required for this particular slip-glaze. The chemical formula of the Albany slip-glaze, calcu- lated from the above analysis, would be : i.o RO 0.689 R 2 O 3 The following slip-glaze, which is also burned at about the same heat with the Albany, becomes of a rich red- brown color ; it is composed of a mixture of fusible clays, the proportions having been established empirically, are kept secret and are unknown to the writer. Per cent. Silica 55-67 Alumina 14.18 Ferric oxid 3-5^ Lime 8.00 Magnesia 2.84 STONEWARE. 85 Per cent. Manganous oxid 0.41 Alkalies 1 5.01 Sulfur trioxid i.n Combined water and carbon dioxid 9.87 100.65 The chemical formula of the glaze resulting from melt- ing the above would be : 0.2290 KNaO ^ 0.0203 MnO I 0.4832 A1 2 3 0.2492 MgO \ . 078l Fe2 o i.o RO 0.5613 R 2 O 3 It appears from these formulas that the common slip- glaze for stoneware must have a chemical formula not differing widely from that of Seger's pyrometric cone number two, namely : o.75CaO/o.iFe 2 O 3 i.o RO o.5R 2 O 3 As clays high in mineral detritus are common, abound- ing especially along the borders of former glaciation, the chemist may frequently be called upon to show the stone- ware potter how a local slip-clay should be mixed in order to melt in a fire most suitable for the body of his ware. It is commonly believed by potters, and also by chem- ists, that if a clay be infusible, it is only necessary to add sufficiently of some basic material, as an alkali, alkaline Combining weight 38.4. 86 THK CHEMISTRY OF POTTKRY. earth, litharge or ferric oxid to obtain a glass. This is not the case, as the relation of the R 2 O 3 elements to silica is vital and a glaze of this type may be infusible from being too basic. The chemist on working out the chem- ical formula of the clay, which it is desired to use, will frequently find it to be too aluminous, requiring the ad- dition of silica quite as much as that of a base. In order to illustrate this, systematic experiments were made with a slip-clay that merely softened and swelled up at the melting down of pyrometric cone eight, but was far too infusible to run at that heat. It analyzed as fol- lows : The portion in- The entire clay soluble in H 2 SO 4 and contains Na 2 CO 3 contains Per cent. Per cent. Silica 56.59 23.56 Alumina 25.96 i.oo Ferric oxid 2.30 0.07 Ljme i. 60 0.06 Magnesia 1.90 o.io Alkalies 1 5.33 0.40 Combined water .... 6.22 Sulfuric anhydrid ... i .07 100.97 25.19 RATIONAL ANALYSIS. Per cent. Clay substance 73-9 6 Feldspathic detritus 5-44 Quartz 19-75 Calcium sulfate i .82 100.97 1 Combining weight 31. STONKWARE. 87 PERCENTAGE COMPOSITION OF THE CI,AY SUBSTANCE. Per cent. Silica ............................. 44-65 Alumina .......................... 33-74 Ferric oxid ........................ 3.01 Lime .............................. i .09 Magnesia ......................... 2.43 Soda .............................. 6.66 Combined water . . ............... 8.40 CHEMICAL FORMULA OF THE MEI/TED CI /iR 0.30 Combined water. 2 5 1347 12.24 13-34 13.10 0.97 13.20 Insoluble in H 2 SO 4 and Na 2 CO 3 100.89 } 2.46 100.64 100.53 0.47 100.32 2-54 99-99 0.40 Many of our white-ware potters being of English train- ing and using formulas calling for English raw materials, the following analyses of three English china-clays com- monly met with in the American market, are given for comparison with our native products : Silica Per cent. Per cent. Per cent. itQ r T 47.90 ^8 2Q 47-1U Ferric oxid ^o.zy O/'OJ O u '// 0.44 Magnesia 4o 5 Alkalies o ^6 U O4 Combined water .... 12.76 1345 5 11.80 Insoluble in ) H 2 SO 4 and Na 2 CO 3 j 100.90 o.oo 99.80 1.24 100.66 8.33 The linear shrinkage of china-clays burned to the melt- ing clear of orthoclase feldspar, about the melting-point of pyrometric cone nine, is from four to eight per cent. They may require as much as forty grams of water for 100 grams dry clay, to make a mass sufficiently soft to allow the Vicat needle to penetrate four centimeters. f |S" *$ #? > *4 RAW MATERIALS OF WHITE-WARE BODIES. 97 Nevertheless the mass of china-clay is of poor plasticity and does not, in the air-dry condition, show a tensile strength of more than 2,000 to 2,500 grams per square centimeter. The commercial value of a china-clay depends largely upon the whiteness of the body it will produce. Its ap- pearance in the clay state is absolutely no criterion of this, as a clay which is very yellow when unburnt may become very white in the fire. Unfortunately the per- centage of iron in china-clays, as shown by their anal- yses, stands in no direct relation to the tints of the burned products. Nor can a true comparative estimate of the clays be made when they are merely once burned, that is, in the biscuit state. It requires the covering of a thin clear glaze to bring out the true tint. For the purpose of comparing the tints of china-clays, they must be made up into bodies by the addition of the same proportions of quartz, and feldspar, of the same lot, burned in the same biscuit fire, and covered to an equal thickness with a transparent glaze, which had best not contain over one-half equivalent of lead-oxide, as a glass high in lead is of a yellowish cast, nor should the glaze have been tinted by an addition of cobalt oxide. To a limited extent, silicates of alumina resembling halloysite, Al 2 O 3 .2SiO 2 .4H 2 O, have at various times found a local use in place of kaolin. A clay of this nature from Lawrence County, Indiana, which has had, perhaps, a more extensive pottery-appli- cation than any other of this kind, can not be disintegra- ted by mere washing, but ground in water it gives a very 98 THK CHEMISTRY OF POTTERY. voluminous mass resembling starch-paste, and is pasty rather than plastic in character, and dries down to a horn-like body. Burned at the heat of running feldspar, it has a linear shrinkage of twenty per cent, from its size in the clay state. The burned clay is much denser than kaolin and has a faintly greenish tint. Partial analyses of the material analyzed at intervals during a number of years, contained silica ranging from 39.74 per cent, to 41.15 per cent., and combined water from 17.21 per cent, to 17.78 per cent. An exhibition specimen, analyzed in full, ran much closer to the proportions of kaolin, though with the com- bined water still high, as follows : Per cent. Silica 44-79 Alumina 38.77 Lime i .00 Magnesia 0.25 Alkalies 0.67 Combined water 15-49 100.97 Insoluble in H 2 SO 4 and Na.,CO ; , 0.98 It is not unlikely that more materials of this nature may be found and applied to special uses in the pottery art. A specimen sent to the writer, as a kaolin from Inyo County, California, appears from the analysis to be of this nature : RAW MATERIALS OF WHITE-WARE BODIES. 99 Portion insoluble Constituents of the in H 2 SO 4 and entire clay Na 2 CO 3 Per cent. Per cent. Silica 44-74 8.99 Alumina 33.23 1.47 Ferric oxid i .08 0.45 Lime 0.77 Magnesia 0.23 Alkalies 2.24 Titanic oxid i.io Combined water I7-5& 100.95 10.91 RATIONAL ANALYSIS. Per cent. Mineral and quartz sand 10.91 Clay substance 90.04 100.95 PERCENTAGE COMPOSITION OF THE Clay Substance and Halloysite. Per cent. Per cent. Silica 39.70 40.74 Alumina 35.27 34.86 Ferric oxid 0.70 Lime . 0.86 Magnesia 0.26 Alkalies 2.49 Titanic acid r 1.22 Combined water 19-50 24.40 Secondary or plastic clays, known to potters as pipe- or ball-clays, occur plentifully, and of a high degree of purity in a number of the tertiary and quarternary expo- sures of the country ; New Jersey, Florida, western Ken- 100 THE CHEMISTRY OF POTTERY. tucky and eastern Missouri furnishing the principal sup- plies. They are necessary additions to white- ware bodies, because of their plasticity. Being more abundant than kaolins, they are cheaper than these, but are also, even in the best varieties, far less white. Plastic bodies being in reasonable limits easier and hence cheaper to mold, ball-clays are used in as large amounts, for their required clay substance, as the quality of the ware will admit. From their formation these clays are very likely to con- tain finely divided quartz and the detritus of fusible min- erals, hence their introduction into a body or substitu- tion for some other ball- or china-clay in a pottery recipe, should only be made with a clear knowledge of their chemical composition. It is, however, a striking fact, that a much larger pro- portion of the American ball-clays approach the compo- sition of kaolinite, than do those of Europe. The purest of our native ball-clays are mined in Florida, and are being sold as ' ' plastic kaolins, ' ' a misleading trade name. An analysis of one of these runs as follows : Per cent. Silica 45-39 Alumina 39. 19 Ferric oxid 0.45 Lime 1 0.51 Magnesia 0.29 Alkalies 0.83 Combined water 14.01 100.67 Insoluble in H. 2 SO 4 and Na 2 CO 3 0.87 RAW MATERIALS OF wtfiT KARE ibl A sample baked at the melting heat of orthoclase, while very nearly as white as a high grade kaolin, had not at all remained porous, but was dense and glossy, having shrunk fully fifteen per cent, in linear diameter. Next in purity to these Florida clays are those very extensively mined in New Jersey. An analysis of a typical one of these clays is the following : Portion insoluble Composition of the in H 2 SO 4 and entire clay Na 2 CO 3 Per cent. Per cent. Silica ..................... 46.18 3-20 Alumina .................. 39.08 0.39 Ferric oxid ............... i.n Lirne ..................... 0.42 0.05 Magnesia ................. 0.35 0.05 Potash .................... 0.23 0.05 Soda ...................... 0.28 0.21 Combined water .......... 13.04 o.oo 100.69 3.95 Burned at the melting heat of orthoclase the clay was yellowish white and shrank fourteen per cent. The clay is of good plasticity and requires 51.5 grams of water for 100 grams of dry clay, in order to be pene- trated by the Vicat needle. However, its binding power is very low. Briquettes formed of the plastic clay would not, in the dry condi- tion, bear a strain of more than 1600 grams per square centimeter. The ball-clays of Western Kentucky and Eastern Missouri, while less pure than the foregoing, are of far greater binding power, giving them, for certain pur- OF POTTERY. poses, much the preference over the former. In fact, be- cause of the deficiency of the New Jersey clays, in this particular, not inconsiderable quantities of English plas- tic clays, that could well be replaced by these western domestic ones, are imported. A ball clay from Jefferson County, Missouri, of estab- lished reputation, has the composition : Portions insoluble in H 2 SO 4 and The entire clay Na 3 CO 3 Per cent. Per cent. Silica 48.5 1 2.85 Alumina 35-i8 0.75 Ferric oxid 0.92 Lime i .01 0.06 Magnesia 1.47 0.48 Alkalies 2.30 0.35 Combined water 10.72 o.oo ioo.ii 4.49 Burned at the melting of pyrometric cone nine, the shrinkage of the clay is fifteen percent., and the body is not only dense from the high contraction and the charac- teristic structure of the plastic clays, but is vitrified as a direct result of the high percentage of fluxing oxides it contains. A plastic clay brought into the Market from Galloway County, Kentucky, gave the following analysis : The portion insoluble in H 2 SO 4 and The entire clay Na 2 CO 3 Percent. Per cent. Silica 59.83 26.02 Alumina 27.80 0.29 Ferric oxid 0.83 0.07 RAW MATERIALS OF WHITE-WARE BODIES. 103 The portion insoluble in H 2 SO 4 and The entire clay Na 2 CO 3 Per cent. Per cent. Lime 0.15 0.05 Magnesia 0.24 trace Alkalies 0.82 0.46 Combined water 10.42 o.oo 100.09 26.89 RATIONAL, ANALYSIS. Per cent. Clay substance 73.20 Feldspathic detritus 1 .92 Quartz 24.97 100.09 PERCENTAGE COMPOSITION OF THE CI,AY SUBSTANCE. Per cent. Silica 46. 19 Alumina 37-59 Ferric oxid i .03 Lime 0.13 Magnesia 0.32 Alkalies 0.49 Combined water 14.25 At the melting heat of feldspar the clay gives a dense body, but having still some suction when applied to the moist tongue ; it is of a yellowish- white color and has a shrinkage of ten per cent. With the exception of the Florida and some of the New Jersey clays, ball clays are put on the market un- washed. They almost invariably contain iron pyrites, some- times in large nodules. 104 THE CHEMISTRY OF POTTKRY. Frequently the more plastic English and Kentucky clays contain sufficient organic matter to appear quite black in their raw state. This may be determined analytically with sufficient accuracy, by dissolving the clay in hydrofluoric and hydrochloric acids on the water-bath, and filtering off the organic flocks on a tared paper, where, after drying, it is weighed. By this method these clays have often yielded the writer as much as four per cent, of organic impurity. Like the china-clays the ball-clays should be exam- ined comparatively for tint, working them up into bodies with pure flint and feldspar and covering the biscuit pieces with a clear glaze, in the manner already de- scribed. The uncombined silica of pottery bodies, in so far as it is not furnished by the clays in compounding the same, is added in the form of finely ground flint or quartz. French flints are cheaply imported as ship ballast and used to a considerable extent, though native flint of a high degree of purity is obtainable and also largely used. Flint is roasted previous to grinding, becoming friable through loss of its organic matter. It should burn perfectly white, though a faint pinkish cast is often not objectionable, as the feldspathic flux of white-ware bodies takes up the small amount of ferric oxid produc- ing it, giving a perfectly white sinter. A native flint of average commercial quality, put on the market by a spar-miller of Trenton, N. J., has the composition : RAW MATERIALS OF WHITE-WARE BODIES. 105 Per cent. Alumina 0.33 Ferric oxid 0.27 Lime o. 13 Magnesia 0.09 Alkalies o. 1 1 Total impurities 0.93 Moisture 0.24 Silica (by difference) 98.83 100.00 Heated to the melting heat of orthoclase, it remained snowy white and did not, in the least, sinter together. A quartz sand, of high purity, mined in L/aSalle County, Illinois, which is ground for pottery purposes, has the following composition : Per cent. Alumina o. 155 Ferric oxid 0.069 Lime 0.026 Magnesia 0.013 Alkalies o. 1 12 Total impurities 0.375 Moisture 0.070 Silica (by difference) 99-555 100.00 There are, in the country, deposits of silica contain- ing but small amounts of accompanying clay and min- eral detritus, that are very finely divided, requiring practically no grinding and burning very white indeed. Such materials are, as yet, not utilized for pottery pur- poses, though they would furnish an excellent and cheap io6 THB CHKMISTRY OF POTTERY. substitute for the artificially ground flint, allowance be- ing made in using them, for the clay and fusible min- erals which they contain. Being powdered by natural agencies, possibly by precipitation, they seem to com- bine more intimately with the clays of a mass, making often a more plastic and perfectly knit body, than is ob- tained with silica reduced by grinding. A material of this character, from along the Cumber- land river in Tennessee, analyzed : The portion in- The entire soluble in H 2 SO 4 sample and Na 2 CO 3 Per cent. Per cent. Silica 85.80 65.17 Alumina 9.75 0.50 Ferric oxid 0.46 0.04 Lime 0.20 0.08 Magnesia 0.23 0.07 Alkalies 0.98 0.35 Combined water 3.07 o.oo 100.49 66.21 RATIONAL, ANALYSIS. Per cent. Clay substance 32.74 Feldspathic detritus 2.94 Quartz 63.27 Soluble silica i .54 100.49 PERCENTAGE; COMPOSITION OF THE CLAY SUBSTANCE. Per cent. Silica 58.31 Alumina 28.25 Ferric oxid 1.28 RAW MATERIALS OF WHITE-WARE BODIES. IOy Per cent. Lime 0.36 Magnesia 0.48 Alkalies 1.92 Combined water 9.37 Baked at the melting-point of feldspar the material has a shrinkage of one and a half per cent., and when it is still porous is much denser than a mixture of a flint and ball-clay according to its rational analysis would become, as a steel point marks but does not scratch it. Still more striking is a material of this character from Galloway County, Kentucky. Quite friable in its natural state, it hardens somewhat when immersed in water, like a weak plaster. This, however, is very easily ground to a mass which no longer sets. Its analysis is : The portion in- The entire soluble in H 2 SO 4 material and Na 2 CO 3 Per cent. Per cent. Silica 90.49 77.00 Alumina 5.45 0.50 Ferric oxid 0.39 0.02 Lime 0.23 0.03 Magnesia 0.30 0.07 Alkalies 1.74 0.33 Combined water i .64 o.oo 100.24 77.95 RATIONAL ANALYSIS. Per cent. Clay substance 18.70 Feldspathic detritus 2.71 Quartz 75.24 Soluble silica 3.59 100.24 108 THE CHEMISTRY OF POTTERY. PERCENTAGE COMPOSITION OF THE CI,AY SUBSTANCE. Per cent. Silica 53.0 Alumina 26.4 Ferric oxid 2.0 Lime i.i Magnesia 1.2 Alkalies 7.5 Combined water 8.8 100.0 Burned at the heat of running spar, the material shrinks but one-half per cent, and is very white. It re- mains more porous than the material just described and can just be scratched with a knife blade. Of the fluxes used in the bodies of white ware, feld- spar is the most important. It is put upon the market by a number of spar-millers of Connecticut, New Jersey, Ohio, and Pennsylvania of satisfactory purity and of different fusibilities. A good commercial potash feldspar gave the follow- ing analysis : Per cent. Silica 65.85 Alumina J 9-32 Ferric oxid 0.24 Lime 0.56 Magnesia 0.08 Alkalies 1 14. 10 100.15 A commercial soda-lime feldspar, materially more fusible than the former, being sold as a ' ' soft spar, ' ' analyzed : 1 Combining weight, 45.9. RAW MATERIALS OF WHITE-WARE BODIES. IOQ Per cent. Silica 66.81 Alumina 21.09 Ferric oxid o. 13 Lime 2.03 Magnesia o. 10 Alkalies 1 9.64 99.80 Often the quartz veins of a spar-bed may be difficult to remove or the workmen are careless in picking over the mineral, causing more or less variation of the com- mercial product, against which the consumer must be on his guard. Again, it may occur, that the more quartzose por- tions of a soda-lime spar^are ground separately and sold as ' ' hard or potash spar, ' ' as was the case with a lot of which the following is an analysis : Per cent. Silica 68.82 Alumina 19-75 Ferric oxid o. 16 Lime i .64 Magnesia o. 1 7 Alkalies* 9.15 99.69 It is of about the same fusibility as the true potash- spar shown in the first analysis, and would deceive one in the mere kiln-fusibility test, which is all that potters usually make to verify the quality of the material. 1 Combining weight, 33.5. 2 Combining weight, 36.4. 110 THK CHEMISTRY OF POTTERY. The practical bearing of the difference in composition of these two feldspars of equal fusibilities, is demon- strated in their varying coefficients of expansion a serious difference if one were substituted for the other in a body having to carry a certain glaze. This difference is practically shown in the fact that the potash feldspar melted in a thick layer upon a pot- tery body, cracks off of one upon which the siliceous soda-lime feldspar remains immovably fixed. . Cornwall stone, a partly decomposed granite, mined in Cornwall, England, is used by English potters as their principal pottery flux and also finds considerable application in the United States. All that is used here is imported, no material resembling it having as yet been commercially developed within our borders. An average sample of a good quality of this material has the following composition : The portion in- The entire soluble in H 2 SO 4 material and Na 2 CO 3 Per cent. Per cent. Silica 73.57 57.69 Alumina 16.47 4-7 Ferric oxid 0.27 0.30 Lime 1.17 o.io Magnesia 0.21 0.12 Alkalies 5.84 3.50 Combined water 2.45 o.oo 99.98 66.41 Combining weights of the alkalies 44.6 38.4 RAW MATERIALS OF WHITE-WARE BODIES. Ill RATIONAL ANALYSIS. Per cent. Clay substance and mica 33-57 Feldspar 25.31 Quartz 4i-io 99.98 PERCENTAGE COMPOSITION OF THE Clay Substance and Mica Feldspar Per cent. Per cent. Silica 47- 2 7 6 5-55 Alumina 35-Q4 18.57 Ferric oxid o.oo i . 18 Ivime 3.18 0.40 Magnesia 0.26 0.47 Alkalies 6.96 13.83 Combined water 7.29 o.oo 100.00 100.00 The figures show that the kaolinizing decomposition of the rock has proceeded to but a limited extent, the " clay substance," as in this sample, consisting in the main of mica. This is further proven by the constant presence of fluorine, which, though it has been ignored as a separate element in the above analysis, has been found present to the extent of 1.66 per cent. It may be justified, in the case of this material, in which mica plays nearly as important a part as the feld- spar as fluxing constituent, to give it a separate place in the rational analysis, for the better guidance of the pot- ter. Cornish stone is by no means as uniform in character and composition as potters generally believe. The portion insoluble in sulfuric acid and sodium carbonate solution is in many cases markedly greater in 112 THE CHEMISTRY OF POTTERY. alumina than in that of which the analysis has been given and not infrequently the silica is either largely soluble in the sodium carbonate solution or is more readily made so by the action of sulfuric acid than quartz commonly is. A sample showing both of these peculiarities analyzed as follows : The portion in- The entire soluble in H 2 SO 4 material and Na 2 CO 3 Per cent. Per cent. Silica 72.99 4 2 '72 Alumina 17*58 7.83 Ferric oxid 0.15 o.io Lime 1.25 0.71 Magnesia 0.37 0.19 Alkalies 6.20 4.31 Combined water 1.77 o.oo 100.31 55.86 RATIONAL, ANALYSIS. Per cent. Clay substance, mica, and soluble silica 44-45 Feldspar 40.68 Quartz 15.18 100.31 PERCENTAGE COMPOSITION OF THE Clay Substance Etcetera Feldspar Per cent. Per cent. Silica 68. 10 67.68 Alumina 21.94 19-24 Ferric oxid o.ii 0.25 Lime 1.21 1.75 Magnesia 0.41 0.47 Alkalies 4. 25 10.61 Combined water 3.98 o.oo RAW MATERIALS OF WHITE-WARE BODIES. 113 The sum of the alkali and combined water in this * ' clay substance ' ' falls far short of what would be de- manded by a mixture of mica and pure clay ; while on sub- tracting the excess of silica, assuming it as uncombined but soluble in sodium carbonate solution, and recalcu- lating the residue on a percentage basis, they assume the proper proportion. This would make the rational analysis : Per cent Clay substance and mica 25. 71 Feldspar 40.68 Quartz 15. iS Soluble silica 100.31 The percentage composition of the clay substance and mica then is as follows : Per cent. Silica 44.85 Alumina 37-93 Ferric oxid o. 19 L/ime 2. 10 Magnesia 0.70 Alkalies 7.35 Combined water 6.88 Direct determinations of soluble silica in a number of such specimens failed to yield anything like the required amount, leading to the conclusion that sulfuric acid may in many cases have more action on some form of silica, in this mineral, as well as fjn other clays, than that already pointed out in the case of quartz, rendering it in greater measure soluble in sodium carbonate solution. 114 THE CHEMISTRY OF POTTERY. But more important than the difference in the physical character of the contained minerals or a variation in the apportionment of the elements to the different mineral groups, is the variation in ultimate chemical composi- tion of Cornish stone, particularly in the proportion of the alkalies, as in the following : Per cent. Per cent. Silica 74.55 73.77 Alumina 17-37 16.05 Ferric oxid 0.26 0.23 Lime 1.68 1.14 Magnesia 0.54 0.22 Alkalies 3.68 7.52 Combined water 2.04 . 1.78 100.12 100.71 While Cornish stone has distinct uses determined by its physical properties, it could, in the vast majority of the cases of its application in the United States, be more cheaply and safely substituted by equivalent amounts of native feldspar, kaolin, and quartz, which run much more uniform in quality. An anomalous material, white, of fine grain, free from crystals of quartz and feldspar, and flakes of mica, resembling a kaolin in appearance, was obtained by the writer from Fayette County, Texas, where it is said to occur in large deposit. It resembles Cornish stone in composition more nearly than any native material which has thus far fallen into the writer's hands. Its analysis is as follows : HAW MATERIALS OF WHITE-WARE BODIES. 115 The entire substance Per cent. The portion in- soluble in H 2 SO 4 and Na 2 C0 3 Per cent. Silica 68.88 43.60 Alumina 16.77 7-9 1 Ferric oxid 0.83 0.32 Ivime 0.99 0.26 Magnesia 0.17 0.18 Alkalies 6.77 2.56 Combined water 5.79 o.oo Sulfuric anhydrid- 0.42 0,00 100.62 54.76 Soluble silica (directly determined) 5,23 RATIONAL, ANALYSIS. Per cent- Clay substance 39.83 Feldspathic detritus 40. 28 Quartz 14.55 Soluble silica 5.23 Calcium sulfate 0.74 100.63 PERCENTAGE COMPOSITION OF THE CI^AY SUBSTANCE. Per cent Silica 50.33 Alumina 23.52 Ivime 1.03 Alkalies - 10.57 Combined water 14.54 The combined water, in this case, seems to belong largely to the soluble silica, which was hydrated so that a portion of it could be extracted with distilled water Il6 THE CHEMISTRY OF POTTERY. alone ; but as there was no way of apportioning it, it was counted in with the " clay substance." This material was substituted for Cornish stone, both in body and glaze trials, with entirely concordant re- sults. Calcium carbonate, which is much used by the continental potters, finds but very limited application with us as a body-flux. Commercial whiting is a pure form of this material, as shown in its analysis in a previous chapter, though a higher grade is put upon the market for the use of potters, under the name " Paris white/' Should the manufacture of a ware covered with tin enamels, make a body containing a large proportion of calcium carbonate necessary, such as that used for mak- ing the tile of the German and Swiss "Kacheloefen," a cheap and reasonably pure form of this material would be found in the fresh water marls of northern Ohio and Indiana, some of which are now exploited in the produc- tion of Portland cement. Samples of these were found of the following compo- sition : Per cent. Per cent. Silica 1.16 10.45 Alumina trace 4.20 Ferric oxid 0.05 trace Calcium carbonate 94-7$ 79-35 Magnesium carbonate 0.19 trace Organic matter and loss. . 3.82 6.00 100.00 100.00 CHAPTER X. WHITE GRANITE AND CREAM-COLORED WARE. HE bulk of the dishes used for table ser- vice and all of those for the toilet and for modern sanitary plumbing, belong to the category of "Granite and C. C." ware. White graniteware or ironstone china differs only from " C. C." that is cream- colored ware and ivory ware, in being made with a larger proportion and better quality of kaolin, so as to be as white in color as possible ; the whiteness being generally heightened by the neutralizing of any faintly yellowish cast, through the addition of cobalt-blue to the body and covering the same with glazes in which, among the basic constituents, the alkalies and alkaline earths at least equal or exceed the equivalent of lead. Beyond this, there is no technical distinction between the wares repre- sented by these and some other trade names. Naturally those which are made with less and cheaper kaolin, having a larger proportion of plastic clay, are more easily fashioned, and when covered with a more plumbiferous and hence more fusible glaze, are commoner and cheaper. The bodies are made mainly from the materials dis- cussed in the previous chapter, their general composition not varying widely from the following : Il8 THE CHEMISTRY OF POTTERY, 50 to 60 per cent, clay substance, 38 to 32 per cent, quartz, 12 to 8 per cent, feldspar. Where Cornwall-stone is used as the flux, its propor- tion is, of course, larger than that of the feldspar given above and is about such that its mica and feldspathic mineral equal in amount the required feldspar, the addi- tional quartz and clay substance replacing their weight of these in the formula. The clay substance is made up of kaolin and plastic clay. In the whiter varieties there may be two-thirds of the former to one-third of the latter, while in the more tinted bodies the proportions may be reversed ; but in this there is no rule technically imperative, for the tint of body desired, the relative plasticity of the clays, and cost thereof determine the mixture. If, as is frequently the case, a plastic clay be used, bearing a considerable proportion of quartz, feldspathic mineral, or a clay substance rich in alkalies and alkaline earths, the amounts of added quartz and flux are propor- tionately less in the body made with it. The bodies in use have grown up from the most hap- hazard of empirical trials, but a recalculation of the mix- tures, with reference to the composition of the ingre- dients, will show most of them to fall within the limits of the general formula given. If this be used as the starting point, comparatively few empiric trials would have to be made to attain a body of any desired charac- ter, with materials of known composition. The materials for a body are weighed off in the propor- tion required, and mixed in the wet way, in the ' 'blunger" WHITE GRANITE AND CREAM-COLORED WARE. 119 already referred to in the washing of yellow- ware clays. Errors in the proportions of the constituents entering into the body often occur from the fact that potters , as a rule , are not accustomed to determine and allow for the vary- ing amounts of moisture, which the materials may con- tain. Thus the powdered flint or quartz will contain from one to four per cent, feldspar, from two to five per cent, china-clays, and Cornish-stone from two to ten per cent., and the various plastic or ball-clays will seldom contain less than ten per cent., and they may often contain as much as twenty-five per cent, uncombined water. Some of the highly plastic English ball-clays, as also those from western Kentucky, may show in different ship- ments considerable variation in the amount of contained carbonaceous matter. This should be determined and allowed for, as well as any moisture. Systematic mois- ture determinations and recalculations of formulas in ac- cord with the results would easily obviate annoying va- riations in the bodies. After the mixture has been stirred or "blunged" to a homogeneous slip, this is passed through a revolving, shaking or vibrating sieve covered with a No. 12 silk bolting-cloth or a wire cloth having one hundred and twenty meshes to the linear inch. By this means, such accidental impurities as chips, grain and cinders, which get into the materials from the railway cars, are removed, as also sand and the particularly troublesome small gran- ules of iron pyrites of the ball-clays. Here, however, is also a point where the composition of the body may materially change. If the slip be too 120 THE CHEMISTRY OF POTTERY. thick, a considerable quantity of the plastic clay will be taken out of the mixture ; but the most important dan- ger arises from insufficient grinding of the fluxes, an ap- preciable percentage of which may not be fine enough to pass the sieve. This is particularly the case since the introduction of dry- in place of wet-grinding by the spar- millers r and is most to be looked for in the case of Corn- ish-stone, the tough and flexible crystals of mica, of which resist reduction in pulverizing cylinders much more than they do the crushing and tearing action of the buhr- stones in the wet-drag mills. Cornish-stone has been found on the market leaving as much as fifty per cent, residue on a No. 12 silk bolting cloth,, and it is seldom so fine as not to leave five per cent. As already stated, it is customary, in the case of the whiter bodies, to add a small amount of cobalt, to neu- tralize any faintly yellowish cast, which they may show. This is usually added in the form of " blue calx," a kind of Thenard's blue, made by roasting a mixture of cobalt oxid with a china-clay and quartz and grinding the re- sulting mass to great fineness in water. As the most productive whitening effect and the greatest homogeneity of color is obtained by the most perfect division and distribution of the cobalt, it is bet- ter to effect this by chemical precipitation than by me- chanical grinding. For this purpose, all that need be done, is to add to the slip a solution of the necessary quantity of a soluble cobalt salt. The natural alkalinity of most waters will be sufficient to precipitate the cobalt perfectly, lodging it on every particle of the charge in WHITE GRANITE AND CREAM-COLORED WARE. 121 even distribution without spots or specks in the finished body. Formerly, the useless precaution was taken to pass the slip through a trough in which a row of magnets was suspended, with the idea that iron was to be re- moved from the clay in this manner ; this belief is now practically past and the prejudice against the more con- venient wrought-iron blunging tubs and cast iron filter- presses is also fast disappearing. The bod)^ mixed, sifted, and thickened, is fashioned into ware, which is dried and placed in the kiln for the biscuit fire. The temperature at which this is finished ranges from the melting of pyrometric cones eight to ten ; the em- piric trials for judging the fire being biscuit rings coated with Albany slip-clay and with feldspar. The successive changes in appearance of the former material indicate to the experienced eye, the progress of the fire, while the melting of the feldspar determines the finish of the burn. The properly burned biscuit-ware should be of such hardness that it is not possible to scratch it with a steel point ; yet it must be of uniform porosity, it being par- ticularly undesirable for the thin edges and raised points of the modeling on the ware to be so dense as not to ad- here when touched to the tongue. A moderately and uniformly porous body, on being immersed in the creamy liquid of the glaze materials, ground and suspended in water, absorbs the latter, which deposits a uniform coat of its suspended solids on 122 THE CHEMISTRY OF POTTERY. the surface of the piece. Where, however, this is too dense to take the water up, its solids are washed off, on the withdrawal of the piece from the dipping tub, and after passing through the glost fire these parts appear dry and rough. The best glazes, for these wares, approach a formula of the following character : 0.25 KNaO) 0.50 CaO 0.3 A1A 0.25 PbO ) i.o RO 3.5 A glost-kiln heat approaching the melting-point of gold is usually given, at which the glaze not only runs perfect- ly bright and smooth but eats itself also slightly into the surface of the body, taking up additionally from the lat- ter a small amount of silica and alumina. Thus thor- oughly burned, the glaze should have the coefficient of expansion of a body of a formula lying within the limits of that already given and burned to proper bis- cuit-hardness. It is also sufficiently bright and of such hardness as not to be scratched by the common table cutlery. Deviations from the formula given,, in the direction of an increase of lead oxid at expense of the lime, and of boracic acid at expense of the silica, and lessening of the alumina and silica occur in many degrees, resulting in relatively more brilliant and easily fusible glazes, but these are also proportionately less white, more easily scratched, and of less range between the crazing and shivering points. WHITE GRANITE AND CREAM-COI/)RED WARE. 123 Brilliance is however a consideration in the sale of ware and most glazes will be found rather higher in lead and boracic acid than that of the formula given. This glaze can, of course, be made in a variety of ways, according to the materials at hand, though the following will serve as a simple example. Because of the alkali and boracic acid a frit rendering these insoluble must first be made. This is simply done by melting together 0.25 equivalent of powdered borax, 47.8 parts ; 0.5 " " boracic acid, 31.0 " 0.25 " " feldspar, 69.6 " 0.5 " " whiting, 25.0 " 0.5 " " quartz, 15.0 " the chemical formula of this frit being o.25Na 2 O 2 o 0.25 K 2 o.2 5 Al 2 3 ' 0.5 CaO i.o RO and its combining weight 141.4. For the glaze it is then necessary to grind together 0.5 equivalent of the frit, 70.7 parts. 0.25 " " whiting, 12.5 " 0.25 " " white lead, 32.9 " 0.175 " " china-clay,' 22.7 " 1.65 equivalents " quartz 49.5 " As in the mechanical application of the glaze mixture in a uniform layer to the surface of the ware by dipping the biscuit pieces in the glaze, it is important that its constituents, which differ widely in specific gravity, must remain uniformly mixed and suspended in the 124 TH E CHEMISTRY OF POTTERY. water for a considerable time, the proportion of clay in the glaze mixture, which helps this suspension, should not be too small. It is best, therefore, not to introduce any clay into the frit, but to reserve it unburned for the glaze mixture itself. Where the amount of alumina required by the glaze is not large it is often best not to introduce feldspar, in order to have all the alumina available for introduction as clay, because of its mechanical use in keeping the glaze materials afloat. The frit is often melted in saggars in the glaze-kiln. In order to prevent its sticking to the same, it is com- mon to pour out the inner surface of the saggar with a thick milk of flint and water, leaving on the bottom and the walls a coating of flint from one-eighth to a quarter of an inch thick. Into the saggar thus coated the frit mixture is tightly packed. After baking in the glost- fire the saggar is easily broken away from the lump of glass which it contains, and the excess of flint adhering to it is chiseled off. Any remaining too firm for removal is generally too insignificant in amount to affect the composition of the product. Previous to its introduction in a mill for fine grinding, the frit should be crushed small on a buhrpaii under buhrstone runners. Where it is necessary to do this in an iron mortar or between the jaws of a rock- or ore- breaker, the crushed glass must be well freed from the not inconsiderable amounts of fine metallic iron, with which it becomes contaminated, with a large magnet well cleaned of rust. WHITE GRANITE AND CREAM-COLORED WARE. 125 The mills used for grinding the frit fine are buhrstone drag-mills and porcelain-lined cylinders filled, with Ice- land flint pebbles. The grinding is done in water. Far better than melting the frit in saggars, is to melt it on the sloping hearth of a special furnace, from which it can be run through a hole at its lowest point. By this means the melting is much more thorough. The frit is removed from further action of the fire as soon as it is liquid enough to run from the kiln and it is at once cooled and broken so fine by falling into water that it can be introduced into the fine mills at once. The gradual and long fire of the glost-kiln is very lia- ble indeed to volatilize alkali and boracic acid from the mixture before they are thoroughly combined into glass, and, furthermore, the slow cooling of the kiln so tem- pers the latter when melted in saggars, that it is much more difficult to grind than when suddenly quenched in water on flowing from a frit-kiln. The various materials enumerated in the preceding chapter all enter into glaze formulas. There the com- position and properties of flint, feldspar, Cornish-stone, whiting, and the clays have been sufficiently discussed to throw the necessary light on their use in this connection. The various lead preparations used in glaze making, namely white and red. lead, and litharge, are easily ob- tainable in the market in a high state of purity. The testing of these is also not difficult, and can be found in books on technical analysis. It remains merely to call attention to a superstition among potters, created by commercial rivalry among white-lead manufacturers, that traces of acetates in white 126 THE CHEMISTRY OF POTTERY. lead are of deleterious influence on the glaze. A real dan- ger lies in the mistaken or ignorant use of lead sulfate in the place of white lead. This sulfate is now produced by direct oxidation of galena and put upon the market under names tending to confound it with the former product. Borax is obtainable of sufficient purity, though it is usually well to examine it for the possible presence of sodium sulfate. Boracic acid may contain ammonium sulfate and prove troublesome by introducing sulfates into the frit. As much as 10.28 per cent. SO 8 has been found by the wri- ter in the light-brown flaky variety. Soda is more conveniently employed in the form of soda-ash, as the large amount of water of crystallization in sal-soda is troublesome in the melting of a frit con- taining it, and furthermore the efflorescence of the salt, when standing in open barrels is liable to cause serious changes in the composition of mixtures made with it. It is important to examine all soda-ash for the pres- ence of sulfate. That made by the Solvay process, and now becoming common in the market, is free from such contamination however. Potash, where its introduction as potash-feldspar is not admissable, because of the amount of alumina or silica necessarily accompanying it, is used as pearl-ash, or where great purity is required as bicarbonate or nitrate. On account of its variable composition pearl-ash should always be subjected to chemical analysis previous to use and because of its hygroscopic character, its moisture should be determined and allowed for whenever it is applied. CHAPTER XI. MAJOLICA AND ENAMELED TILE. ITH exception of the printing and hand paint- ing in vitrifiable colors and gold, and, to a more limited extent, in underglaze colors, on the useful articles of manufacture coming un- der the head of the kinds of ware described in the last chapter, the principal commercial development of ornamental pottery in the United States has been in the use of transparent colored glazes on modeled surfaces. In so-called majolica, of which many articles of an or- namental and semi-useful nature are made, a decorative effect is sought to be attained by the application of dif- ferently colored glazes to different parts of the piece. The colored glazes are put on thin, and the ware is gen- erally very imperfect technically, the glazes being mi- nutely crazed and the body very porous and brittle. In enameled tile the individual pieces, modeled as a rule in bolder relief than the ornament of the former, are colored in monochrome, through a very heavily glazed surface. The decorative effect being attained by the variation in depth of the glaze through the approach to or recess of the contour of the design from the surface of the glass, lighter and darker tints of the color appear and bring out the design in light and shade. To a certain extent also, mottled colors in imitation 128 THE CHEMISTRY OF POTTERY. of marbles are used, but in all cases the decorative effect sought is dependent on the thickness of the glaze, which in such measure refracts the light and makes the glaze brilliant. This ware, manufactured and used very extensively for interior decoration in kitchens, halls, bath rooms, about hearths and mantel pieces, is artistically as well as technically much better than that previously men- tioned ; but the heavy glazes employed to make it dec- orative, offend good taste as well as make it practically impossible in manufacture to steer the product uniformly perfect through the "Scylla and Charybdis" of crazing and shivering. The body and glazes of both of these kinds of ware are very much the same. The former coincides in composi- tion with the body of cream-colored wares, but is as a rule not so hard, because the manufacturers of ornamen- tal articles are more careless in the giving of a thorough biscuit fire, than those who are faced with the conditions imposed on useful products, Tile, which must burn perfectly level and true in size and shape, are not formed from the mass in a plastic state, but this is dried and reduced to a powder contain- ing from eight to ten per cent, moisture, which is pressed into shape in metal dies with powerful screw, cam, or hydraulic presses. By this means, bodies are obtained that are little liable to warp or become otherwise untrue in drying and burn- ing, but the clay particles not being nearly so closely bonded as in drying down from a plastic condition, the MAJOUCA AND ENAMELED TILE. 1 29 larger pieces are with many clays subject to cracking or " dunting" as it is called. In addition to the other con- ditions that obtain in mixing white pottery bodies, this danger must be kept in mind in selecting a body suita- ble for tile. The clear glazes used as the bases of the colored ones, for these -industries, are both raw and fritted. The for- mer resemble closely those used in yellow ware; the lat- ter must be more fusible and brilliant than those em- ployed for white granite and cream- colored wares and contain, therefore, more lead oxid and boric anhydride and rather less silica and alumina. However, the tints imparted by the chromogenic oxlds to glazes are so profoundly influenced by their chemical composition and the relative proportions of their basic and acid oxids, that the widest divergence in composi- tion must be looked for. It is only possible, therefore, to give for illustration a common type of a good glaze of this class, as the follow- ing : 0.25 Na.Cn ( 2 c S iO 0.25 CaO k 2A1 'loifx> 0.50 Pbo y (- 5 ^ u * i.o RO It may be prepared by first melting a frit of 0.5 equivalent borax, 95.5 parts. 0.5 whiting, 25.0 " 2.0 " flint, 60.0 " its formula being 0.5 Na 2 O ) 2.0 SiO 2 0.5 CaO )i.oB 2 O 3 130 THE CHEMISTRY OF POTTERY. and its combining weight 124.5. * n order to make the glaze it is then necessary to grind together 0.5 equivalent of the frit, 62.25 parts. 0.5 " " white lead, 64.85 " 0.2 " " china clay, 25.9 " i.i equivalents " flint, 33.0 " For coloring the glazes, cobalt, nickel,' copper, chrome, manganese, uranium, and iron oxids are in use. It is customary to grind these with quartz and china- clay previous to their addition to the glaze, in order to insure their perfect division, that they may dissolve in th glazes without producing spots of deeper tint, but produce a uniformly colored glass. As a rule there is no proportion of these admixtures bearing any rational relation to the glazes that is observed ; in fact, the only thought connected .with this addition, beside that of divi- sion of the coloring oxid, is cutting down its tinctorial power, so that those of higher coloring property receive the larger additions. As a result, the addition of these colors changes the formula of the glaze to a greater or less degree. It would, therefore, be a more rational proceeding to add to the various oxids, in grinding the corresponding colors, the quartz and china-clay in such proportion as to give compounds of the same acidity as the glaze in which they ar,e to be used. Any desired proportion could then be added to the glaze without ma- terially disturbing the relations of its chemical formula, which is of some moment both with reference to the fusi- bility and the coefficient of expansion. Such mix- tures, for the above glaze, would be composed as follows : MAJOLICA AND ENAMELED TILE. 131 For cobalt color i.o equivalent cobalt oxid (Co 3 O 4 ), 40.0 parts. 0.2 " china-clay, 25.9 " 2.6 equivalents flint, 78.0 " and for the other colors, like amounts of clay and flint, with the following amounts, respectively, of the oxids : 37.4 nickel oxid (NiO). 39.8 copper oxid (CuO). 38.3 chromic oxid (Cr 2 O 3 ). 38.2 manganese oxid (Mn 3 O 4 ). 40.0 ferric oxid (Fe 2 O 3 ). 70.7 uranium oxid (U 3 O 4 ). A better method than this, insuring the mo'st perfect division of the coloring oxid and no disarrangement of the chemical formula of the glaze by its addition, is to substitute the oxid for one-half of the basic equivalents of the glaze formula and melt the resulting mixture. A series of colored frits, each containing one of the color- ing oxids in definite chemical proportion can thus be made, which may be added in any desired measure, to the clear glaze, in order to produce the various colors. The formula of these colored frits would then be, for the clear glaze given 0.125 Na 2 O) C 0.125 CaO o.2Al 2 9 ]oc 0.25 PbO ) (-,5. 0.50 RO The different coloring oxids take the place of RO. Thus, the copper frit would be made by melting together the mixture : 132 THE CHEMISTRY OF POTTERY. 0.125 equivalent powdered borax, 23.9 parts. 0.125 whiting, 6.25 " 0.25 " white lead, 32.4 " 0.5 copper oxid, 19.9 " 2.1 equivalents flint, 63.0 " 0.25 equivalent boracic acid, 15.5 " This would yield 135.7 parts of glass, that are then ground with 0.2 equivalent of china clay, 25.9 parts. In the other frits the copper is replaced by the equivalent weights of the respective coloring oxid. As these frits are used in small amounts, melting them on the hearth of the frit-kiln would not be practical. The advantages of the frit-kiln on a small scale can be at- tained, however, by melting in large crucibles, pierced in the bottom with a hole three-eighths of an inch in di- ameter. As many of these as frits are desired to be melted at one time, are placed side by side on the flat hearth of a furnace having a hole two and one-half inches in diameter under each crucible. Into each of these, which must be easily reached and filled from above through corresponding openings in the crown of the fur- nace, its frit-mixture is introduced, which on becoming sufficiently fluid, runs from the hole in the bottom, falling through the opening in the hearth, into a suitable recep- tacle filled with water below. The principal raw materials of these glazes, with ex- ception of the coloring oxids, have been previously dis- cussed. Cobalt is commonly met with in the market as black oxid, prepared for use in Saxony and Wales. All commercial preparations contain nickel oxid in larger or smaller proportion, as the impurity affecting the tint of MAJOLICA AND ENAMELKD TII^. 133 the glass in which the oxids are used. The amount of nickel oxid contained in the Saxon brands R. K. O. and F. K. O. is from five to six per cent. ; in G. K. O. there is from two to three per cent., and in F. F. K. O. one- half per cent. The Welsh oxid, seen in this market, generally con- tains about five per cent, nickel oxid. Nickel oxid itself is not used very extensively from the fact that in the more convenient glazes, which are rich in lead oxid and bo- racic acid, it is very liable to cause turbid loam-colored separations, and to shade badly from brownish to green- ish tints. It is, however, serviceable in the more alka- line glazes and those poor in boracic acid. Nickel oxid is obtainable quite pure of domestic man- ufacture. Copper oxid in the form of copper scale or black oxid, is of such varying origin as found in the market, that it is advisable to always subject it to examination previous to use. As only glasses containing lead oxide in at least some proportion are in use, and the glost-fire is of neces- sity oxidizing only the green and blue-green tints of the oxid are known in our domestic pottery. The sub-oxid supplied to glass makers for ruby glass, finds no appli- cation. Chromic oxid is very commonly made by the potters themselves, by roasting an intimate mixture of potassium bichromate and sulfur in the top-kiln or flue of the glost- oven, and grinding and washing the product until free from alkaline sulfid. Used in larger proportion, chromic oxid produces intransparent green enamels, and even in smaller amounts it is liable to separate from the clear 134 THE CHEMISTRY OF POTTERY. greenish-yellow glass in deep green intransparent flecks. It is frequently the custom, therefore, because of the bet- ter division of the oxid, to introduce it in the form of mercuric chromate or lead chromate. The latter should never be used without previous examination, as it is manufactured mainly as a pigment, and as such is fre- quently adulterated to render it lighter in tint. Manganese is an important chromogen for pottery gla- zes and is used in several forms. Where brown tints are desired, the presence of iron oxid is not only not objec- tionable, but is needed, the amount depending upon the required tint. The most common form in which man- ganese finds application is as black oxid or pyrolusite. It is sold in a variety of grades, varying considerably in the amount of contained sandy matter and ferric oxid, so that where exact and economical work is contemplated, it should always be analyzed before use. Where yellow- ish-brown tints are required Turkey umber is often em- ployed, though its use is neither economical nor rational, inasmuch as the color imparted by this substance is due to the contained manganous and ferrous oxids. It would be better and cheaper to add these substances in approx- imately pure form, as commercial umber varies consid- erably in the percentage of these present. Nor is it graded, commercially, so that one could form a rough idea of the proportion of its important constituents ; for being mainly used as a pigment, physical rather than chemical properties determine its grade. Umber usually contains from sixteen to twenty per cent, of manganous oxid and from thirty to thirty-eight per cent, ferric oxid. It generally contains also quite MAJOLICA AND ENAMELED TILE. 135 an appreciable amount of gypsum, which is objectionable on account of the introduction of sulfates into the glaze. Where the reddish and wine-colored tints of manga- nese are desired, it is essential to use preparations en- tirely free from all traces of iron. In this case it is com- mon to use manganous carbonate. It is important, how- ever, not to trust commercial preparations, which often contain as much iron as common pyrolusite. Not infre- quently the commercial carbonate contains considerable amounts of calcium carbonate, not intentionally added, but derived probably from a recovered manganese. The following analyses are of preparations of this char- acter : Per cent. Per cent. Sandy matter o.io ^Z- 1 ! Alumina o. 1 1 Manganous carbonate 52.15 63.17 Calcium carbonate 47.07 21.55 Magnesium carbonate 0.57 2.11 100.00 100.00 As pure manganous carbonate is by no means a stable product, but loses moisture and carbonic acid on expos- ure to the air, it should be assayed from time to time, or it should be converted, by heating in the presence of air, into mangano-manganic oxid and weighed as such. Uranium is offered in the market as oxid, ranging in color from a bright canary to an orange-yellow. The preparation is really a hydrate, and a difference of five per cent, in the amount of loss on glowing different specimens under similar conditions has been found by the writer. The colors obtained with uranium vary with the composition of the glaze from greenish lemon- 136 THE CHEMISTRY OF POTTERY. yellow to orange. The black color imparted to certain fluxes, used in the verifiable colors of overglaze paint- ing, is not producible in the glazes and under the firing conditions of pottery. Pitch-blende, used by European potters, is not known to those of the United States. Ferric oxid is a useful chromogen, giving orange and brownish-yellow tints. It is readily obtained of suffi- cient purity, but being mainly prepared for the market as a pigment or polishing powder, physical properties determine its commercial grading and the higher priced samples are often the poorest for the potter's use. Fer- ric oxid should never be used, therefore, without pre- vious chemical examination. Specimens prepared directly from hematite and other iron ores generally contain more or less sandy matter ; those made by roasting copperas often contain injurious residues of sulfuric acid. The crocus powders to which potters are particularly partial, as most English pottery receipts call for " cro- cus martis" in place of ferric oxid, are treacherous be- cause of the foreign additions they almost invariably contain, which have been added to brighten the color or improve their character as polishing agents. A ' ' crocus martis' ' sold to a pottery was found by the writer to con- tain : Per cent. Ferric oxid 47-14 Alumina 2.31 Silica 13-65 Barium sulfate 37-4* 100.51 MAJOLICA AND KNAMKLED TILK. 137 Underglaze crimson is used instead of copper and gold for pink and red glazes, the colors being much more easily obtained, though far inferior in quality to those producible with the latter agents. The color consists mainly of stannic oxid, containing a trace of chrome in minute division. The analysis of a good quality of the color of English manufacture ran as follows : Per cent. Silica 10.02 Alumina 1.57 Chromic oxid strong trace Lime 20.32 Carbon dioxid 1 .00 Stannic oxid (by difference) 67.09 100.00 Colors with increased lime and a reduction of silica impart a disagreeable purplish cast to the glaze. The composition of the latter also affects the tint considera- bly ; boracic acid, particularly, can only be used in small proportion. Stannic oxid is used to convert the clear glazes into white intransparent enamels. As its effect depends upon its being insoluble in the glass, and clouding it by re- maining suspended in it, its physical character is all im- portant. Many commercial preparations of the oxid are much too dense, collecting in little concretions in the glass and giving it a curdy appearance without much covering power. To insure a perfect division of the tin oxid in the glaze it is best to introduce it in the form of a ' ' putty- 138 THE CHEMISTRY OF POTTERY. powder," with the lead oxid. This is done by first pre- paring an alloy of the metallic lead and tin and carefully oxidizing it in a flat pan or on the open hearth of a fur- nace, with plentiful excess of air, raking off the oxid from the surface of the molten metal as it forms. The oxid is then ground in water for a short time in a tumbler mill and floated from the flattened grains of unoxidized metal. The dried powder must be assayed and its addition to the glaze regulated by its proportions of the respective oxids. CHAPTER XII. WHITE ENAMELED BRICK. HE development of fire-proof building has made a permanently white covering for walls, imperative as a substitute for lath and plaster. The light-conditions of our large buildings in crowded city quarters, demand that this covering be highly reflecting, easily cleaned, and as permanent as the structure itself. Answering these and many minor conditions, brick with one or two faces covered with a white glazed or enameled surface are being supplied to meet the demand for such walls for ware-cellars and subways, the halls and corridors of public buildings and railway depots, the operating rooms of hospitals and light-shafts and light walls of the narrow courts of office buildings. The technical condition of the manufacture of these brick, namely the melting of a solid enameled surface on a bulky ware of coarse clay, together with the cost of fre- quent and careful handling, has presented such difficulties that the manufacture has not kept pace with the demand ; hence a large portion of the supply comes to us from abroad. The methods of making these brick are two. The one consists in covering the surface to be exposed, with a tin enamel of sufficient covering power to present a 140 THE CHEMISTRY OF POTTERY. smooth white surface completely concealing the clay un- derneath; the other, in coating the face to be exposed, with an engobe or "slip" of china-clay or of a body re- sembling that of white ware and melting a transparent glaze over this. Where the character of the clay is such that it can be burned at the same heat as the glaze or enamel it will bear, the aim is to finish the piece with its coating in the clay state and subject the product to but one fire. Tin enameled brick are best made on the plan of the white stove tile used on the continent of Europe. For these, clays rich in the carbonates of lime and magnesia are best suited. The enamel will lie on them as a smooth and uniform glass without danger of beading up and leaving portions of the surface uncovered, and if the car- bonates be present in sufficient amount, without crazing. Such clays should bake to the hardness of good building brick at silver-melting heat, which is the best tempera- ture for flowing the lead-tin enamels. As a type of such clays, burning to a light buff color and of sufficient hardness at the melting-point of silver (960 C.), one from Hamilton county, Ohio, may be taken, which gave the following analysis : Silica The entire clay Per cent. 41 66 The portion insol- uble in H 2 SO 4 and Na 2 CO 3 Per cent. 25.50 . 2.03 0.00 0.17 0.15 O.^l 93 T ^ 77 6 07 Potash . . 2. it; WHITE ENAMELED BRICK. 141 The portion insol- The entire uble in H 2 SO 4 clay and Na 3 CO 8 Per cent. Per cent. Soda 0.50 0.31 Carbon dioxid 13.24 o.oo Combined water 3.88 o.oo 99.40 28.69 This would bear an enamel of German manufacture, very white, of excellent covering power and entirely frit- ted, that analyzed as follows : Per cent. Lead oxid 28.56 Stannic oxid 9.4 Alumina 4.79 Ferric oxid 0.08 Lime * 0.60 Potash 8.82 Loss on ignition 0.93 Boracic acid none Silica (by difference) 46.82 100.00 The chemical formula of the glass then is: 0.551 1 PbO ) 0.4029 K 2 O Y 0.2035 A1 2 O 3 3.358 SiO 2 0.0460 CaO ) i.o RO with 9.4 per cent, suspended SnO 2 . But tin-enameled brick made with calcareous clays and finished in one fire are not yet common in the United States. In most cases the enamel is applied to ordinary buff and even red brick and flowed in a second heat. 142 THE CHEMISTRY OF POTTERY. A brick of American manufacture was found by the writer, which did not quite conform to either of the above. The enamel had been applied to the "biscuit" or once baked body, but between the brick proper and the enamel was an engobe, apparently applied for the twofold purpose of helping to conceal the body, as the enamel in itself seemed of insufficient covering power and for supplying a surface to which the enamel would ad- here without beading and crawling, as the body of the brick contained insufficient lime to insure this. The material of the brick consisted of Per cent. Silica 67.99 Alumina 24.97 Ferric oxid 1.12 Lime 3.63 Magnesia 1.46 Alkalies 1.24 100.41 The enamel carefully dressed from the surface gave the analysis : Per cent. Stannic oxid 1 1 .42 Lead oxid 33. 1 1 Alumina 7.86 Lime 4.53 Magnesia 0.36 Alkalies 4.87 Boracic acid none Silica (by difference) 38.85 100.00 WHITE ENAMELED BRICK. 143 the chemical formula of the glass then being : o.i798K 2 O ^ 0.3037 CaO V 0.2654 A1 9 O 9 2.252810, 0.5165 PbO J i.o RO with 11.42 per cent, suspended SnO 2 , though so poorly divided that the enamel looked curdy and was, as already stated, of poor covering power. The alumina, as found by the analysis, is unquestion- ably higher than was originally introduced in the glaze ; but in fusion some would be dissolved and taken up from the surface of the body or engobe beneath. Taken as the analysis shows it, the enamel would be produced by the following formula : 0.5165 equivalent white lead 66.98 parts. -337 " whiting 15.20 " 0.1798 " feldspar 50.08 " 0.0856 ' china clay 11.10 " T.oooo " flint 30.00 " Tin oxid 19.86 " The engobe was so eaten into by the glaze that it was difficult to chip off a pure specimen. The best that could be obtained was of the following composition : Per cent. Silica 44.56 Stannic oxid 2.75 Lead oxid : 25.22 Alumina 19.50 Lime 0.57 Magnesia 2.10 Alkalies 5.30 100.00 144 THE CHEMISTRY OF POTTERY. Taking the stannic oxid, in the above, as indicative of the amount of enamel present in the specimen analy- zed and deducting the proportionate amounts of the other constituents, the engobe proper must have had (in the burned condition) the percentage composition : Per cent. Silica 46.51 Lead oxid 22.79 Alumina 2 3- 2 7 Lime and magnesia 1.97 Alkalies 5.46 99-99 which wo.uld approximately be given by the mixture : Feldspar 50 parts. White lead 25 " China-clay 33 " The other type of " enameled " brick mentioned, hav- ing a clear glaze over a white engobe which covers and conceals the buff-body of the brick, is made from a sandy fire-clay and finished with its two coatings in the clay state, being subjected when dry to but one fire. The temperature of this is quite high, sufficient to melt the pyrometric cones nine or ten, the glazes used not being plumbiferous, but of the porcelain type. The body of an English brick of this character had, in its burned condition, the composition: Per cent. Silica 72.36 Alumina 24.97 Ferric oxid 0.74 Lime i .34 WHITE ENAMELED BRICK. 145 Per cent. Magnesia 0.30 Alkalies i.oi 100.72 The engobe was a pure china-clay. The glaze was composed of Per cent. Silica (by difference) ....................... 66.67 Alumina ................................... 20.64 Lime ...................................... 7.68 Magnesia ................................... 0.33 Potash ..................................... 4.68 IOO.OO its chemical formula being ^ 5-76Si0 8 . In the severe fire, to which the ware has been sub- jected, the glaze has, of course, taken up alumina and silica from the engobe, so that its original composition would not accord, in the proportion of these elements, with the formula derived from the analysis of the fired product, as above given. A comparison of this formula with those of the stoneware slips, shows by how much approximately the proportions of alumina and silica are perhaps too high. The formula as it stands would be produced by the following mixture : 0.26 equivalent feldspar .............. 72.41 parts. 0.74 " whiting ............... 37.00 " 0.78 china-clay ............ 101.01 " 2.64 eqi valeuts flint .................. 79.02 ' ' 146 THE CHEMISTRY OF POTTERY. Trials made on the basis of the above formula with systematic reduction of the china-clay and if need be of the flint, would in a very few trials give the composition of a glaze such as a particular clay and the fire required to bake it would need. The clear glaze chipped from a brick of similar char- acter of American manufacture gave the analysis : Per cent. Silica (by difference) ....................... 65.67 Alumina ................................. . . 20.20 Ferric oxid ................................. 0.49 Lime ...................................... 6.27 Magnesia ................................... 0.88 Alkalies 1 ................................... 6.49 IOO.OO the chemical formula then being o.6 45 6CaO j i.o RO The adaptation of a clay and enamel or of a clay, en- gobe, and glaze to each other, in both of these kinds of brick, involves considerable empiric experiment, par- ticularly if it is aimed to finish them in one fire. Not only do the difficulties due to differences of their coeffi- cients of expansion arise, causing crazing or shivering of the glaze or ehamel which must be avoided, but also the shrinkage conditions in the clay state must be so met that the surface coverings do not shell off in the drying or loosen so as to turn up in the fire, melting back in 1 Combining weight, 46.3. WHITE ENAMELED BRICK. 147 beads with exposure of uncovered patches of the brick surface. A consideration that has received very little attention thus far, but is of considerable moment, is that the glaze or enamel itself must resist all action of the atmos- phere and the body of the brick should be of such den- sity that the combined action of moisture and frost does not cause the white coating to split off. The majority of the brick now placed, it is true, are not subjected by their position to the latter danger, but it is not unlikely that in time an important use will be the facing of the walls of light courts, which are not under roof. The glazed surface cannot, in itself, be depended upon for preventing the rain-water from penetrating the brick, as it would find ample access to the body of the same through the mortar-joints. From such brick, saturated with water and subjected to several frosts, the impervious surface not allowing the expansion of the ice through absence of pores, is very soon pushed off. Hence, glazed or enameled brick, in order to be of prime quality, should be very nearly vitreous in body. It is, of course, out of the question to attain this with the soft- fired calcareous body of the first type and these brick should only be used for interior work. A specimen of the kind described, having a feldspar-lead china-clay engobe between the body and the tin enamel, absorbed eighteen per cent, of water. It showed cracking and loosening of its enamel when saturated with water and frozen and thawed twice. In five freezings and thaw- ings it was completely ruined. 148 THE CHEMISTRY OF POTTERY. Brick of the second type with even a porosity allowing an absorption of five per cent, of their weight of water, withstood thirty freezings and thawings, without dam- age to the coated surface. There are few brick of even this type in the market, whether of foreign or domestic manufacture, as dense as this. Most of them will take up ten per cent, by weight of water, which is too much for safe out of door exposure. The hard alkali-alkaline earth glazes are far the best for resisting atmospheric influences, but even lead glazes and enamels, if of suitable composition, may be relied on for retaining a bright reflecting surface. In order to test if such glazes are in no danger of suf- fering surface decomposition, in the course of time, by the action of moisture and carbon dioxid, the most rapid and practical method is that of Professor Rudolf Weber. The glazed surface is exposed for twenty-four hours under a bell jar to the fumes of highly concentrated hydrochloric acid. Dried in an atmosphere free from dust, the surface must remain perfectly bright or show at most the very faintest clouding. Marked clouding indicates a decomposition of the surface, which may be further seen ori wiping off the cloud of salts and separa- ted silica in a decided iridescence of the glass. CHAPTER XIII. FLOOR-TILE AND TERRA-COTTA. HE raw materials of the unglazed wares for architectural use and ornament are almost co- extensive with the argillaceous minerals of the earth. Their final selection for the sev- eral purposes depends upon the conditions of use and upon the necessary or accidental con- ditions of manufacture. The first considerations are with reference to physical properties, allowing facile shaping and the faultless dry- ing and burning of the ware. In plastic-formed objects, particularly in very large pieces of architectural terra-cotta, the degree of plas- ticity and the binding property of the clay are mat- ters of very great moment, which in the present state of knowledge can only be safely determined empirically and on a scale of actual work. The plasticity of the clay must be ample to allow for the easy forming of the mass in molds, or its modeling free-hand, and also allow heavy masses in bold relief to sustain their own weight. Yet the clay dare not be so plastic that it be strained in the shaping and twist and be contorted in drying and in the fire. To a certain degree the evils of an over-plastic and a very fine grained ma- terial are overcome by adding once fired clay (''grog") in coarse grains to the mass. 150 THE CHEMISTRY OF POTTERY. Such material must have sufficient binding property to resist, without cracking, the strain of a shrinking sur- face on a less shrunken interior, during the operations of drying and burning, both attended by appreciable re- ductions of mass from the surface inward. Dust-pressed ware, such as flooring tile and molding brick for belt-courses, require a material that does not form center-cracks, from the surface-sealing of the clay during pressing, before the air contained in the powder has escaped, a phenomenon known as "busting." The material must further allow the ware to dry and be burned without warping, cracking through ("dunting") or becoming covered with surface-cracks ("checking"). The selection of a proper kiln-temperature for burning the ware is a difficult matter that can also only be done empirically. As a rule in these manufactures a variety of colors are required and the problem of temperature is the finding of the heat that will bake to proper hardness the largest number of the available clays, so that by mix- ing and suitable disposition in the kilns, as great a variety of colors as possible may be burned in one and the same fire. The temperatures employed in these industries vary from that of the melting-point of silver to that of pyro- metric cone ten. As a rule, too, they have been derived haphazard from the heat found best for one clay and one particular kind of ware and not from systematic trial of all the clays that are likely to enter into the manufacture. As a result many establishments labor under the serious difficulty that a part of their products are far less perfect FI^OOR-TII/E AND TKRRA-COTTA. 151 than others, because the whole are burned in the same fire, which is only suited to ware from one'or two of the clays. Or, again, in order to meet this difficulty, the colors which can not be properly produced from the local clays at the established temperature, must be gotten from clays obtained from a distance, thus burdening the manufacture with great expense. The conditions thus imposed by chance are in the ma- jority of cases accepted as inevitable to the manufacture and no systematic effort is made to readjust and correct them. While it is true that mere analyses of materials are of little or no avail in the solving of such problems, the training of -the technical chemist fits him better to carry out systematically the many empiric trials necessary for the readjustment of the conditions or their proper estab- lishment originally in these works. The opportunity afforded workers of such training, to study the problems of these arts in actual manufacture, will alone lead to the finding of physical data, obtainable in the labora- tory, by which the behavior of the various clays can be foretold and bases of rational manufacture worked out before the practical establishment of plants. All wares for architectural use must adhere closely to definite sizes. The shrinkage of the clays employed must therefore be definitely known for the temperatures to which they are to be subjected, so that the molds and dies can be made of the proper dimensions. Frequently pieces of the same pattern, but of different colors, alter- nate in a frieze or floor. It is therefore a great conveni- 152 THE CHEMISTRY OF POTTERY. ence if the shrinkage of all the adopted colors be adjusted to the same scale, as it would save carrying molds and dies of the same design in a variety of sizes. The question, however, is entirely one of convenience and relative ex- pense. In the case of dust-made brick with plain and simple molding faces and geometric floor tile, it is per- haps best that the bodies employed have different shrink- ages ; as in that case the dies, after wearing too large for the material of least shrinkage, can be chopped out to the size required for the clay of next greater shrinkage, and so continue until they have done service for the entire color scale. Were all the materials on the same scale of shrinkage, the life of a die would be comparatively short, as it could no longer be used when it became too large, by wear, for the standard size. The question of shrinkage is most aggravated in the case of encaustic tile, whether made plastic or of dust- clay. In these, the colored clays forming the design are inlaid in the clay making the body of the ware. As the product must be absolutely level, the least difference in the shrinkage of the several clays, both in the drying and burning, would unfailingly warp the tile, either causing them to " buckle" or " dish." In this manu- facture it is also absolutely necessary that each colored clay require the same heat and attain an equal hardness with the others, for it must be possible to inlay any combination of the color scale into the same piece, and therefore occupy the identical position in the kiln. Ad- vantage, therefore, cannot be taken of the greater or less difference in temperature that always exists between FI/)OR-TIIvK AND TERRA-COTTA. 153 certain parts of a kiln, for burning particular colors, as is always done, where individual pieces are made of but one mass. Soluble salts of the alkalies, alkaline earths, or of iron, whether pre-existent in the clay or formed while it lies as a damp mass or in the fire, are dangerous in a variety of ways ; chemical examination for their pres- ence or the likelihood of their formation, is an important though greatly neglected step in the selection of clays, for ware which is not to be glazed. Gypsum, if present, and ferrous sulfate often formed in the damp clay by oxidation of the finely divided iron pyrites it may contain, are on the drying of the finished pieces brought to the surface, particularly on prominent points and ridges, where, on account of the readier evaporation, larger quantities of the contained moisture with its dissolved salts, are drawn by capillarity. Thus, on the eye-brows, tip of the nose, lips, and chin of a terra-cotta head would most of the salts be lodged. On burning, these surfaces, which in the clay state may have shown no defect, become bodily discolored ; red and similar colors will, through the presence of lime salts, become buff, and where there are larger amounts may be even coated with a whitish, irremovable scum ; while light-colored clays would, through the presence of the iron salts, stain brown in these parts. Where the use of such clays cannot be avoided, two methods of treatment are employed to escape these re- sults. One consists in adding to the clay barium chlorid and carbonate, in order to render the salts insoluble by 154 THB CHEMISTRY OF POTTERY. double decomposition, preventing their carrying to the surface with the movement of the water in drying. Bighty-five to ninety per cent, of the barium necessary, as shown by the determination of sulfuric acid in the clay, is added in the form of chlorid, dissolved in water, and the remainder with a small additional excess, in the form of carbonate. The other method consists in pre- venting evaporation from the right side of the ware, by coating it, after it has been formed and finished, but be- fore drying, with crude petroleum or tar. As the evap- oration is forced to take place entirely from the uncoated surfaces, the salts are, in the drying of the ware, en- tirely carried away from the surface and lodged where, in the setting of the piece in the wall, they are unobjec- tionable. The coating material of the face must be of such a nature that it burns away without leaving a blem- ish. In the case of ware burned at higher heats and not protected from the immediate contact of the flame, sur- fa*ce discolorations are likely to occur from the flying ash of the firings. Where this is a matter of serious objection, it is necessary to burn with fuel gas or crude petroleum, or at least with a solid fuel containing less iron than the ash of our commoner coals. Where the temperature reached is not so great as to destroy the sulfates of the clay and fix their bases as undecomposa- ble silicates, or where the burning is conducted with fos- sil fuels, containing sulfur, with a practically continu- ous oxidizing flame, so that the clay takes up sulfuric anhydrid from the fire gases, soluble salts are lodged in the ware. These tend seriously to deface it if, as in FI/)OR-TlIv AND TERRA-COTTA. 155 the case of bricks and terra-cotta, it is exposed in outer walls to alternate damp and drying. Through these processes the salts are brought to the surface as a white crystalline efflorescene in dry weather, which is not washed off but returned by absorption to the ware with the first succeeding rainfall. The efflorescences, con- sisting mainly of potassium, calcium, and magnesium sulfates, cannot be removed except by scraping the sur- face after a period of drought has brought them thor- oughly out of the body of the ware. It is then possible, by careful application ot a dilute solution of barium chlorid, to fix what still remains in the body of the wall. All the efflorescences of salts found on walls are by no means derived from the brick or terra-cottas alone. Frequently the mortar or cement setting is responsible ; often in stables or outhouses, where nitrogenous liquids are absorbed by the brick or ammoniacal fumes are plentiful, the ware may be the seat of a true nitrif active process and the salty efflorescences contain considerable nitrates. Soluble lime salts may be introduced into clays by the materials used to produce certain shades of color. Thus for chocolates, browns, black, recovered manganese, and umber are frequently employed. The latter is sel- dom free from gypsum. The physical properties of the finished wares are re- ceiving greater attention than formerly, though the re- quirements of architects should be more rigorous. It is very important that brick and terra-cottas be burned to such density that they absorb very little water 156 THE CHEMISTRY OF POTTERY. and suffer no decay through the rigor of our northern climate. Roofing tile, which on account of the light roof -con- struction long in vogue in the United States, have but recently been introduced in larger amount, with sub- stantial building, both on account of their technical ex- cellence and decorative character, should be particularly dense and resistant to weathering influences. The same applies to paving material. Sidewalk blocks and paving-brick are now frequently required to show less than one per cent, absorption of water by weight. Floor-tile are not subjected to as severe conditions, but if they do take up more than from three to four per cent. by weight of water, they become difficult to keep clean from the grinding of dirt into their pores. The following data, averaged from a considerable num- ber of tests made on commercial wares of all makers rep- resented in our market, will show that while there is much that amply fills all requirements, the average by no means does. WATER ABSORPTION OF FI,OOR-TII V E. Percentages by weight. Color of the clay. Extremes. Average. Salmon 1.5 to 9.1 5.8 Buff i. 9 to 7.2 4.6 Light gray 1.9 to 8.5 5.8 Dark gray 2.0 to 5.8 4.4 Chocolate o.o to 7.4 4.8 Red J. 5 to 8.4 6.0 Black 4.4 to 10.3 7.5 Fawn 8.3 In encaustic tile, it is important that the clay consti- tuting the "backing material" be fully as dense as the FLOOR-TILE AND TERRA-COTTA. 157 clays making the design inlaid in the face. Where this is not observed the tile are in danger of certain destruc- tion wherever they are exposed to wet and frost, as on the floors of open vestibules and verandas. The process of destruction is quite the same as that already pointed out in the case of porous glazed brick, subjected to similar exposure. A porous " backing- clay" takes up considerable water, which on freezing does not find room for expansion through a nearly im- pervious surface. In consequence these nearly vitreous surf ace- clays, making the design are very soon spalled off, through ice-pressure. This is particularly the case in ware made by the English plastic process. In this the surface bodies, known as "Jaspers" are very dense, while the backing, to make it as little liable to warp as possible, is made very porous by the addition of a large proportion of fired clay in coarse granules ("grog"). Exposed to the conditions indicated, the tile are, in the course of one or two winters, ruined completely. Prime requirements of paving-ware are, further, hard- ness and toughness. It should certainly not be possible to scratch anything for such use with even hardened steel. Grinding tests and tumblings for a certain length of time in a "rattler," as is used in the foundries for cleaning castings, have been made, with a view to getting factors representing both of the necessary qualities. The results obtained are valuable, but mainly comparative, and no standards have, as yet, been set up. 1 1 Geological Survey of Ohio, vol. vii, part i, p. 192. Edward Orton, Jr. : Clay-working Industries. CHAPTER XIV. REFRACTORY MATERIALS. HK subject of refractory materials is receiving much serious attention from engineers, and the quality of such prod- ucts, furnished for the various metal- lurgical operations, is improving with rapid strides. Recognition that the term refractory is altogether relative has gained ground and leading to a more exact formulation of demands, is becoming the proper basis of improvement. Refractory bodies being used as apparatus and con- tainers for materials and products subjected to various mechanical and chemical operations at high tempera- tures must, above all things, be infusible at the temper- atures to which these are to be subjected ; they must not soften, swell up, or otherwise lose shape on fre- quent repetition of this condition of temperature or such excess of it as the accidents of operating may make lia- ble. Refractoriness to temperature very much beyond this point is not only useless but is often bought at the expense of other qualities, which from this point on, be- come quite as essential. Such qualities are the ability to resist abrasion, chemical action of the material, or the fuel and disintegration by repeated heating and cooling. It is not possible, in the compass of this work, to treat REFRACTORY MATERIALS. 159 of the subject of refractory materials other than as the potter himself makes use of them for the building of kilns, the making of saggars for holding ware that must be protected from direct contact with the flame and the wads for luting the same against the entrance of fire gases. The potter, then, has nothing to do with any materials, other than clays, belonging to this cate- gory. The problems concerning him, in this connec- tion, are those relating to the resistance to temperature, which is seldom as high as in many metallurgical opera- tions ; resistance to disintegration under the movements of repeated heatings and coolings, and under the burden of often very great weights, but of no abrasive action, as in the blast-furnace and lime-kiln ; resistance to the action of the flame and the fluxing properties of the ash of the fuel, but of no strong chemical influences requir- ing decidedly basic or entirely acid refractory bodies. Nevertheless, inattention to the demands of the condi- tions, although these are not as stringent as in most chemical and metallurgical operations, and are, perhaps, for this very reason more often neglected, becomes the cause of a serious drain on many establishments, in the incessant repair of kilns and the loss of saggars. Practical trials, so-called, are usually quite insufficient to draw from them conclusions as to the behavior of the clay on a working scale. Frequently the potter merely places a lump of the clay, that he intends to apply for re- fractory purposes, on the bag-wall of the kiln or in the " cut " of the fire, and notices if, after its baking in the place of most intense heat, it remains without trace of 160 THE CHEMISTRY OF POTTERY. fusion or even sintering. Often he may even make a brick or saggar of the clay and expose it similarly to the most intense heat of his kiln, believing that he is thus making a trial under severe working conditions. This is, of course, far from being the case. Only when a large number of pieces are tested continuously under actual conditions of work, are the accidental irregulari- ties of the making of single pieces lost in an average ap- proximating the truth and only subjection to the many destructive influences of continued use, can answer the question of serviceableness. Such practical trials, if properly carried out, are, contrary to the common belief of those unaccustomed to systematic experiment, expen- sive and long in giving a conclusive answer. The question, whether the most serviceable refractory materials are being employed, is an open one in most potteries, which seldom reaches a satisfactory answer or is dismissed from all consideration, and the losses en- tailed are accepted as inevitable, without attaining the assurance that they are so. A knowledge of the chem- ical composition of the fire-clays and their pyrometric value by direct determination would dispel all such doubts and enable their selection on the most rational basis that the circumstances will allow. In addition to an exact knowledge of the character of the material, it is important to satisfy one's self that the ware made of it, the bricks, tile, saggars, etc., have been burned at a temperature higher than that to which they are after- wards likely to be subjected, in order to prevent the al- most certain danger of cracking and the sinking of REFRACTORY MATERIALS. l6l arches in fire-brick structures, when ware shrinks on being subjected in use to temperatures higher than those at which it was originally burned. To this important matter very little attention has been paid. Engineers have never as yet made a specific demand of the manufacturers of fire-brick and tile, that the ware for different purposes be subjected to tempera- tures which, according to their use, must not fall below certain definitely prescribed minima. Furthermore, potters very seldom burn their saggars in a special fire at heats higher than they will meet in future use, but burn the green stock continuously in their regular ware-kilns, placing it on top of the bungs, where it usually gets less heat than in the places where the saggers are later likely to be exposed. Experience has shown that in order to resist repeated heatings and coolings most successfully, refractory wares must remain as porous or even spongy in character at the temperatures reached, as is consistent with the necessary firmness of body to carry the weights they have to bear. This is particularly the case where from position in a furnace or elsewhere the heating is of necessity rapid and the heated surfaces are liable to be struck by cold air. In order to attain this structure it is customary to mix with the plastic fire-clay, which should in itself be so refractory as to burn quite porous at the heat given, at least an equal weight of the same or a similarly refractory clay which has been burned and ground and so sifted that only particles from the size of split peas to beans are used ; the finer material 162 THE CHEMISTRY OF POTTERY. and dust, which would fill up the pores and interstices, producing too compact a mass, are discarded. This material is known to potters as "grog." About pot- teries there are always sufficient broken fire-brick and saggars to supply the "grog" needed. In the fire- brick yards it is almost entirely substituted by coarsely ground unburned flint-clays. For the best products it is sometimes burned before use. Clays markedly siliceous remain porous even in the higher heats reached in pottery burning and can be given an open structure by the addition of pure sili- ceous sand, where they do not already contain it in sufficient amount to bum of that character in themselves. The majority of refractory clays at command contain more or less finely divided free silica and are not only of ample heat-resisting power, but some of these are of notable quality. Yet the reverse of the popular belief, that the refractoriness is in proportion to the free silica present, is true : a belief unfortunately kept alive by the statements in many economic geologies. The origin of this mistaken belief is, that in the comparatively low heats of pottery kilns the greater the portion of uncom- bined silica, whether occurring naturally in a clay in which the basic oxids are not excessive, or added arti- ficially, the more porous and seemingly more refractory it is. Seger, however, has shown 1 that the syste- matic addition of silica to a pure silicate of alumina Al 2 O 3 .2SiO 2 , whether artificial or a natural kaolinite, reduces the melting-point until the proportion A1,O 8 . 1 Thoniudustrie-Zeitung, 1893 p. 391. REFRACTORY MATERIALS. 163 iySiO 2 is reached, which melts with the lowest members of the scales which Seger and Bischof have severally set up as refractory standards. Beyond this point the addi- tion of silica causes a uniform rise in refractoriness until the proportion of alumina is so small that it practically disappears from the mixture : proving the high charac- ter of both kaolin and silica separately, as refractory agents and the degree of quality lost in their mixture. It must not be thought that because a clay low in basic oxids in proportion to the contained free silica, remains porous in even the hardest pottery fires, that for the refractory products used by the potter, the more silic- eous his fire-clays are, the better ; particularly as the more siliceous clays are much cheaper and as the use of a sandy clay or the addition of quartz sand to his fire-clay will give the necessary open structure as well as the more expensive " grog." Practical experience, where made in conjunction with a knowledge of the composition of the fire-clays, has proven that this is not the case. Mechanical reasons, mainly, unfit very siliceous clays for use as saggars ; chemical reasons, in many cases, make them unserviceable for the fire-brick of the kilns. It has already been pointed out that the designating of a material as refrac- tory is properly being done more and more in a relative sense, with a statement as to the conditions under which it is refractory. No chemist or engineer would for a moment think of lining a lime-kiln with dinas-brick. In nearly all treatises accessible to the manufacturer and the public at large, however, refractoriness is still 164 THE CHEMISTRY OF POTTERY. treated as if it were an absolute quality, inherent in the material itself and without discussion of the extraneous conditions that may break it down. To value a clay directly and solely by its "oxygen ratio" calculated from the analysis, so commonly taught the public by chemists, interesting themselves in the subject is, to say the least, very misleading. All that this ratio is sup- posed to teach, the relative fusibilities of clays, is more expeditiously and accurately determined by direct fusion in the Deville forge ; for the structure of a clay plays quite as important a part in its fusibility as its chemical composition. The chemical analysis gives a general but not a very accurate clue to its fusibility ; but it gives very positive information of the behavior of the clay toward out- side chemical agents to which it may be subjected under the influence of heat. As far as any danger from fusing is concerned, any highly siliceous clay reasona- bly low in basic oxids, would sufficiently bear a pot- tery fire in itself. But the loss that the potter fears in his saggars is not from fusing but from cracking. Their walls, compared with brick, are comparatively thin and being filled with ware and piled in " bungs " to the height of fifteen feet and more, often have to bear considerable weight. While a brick, forming part of a solid wall of masonry, presents but one front to the fire, saggars are surrounded with the fire-gases and upon cessation of the fire, with the strong draught of the cooling air. They are therefore subjected to more or less rapid changes of temperature, particularly when REFRACTORY MATERIALS. 165 placed opposite the kiln-mouths and spyholes. A sag- gar made of clay burning dense at the heat of the kiln, will not remain intact under such conditions without cracking, and one of a siliceous clay, even when left quite porous in the fire, soon succumbs also, particularly if a portion of the silica is in the form of grains of quartz. It is probable that the well-known property of the swell- ing of quartz in the fire produces strains, which under the severe conditions the ware cannot stand. A clay from Perry county, Ohio, which has found not inconsiderable use for making saggars, fire-brick, wads, and other refractory wares for pottery use, has the com- position : The portion insoluble in The entire H 3 SO 4 and clay Na a CO 3 Per cent. Per cent. Silica 74.93 5I-52 Alumina 17.19 0.42 Ferric oxid 0.79 0.09 Lime 0.29 0.04 Magnesia 0.46 0.02 Alkalies 1 i .61 0.38 Combined water 5.44 o.oo 100.71 52.47 RATIONAL ANALYSIS. < Per cent. Clay substance 48.24 Quartz 49.72 Feldspathic detritus 2.75 100.71 1 Combining weights respectively 34.9 and 34.5. 1 66 THE CHEMISTRY OF POTTERY. PERCENTAGE COMPOSITION OF THE CIAY SUBSTANCE. Per cent. Silica 48.53 Alumina 34-77 Ferric oxid 1.45 Lime 0.51 Magnesia 0.91 Alkalies 2.55 Combined water 11.28 100.00 Thirty per cent, of the quartz would not pass a sieve with eighty meshes to the inch. This clay, burned at the heat of melting orthoclase, gives a yellowish white body, which is very porous and shrinks but four per cent. Saggars made of it, filled with heavy ware, and set in high bungs have an average life of but three and one-half burnings, at the melting temperature of feldspar (orthoclase). The same clay, made into brick, stood sufficiently well in many instances as the lining of pottery kilns, but whenever brought in contact with basic oxids the destruction was rapid and complete. Used in the mouths, bottom -flues, and arches of a muffle-kiln for slip -glazed stoneware, fired with coal yielding a ferruginous ash, these parts broke down in a few firings. The fire-arches of two boilers built of them, under which the slack of a similar coal, high in iron pyrites, was used, lasted but a few weeks. As the lin- ing of a salt-glaze kiln they melted away very fast. The reason of the clays succumbing to the action of basic oxids is readily explained from the composition of stoneware glazes themselves. REFRACTORY MATERIALS. 167 It was seen that the most fusible silicates containing no reducible metals as lead or zinc, were those approx- imating the formula iRO. o.5R 2 O 3 .4SiO 2 and that more aluminous compounds formed puckered viscid crusts, but not fusible glasses. In this clay the relation of alumina to silica is as i.oto 8.34 and under the action of ferrous oxid and of soda, it would give fusible glasses of about the composition iFeO. o.5A! 2 O 3 . 4.i7SiO 2 , and iNa 2 O. o.5A! 2 O 3 . 4.17 SiO 2 , which on running off the face of the brick would constantly expose more of it to their destructive action. An aluminous clay would ab- sorb some of the base, forming in time a superficial vitreous coat, but not a fusible glass : it would remain in place as an effective protection to the material be- neath. The wads and plugs used to support the bungs of salt-glazed stoneware and keep them separate during burning, should not be made of siliceous clays resem- bling the one above discussed, for the reason that the more readily such a clay combines with the alkali fumes the more will it stick to the ware, from the for- mation of glass at the outer line of contact with it. These plugs cannot, in this case, be struck off without more or less damage to the pieces to which they were attached. Plugs of aluminous clay remain dry and give no cause for sticking. Stoneware potters using clays that salt-glaze readily are often troubled with the rims of the pieces, that are set on each other, adhering, causing some breakage w r hen the ware is taken apart. Such as have tried it, l68 THE CHEMISTRY OF POTTERY. will have found that the washing of the surfaces of con- tact with flint, as is done with saggars in the white- ware potteries, may often aggravate the trouble. The reason for this is patent from the experiences and argu- ments given, as is also the remedy sought. China-clay or a fire-clay, as near the composition of kaolinite as possible, would give a proper wash for the purpose. On the other hand, wads used for luting the saggars holding ware that is not burned in the open fire, which serve mainly as a yielding bed for each saggar as it is placed on top of that last set and enables the kilnman to run his " bungs" up straight, should be as siliceous as possible. Made of such clays they will, by their lower shrinkage and binding action, from their remaining more open in a fire free from basic fumes, adhere to and break the saggars less than other clays. For such purpose the clay cited proved excellent. The example of this material sufficiently illustrates the value of the chemical analysis in estimating the suitableness or unsuitableness of a clay as a refractory material under different conditions, entirely aside from its ability to bear mere temperature without fusing. Use of refractory materials is still made with most un- necessary waste, both in the employment of clays that are too good and of those which are altogether inade- quate for their application. The responsibility for this must rest largely with the very one-sided study that chemists have, as a rule, made of fire-clays, in looking solely to their relative fusibilities and failure to instruct the public in the physical and chemical conditions that REFRACTORY MATERIALS. 169 a refractory material has to resist in different places of its application, which often are of such importance that the relative fusibilities are, beyond a certain point, of very minor significance. The best plastic, clays for saggar-making have thus far proven to be the tertiary clays of New Jersey , * par- ticularly those that with not too high a percentage of uncombined silica contain considerable organic matter that burns out in the fire giving the clay a very porous structure. Clays of low-binding power when dry, how- ever, as New Jersey clays are apt to be, must be worked very carefully that the green ware does not crack before it comes into the fire. The flint-clays of south-eastern Ohio form, when burned, an excellent material as "grog." Some of these are very pure, having almost the composition of kaolinite, as the following : Per cent. Silica 46.54 Alumina. 38.47 Ferric oxid 0.77 Lime 0.29 Magnesia 0.23 Alkalies' i .38 . Combined water 12.98 100.66 The clay, in spite of its composition was, when very finely pulverized, extremely difficult to decompose with 1 For a valuable description of New Jersey clays see " Report on the clay deposits of Woodbridge, South Amboy, and other places in New Jersey." 1878, by George H. Cook, State Geologist. 2 Combining weight, 40,8. 170 THE CHEMISTRY OF POTTERY. sulfuric acid. Even after a treatment of fifty hours with the acid, it left, after washing with sodium carbonate, a residue of 7.54 per cent. From the fact that in many instances these flint-clays are more refractory than the available plastic clays used as the bonds in the fire-clay products, and that in con- sequence fire-brick are commonly graded into first and second qualities, as they contain a greater or less amount of flint-clay, the idea has become rather com- mon that all flint-clays are highly refractory. This is far from being the case ; not a few of low quality, have from this mistaken idea, come extensively into use. An example of one so used is a flint-clay of Beaver county, Pennsylvania, of which the following is the analysis : The entire clay The portion insoluble contains " in H a SO 4 and Na 2 CO 3 Per cent. Per cent. Silica 65.85 36.86 Alumina 22.87 2.19 Ferric oxid 1.14 o. n Lime 0.53 0.42 Magnesia 0.37 0.07 Alkalies 1 2.01 0.47 Combined water 6.93 .... 99.70 40.12 RATIONAL ANALYSIS. Per cent. Clay substance 59-5$ Quartz 30.42 Feldspathic detritus 9.70 99.70 1 Combining weights respectively 45.5 and 39.7. REFRACTORY MATERIALS. PERCENTAGE COMPOSITION OF THE CLAY SUBSTANCE. Per cent. Silica 48.66 Alumina 34-?i Ferric oxid 1.73 Lime 0.18 Magnesia o. 50 Alkalies 2.59 Combined water 11 .63 100.00 In a kiln burned to the melting of feldspar the clay becomes very dense and it melts in the Deville furnace together with Seger's pyrometric cone twenty-six, so that it would be looked upon as a fire-clay of the lowest grade. According to the oxygen ratio (Feuerfestigkeitsquo- O in A1 2 O 3 1 tient) of Bischof ol^So which is 1.24 in this clay, O in A1 2 O 3 it is not even worthy of being designated a fire-clay, as the lowest member of the scale he admits to this clas- sification is the clay of Niederpleis on the Sieg, with the ratio 1.64. In the examination of a fire-clay, first a chemical analysis is required to determine how it will resist chemical and mechanical influences incident to its use. From this the oxygen ratio (Feuerfestigkeitsquotient) may be calculated, according to Bischof, 2 to give a gen- 1 Using, however, the old formulas and equivalent weights and calculating iron as ferrous oxid with the RO elements. 2 C. Bischof, Die feuerfesten Thone, Leipzig, 1876 ; Transactions of the American Institute of Mining Engineers, Virginia Beach meeting, February, 1894 ; H. O. Hofman and C. D. Demoud : Some Experiments for Determining the Refractoriness of Fire-clays. 172 THE CHEMISTRY OF POTTERY. eral idea of its relative fusibility. The only value of this is to be able to institute a comparison of the clay with others of which the analyses are available, but of which no direct determinations of fusibility are recorded. Bischof's quotient is preferable because it is best known and because no other can, after all, bring in the main factor of variation, which depends on the physical struc- ture of the clay. Next, the direct determination of the relative fusi- bility of the clay is made. It is most satisfactory if the standing of the clay is referred to Seger's pyrometric scale. This direct determination is made according' to Seger and Cramer 1 by forming small tetrahedra of the clay to be examined, two centimeters high by one centi- meter on the edges of the triangular base and inclosing these, together with similarly formed pyrometers of the higher members of Seger's pyrometric scale, in a small covered crucible, five centimeters high by four and one- half in diameter, the walls and cover being five millime- ters thick. The material of which these crucibles are made is a mixture of equal parts of alumina and china- clay, burned to a "grog," in as high a heat as is at command ; for forming, the material is rendered plastic by the addition of the necessary amount of unburned china-clay. In the bottom of the crucible a layer of the alumina-china-clay grog, powdered fine, is packed and the bases of the trials pressed into it, to prevent their being upset. The crucible with its trial pieces is set in 1 Thonindustrie-Zeitung, 1893, p. 1281. Also, Transactions of the American Institute of Mining Engineers, Florida Meeting, 1895. H. O. Hofmau : Fur- ther Experiments for Determining the Fusibility of Fire-clays. REFRACTORY MATERIALS. 173 the Sefstrom or Deville furnace upon a block of refrac- tory material of sufficient height to raise it into the focus of the heat generated by gas-retort graphite in pieces the size of a hickory-nut and maintained by a gentle blast from a rotary blower. The member of the pyrometric scale melting coin- cidently with the clay being tested gives the numerical index of its refractoriness. As the trial pieces are not under observation during the heating several tests at greater and less heats are necessary to fix the right degree. The composition of these higher pyrometers of Seger is as follows : ! Cone number. Chemical composition. } "A1A. 44SiO,. a " 54SiO s . '- 66SiO 8 . 26 ' : - 2 ^ '. 73810,. 1 Thonindustrie-Zeitung, 1893, p. 1252. 174 THE CHEMISTRY OF POTTERY. Cone number. Chemical composition. 28. A1 2 O 3 . ioSiO 2 . 29. A1 2 3 . 8Si0 2 . 30. A1 2 3 . 6Si0 2 . 31. A1 2 3 . 5 Si0 2 . 32. A1 2 8 . 4SiO a . 33- A1 2 O 3 . 3SiO 2 . 34- A1 2 O 3 . 2. 5 Si0 2 . 35- A1 2 O 3 . 2SiO 2 . 1 36. A1 2 S . 2SiO,. 2 To receive the designation " fire-clay," a clay should not melt more easily than pyrometer twenty-six. High grade fire-clays seem often of low binding pow- er. While sufficient in this property, for the making of brick and of the ordinary saggars a considerable amount of bonding clay is yearly imported from Germa- ny for the making of glasspots and other large ware. It would certainly seem that our great wealth in clays must include deposits having the necessary physical qualities coupled with high refractoriness and it is im- portant that systematic examinations be made in this direction also the data being furnished with the analy- sis and relative fusibilities. Finally, in accordance with the character of the clay and the use of the products which are to be made of it, the chemist should recommend the minimum tempera- ture of their burning. 1 Zettlitz kaolin, with which also the preceding ones are made. 2 Shale of Rakonitz. CHAPTER XV. BURNING THE WARE. HK important chemical operation to which all pottery wares are subjected, is that of burning or firing. This operation is carried on in kilns, that vary widely in character and construction, a variety, for which the differ- ent kinds of ware and the questions of hand- ling and setting in the kilns, to best advantage, present many good reasons. But it cannot be denied, that in many instances, the mechanical considerations which have determined the style of kiln, have been accepted without a clear understanding or due appreciation of the operation of burning in itself and its effects on the ware. These are among the most important of the subject. The construction and size of the kiln, the character of its fire-places, and the means of creating and regulating the draught, must all be such, that for the purpose for which the kiln is built, it is possible : 1 . To obtain the necessary temperatures required in the proper intervals of time and to hold them at pleasure, as the work may demand. 2. To get as uniform a temperature throughout the kiln as possible by the ability to direct of advance the heat in its different parts, as the progress of the fire may require. 176 THE CHEMISTRY OF POTTERY. 3. To be able to get the chemical quality of flame needed for any operation at will. 4. To get the maximum heat effect with a minimum of fuel. From the standpoint of the chemist, who accepts the methods and apparatus of empirical work, but must modify or add to them, so as to eliminate the elements of chance, the first consideration in firing, bears on the question of production and regulation of the draught. As the ultimately high temperatures produced in all pottery kilns causes a sufficiently rapid movement of the gases to bring an abundance of air into the firings for the combustion, many forms of kilns have no regular chimney. The smoke and products of combustion either escape from the imperfectly closed top, as in the old brick-clamp or from openings in the arched crown, as in the English pottery kiln. The latter, it is true, is as a rule, topped with a cone, intended to collect the gases from the various openings and convey them above the roofs of the surrounding buildings. It acts, in a measure as a chimney but having a large radiating surface it acts very poorly, when a positive draught in the kiln is most needed, namely, when the temperature in the kiln is still low. Besides, its form precludes the possibility of regulation with it except to a very limited extent. As the admission of air into the kiln is deter- mined by the speed with which the products of combus- tion are withdrawn and by the greater or less obstruction presented to the air at its entrance; in kilns of the above type this regulation is mainly effected in the latter way. BURNING THE WARE. 177 Such a method of regulating the volume of air needed, by more or less obstruction at its entrance, is very troublesome and uncertain. There must of necessity be many places of entrance to attend to affecting the draught more or less by their size, shape, and the position of the fuel in the firings. The regulation of these, individu- ally, must at most be the veriest guesswork. By the withdrawal of all the products of combustion through a single stack, of such dimensions to produce at all times a draught in excess of the needs, and controlling this with a single damper, positive in its action, the pressure in the kiln can at will be brought to any definite degree below that of the atmosphere and produce a proportion- ate influx of air through all openings. The individual air inlets can then be adjusted separately, if need be, to the requirements of each fire-place, or manipulated to throw the flame to the top or crowd it to the center and bottom of the kiln, as the progress of the heat in differ- ent parts of the kiln may require. The draught in the chimney determines the propor- tion of air entering the kiln, and therefore the character of the combustion and the quality of the flame. Hence, it is necessary to determine, by analysis, the composition of the gases leaving the kiln through the stack and at the same time to note, by a continuously acting kiln-barometer, the pressure in the kiln below that of the air, at which the gases are of the composition found. A few of such analyses, during the progress of a firing, at times when the kiln- mouths have just been charged with fresh fuel and when the same has well 1 78 THE CHEMISTRY OF POTTERY. burned down, will establish the proper kiln-pressures for admitting the needed amount of air at the different stages of fire, so that the damper can be handled alto- gether by the reading of the kiln-barometer and gas- analyses will only be needed as occasional checks. For the gas analyses, the Orsat apparatus is conven- ient, though even a simple Bunte or a Cramer carbon- ic-acid pipette is sufficient to gain the necessary knowl- edge of the quality of the gases leaving the kiln. The most convenient kiln- barometer or draught me- ter is a modfied form of that of Scheurer-Kestner. 1 It consists of a tin box having a glass guage tube at the side, opening into the box at its lower end and into the air at the upper, inclined at an angle of nine degrees, through KILN-BAROMETER OR DRAUGHT-METER, which the rise and fall of the liquid in the box is amplified ten-fold, making the readings of the slight changes of level possible with the unaided eye. The guage is supplied with a 1 Thoniudustrie-Zeitung, 1891, p. 696 ; Dr. Julius Post: Chemisch-Tech- niche analyse, vol. II, p. 73 ; Braunschweig, 1890-1. BURNING THK WARE. 179 scale and suitable metallic fittings attaching it to the box and protecting it against breakage. The box is filled with carbon oil to the zero mark on the guage- glass, when it hangs level. The opening through which the box is filled is closed with a thumb- screw pierced with a piece of brass tubing. By means of a piece of rubber hose attached to this and a porcelain tube walled air-tight into the kiln or chimney, the at- mosphere of the latter is connected with that above the oil, communicating the kiln-pressure to it. Instead of a porcelain tube a duct may be walled from the kiln to the top of the ' ' hob ' ' surrounding the same and the bell- shaped iron foot of the stand of the instrument placed over it. The rubber hose connection of the kiln-barom- eter is then attached to the gas-pipe upright, as shown in the cut, and the communication with the kiln estab- lished through the duct, bell, upright pipe, and hose to to the box. Through the porcelain tube used to com- municate the kiln-pressure, where this is used, the sam- ples of gas for analyses may be taken, though it is also well to take such samples from the top and bottom of the kiln, by other porcelain tubes, properly let into the walls of the same at the respective places. Unfortunately, these simple devices for attaining a posi- tive knowledge of the character of the fire and helping it in the condition desired by regulating the damper with reference to distinct stack-pressures, have not, as yet, found entrance into our potteries. Firing is looked upon as a mysterious operation and even experienced burners do not attack it, in new kilns, or with new l8o THE CHEMISTRY OF POTTERY. wares, with the confidence that the possession of positive data and a knowledge of their meaning would give them. The majority of even the best kilnmen waste con- siderable fuel by the admission of excessive amounts of air to the kiln. How great this waste is, few potters perhaps realize. A glost kiln fired with crude petro- leum sprayed into the kiln-mouths with compressed air from atomizing burners required 1,050 gallons or 3,135 kilograms of the fuel. The kilo of oil required about 14.35 kilos of air for its combustion and probably had a calorific value of 10,000 heat units. Analyses of the flue gases proved that throughout the burning about 175 per cent, of air was used, as the kiln was ordinarily fired. Subsequent firings showed that although lead glazes were being burned an average consumption of no per cent, of air was not only ample for smokeless combustion, but did not endanger a loss of ware by reduction of the lead of the glaze. Hence, sixty-five per cent, of the theoretically necessary amount of air was admitted to the kiln, above its practical needs, amounting to 29,239 kilograms. At the finish of the burning the gases left the kiln with a temperature suffi- cient to melt an alloy of twenty per cent, silver, eighty per cent, gold, or about at the temperature 1,045 C. The average heat of the gases leaving the kiln from be- ginning to end of the fire was therefore 523 C. Tak- ing 0.2377 a s the specific heat of air, each kilogram of the excess of the same took away needlessly 124.32 heat units or 3,634,660 in all: a quantity which required the combustion of 122 gallons of the petroleum to sup- BURNING THK WARE. l8l ply. At subsequent firings, a careful regulation of the draught, according to the practical needs, effected a saving of about three barrels of oil to each burning. The above example is by no means an extreme case ; on the contrary it is probable that in coal-firing an ex- cess of upwards of 100 per cent, of air is the practice. In an ordinary boiler-firing it is customary to allow an excess of thirty per cent, of air, in order to insure am- ple contact of the solid fuel with the same, on the grate- bars. In firing a pottery kiln, the excess of air may be less, because of the better opportunity for combination in the greater distance of travel under far higher tem- peratures than in the boiler, though how much less, is a matter of experience, depending upon the fuel, the kiln temperature, the character of the firings, and the dis- tance the gases have to travel in the kiln, before reach- ing the exit. In order to distribute the heat as uniformly as pos- sible throughout the kiln, it is important that the com- bustion take place mainly beyond the fire-places and that these act rather as generators. Those parts of the kiln exposed to radiation from the fire-places, known as the "cuts," are invariably hotter than the others and this condition is greatly aggravated, when the fuel lies shallow and is mainly consumed on the grate-bars. The heating of the kiln does not, in this case, it is true, take place entirely by radiation and conduction, for through the dissociation of the products of combustion in passing this heated zone, combustible gases are formed which recombine at the lower tempera- 1 82 THE CHEMISTRY OF POTTKRY. tures in the interior of the kiln. But the absorption of heat in the ' ' cuts ' ' by the dissociating gases is never sufficient to reduce the heat to a moderate difference from that in the rest of the kiln. For this reason, the fire-places, instead of being broad and shallow, should be narrow and deep, approaching the character of fuel- gas generators. The gases leaving them being mainly reducing in character, burn in their progress through the kiln, as by diffusion or by striking solid objects ob- structing their way, they become thoroughly mixed with the air entering through doors, air-openings, and spy-holes. It is only where the firings are of this char- acter that it becomes possible to burn with a relatively small excess of air and effect the economies incident to so doing. For where the fuel lies shallow on the grate-bars air channels soon work through it, admitting excessive amounts of air ascending in compact columns, which afterwards comes but little into play in com- bustion, where the firings send the complete products of combustion and not combustible gases along with this air into the kiln. As there is in this case ab- solutely no way of knowing the proportion of grate sur- face sending nearly pure air into the kiln, and as there is no way of regulating it, a reduction of the general draught of the kiln until the flue-gases have the proper composition, would often be attended by a cutting down of the combustion until insufficient for producing the necessary rise in temperature. Only where the kiln- mouths send reducing gases into the kiln and the needed excess of air enters through openings that can be BURNING THE WARK. 183 manipulated and controlled is the fire absolutely in hand as then, by closing the air inlets more or less, the draught of the stack can be made to act more or less on the fire, sucking the fuel-gases in the proportion wanted. It is a familiar fact to all who have had to make analyses of furnace-gases, that columns of gases, even where great chemical affinity exists between them and the temperature is favorable to combination, may travel side by side with little diffusion and combination. It is necessary, therefore, to make the distance of travel in the furnace of sufficient length and provide that the gases strike solid objects and are compelled to turn on them- selves to effect mechanically as thorough a mixture as possible. The most rational kilns, then, are those working on what is known as the " down draught" prin- ciple. In these the gases rise from the fire-places to the crown against which they strike and are compelled to descend between the bungs of saggars or of ware to the flues under the floor which lead to a center tunnel con- nected with the stack. The striking against the crown of the kiln, the horizontal movement under the same, with the mixture effected by the impeding tops of the bungs of ware and the down ward movement, most effect- ually breaks up any tendency of the gases to move in separate channels. The building up of the ware in the kilns and the ar- rangement of the flues and the tunnels to the stack must be made with careful consideration to the end that the gases from the different fire-places meet similar fric- tional resistances throughout their long distance of travel ; 1 84 THE CHEMISTRY OF POTTERY. otherwise the combustion in different parts of the kiln will vary so as to make it impossible to burn it with any approximation to uniformity. And, furthermore, it must not be forgotten, that as the gases have to travel at least three times the distance they do in an up-draught kiln, and where the stack is outside of the kiln, at least four times this distance, the frictional resistance they encounter is proportionately as much greater and must be overcome by a proportionately greater draught or a corresponding reduction in the speed of combustion and cooling will have to be expected. This fact has often been overlooked where an up-draught kiln has been changed into a down-draught, by closing the crown holes and carrying a center stack in the kiln to just above the crown, the cone carrying off the smoke and creating the draught as before, expecting an increase of this draught without any alteration. Where in such a case it was attempted to burn glazes high in basic oxids, particularly sensitive ones and such of a composition causing devitrification if they are not rapidly cooled, disappointment at the down-draught principle has unjustly ensued. With a suitable provi- sion for the necessarily greater draught required to draw the fire-gases through their longer and more tor- tuous course, all the seeming advantages of the up- draught kiln can be obtained, without sacrifice of the more thorough combustion aimed at in the other con- struction. The dissociation of the products of combustion in the " cuts" of the kiln as well as the production of combus- BURNING THK WARE. 185 tible gases in the producer-like fire-places yields sufficient length of flame for the longer distance of travel, in the down-draught kiln, even with short-flamed fuel. This cannot, however, be counted on to extend the effect of the fire with practical uniformity of temperature to an indefinite distance in the kiln. As the flame, after some distance of travel toward the outlet, becomes freighted with the products of combustion, these, together with the inert gases that it necessarily carries with it, nitro- gen and the water vapor derived from the moisture in the fuel and the oxygen which it contained, may grow to such proportions as to extinguish it, even though it still contains combustible products and an excess of oxygen. There is then a practical limitation to the dis- tance of travel of the fire-gases, consistent with their being still effective, which fixes the dimensions of kilns. When these exceed a capacity of 3,500 cubic feet it is difficult to regulate them so as to burn sensitive products uniformly. The regulation of the English pottery kilns, which are the common type in use in most of our ceramic indus- tries, must be effected altogether by the manipulation of the air inlets. In up-draught kilns, the draught being as a rule much too strong, the bottoms are most liable to be hard, while in down-draught kilns the reverse is the case and it is often difficult to bring the lower ware to the required temperature. In the latter case the flame must be crowded to the center and bottom of the kiln, which is done in admitting air over the kiln-mouths by withdrawal of the regulating brick. The air sucked in 1 86 THE CHEMISTRY OF POTTERY. through the exposed opening rises in a column next to the kiln-wall to the crown and forces the fire-gases in- ward. Where this source of air is insufficient the fire- doors will often have to be given more or less opening. Amounts of air, largely in excess of the needs of com- bustion, are in this manner often necessarily introduced into the kiln for regulation, and as previously pointed out, involve a correspondingly greater consumption of fuel. Where the bottoms of the kiln get too hard it is, of course, necessary to cut off as near as possible all air supplies over the fire, which tend to crowd it down. Serious defects interfering with the regulation must be met by permanent alterations in the kiln, as increasing or reducing the size of the crown holes, the flues, and of the cone, or of admitting air into the latter over the crown. But the variations in draught due to the wind, and the changing pressure of the atmosphere, and the rise in temperature of the kiln, which in a kiln with a stack and damper can be met by the setting of the lat- ter, can in these be counteracted by the manipulation of the air inlets alone. The rapidity of firing a kiln is determined by the draught, the size and number of the fire-places, the in- tervals of time in which fuel is supplied to the fires and the weight and character of the ware contained in the kiln. Thin-walled ware, that is reasonably uniform in sec- tion, as most dish ware, can be burned and cooled quite rapidly without danger of loss in the kiln or creation of a brittle product as one with internal strains tending to shatter it. It is well to have kilns for the manufacture of BURNING THE WARK. 187 this, with a very strong draught, providing the kiln- mouths be correspondingly large or numerous to take the fuel for a plentiful combustion. It is then possible to push the productivity of a kiln, if the manufacture demands it. Glazes having a tendency to cloud or devitrify must be fired and especially cooled rapidly. With these it is not infrequently the custom to " draw the fires," as it is called. As soon as the required temperature has been reached the grate-bars are let down, the fire raked out of the mouths and extinguished. Sanitary ware, terra-cottas, tile, and all varieties of brick should be fired and cooled slowly. In all of these the bodies are thick and opportunity must be given the hygroscopic and especially the combined water of the clay to leave it slowly so as not to incur the danger of warping, shattering, or cracking the pieces. It is like- wise important that such products be cooled slowly, whereby they undergo a sort of annealing process that is very essential to their life. Building terra-cottas and the vessels used in sanitary plumbing often have thin walls attached to thick ones ; the unequal cooling of which is unavoidably associated with the rapid cooling of non-conducting products, brings about unequal and severe strains, making these pieces sensitive to sudden changes of temperature and light blows or jars. All paving material is sensibly toughened by slow cooling. The kilns for these products need not, therefore, have so great a proportionate fire-place volume to the volume of kiln space and the draught may be correspondingly 1 88 THK CHEMISTRY OF POTTKRY. less. The slowing of the fire by charging the furnaces with baitings of fuel at greater intervals is only possible with a corresponding reduction of the draught. Where the fire is maintained with solid fuel the temperature should rise in regular throbs or pulsations correspond- ing with the baitings. These should burn down to clear glowing coals, before the mouths are again charged, but not so far that the temperature begins to recede from the point reached. The progress of the heat must be continuously upward and as uniform in the periods of its rising as possible. Where then this progress is too great the fuel must be made to burn more slowly by cut- ting down its draught. The stormy evolution of gaseous products, when fresh baitings of fuel are put on the fires, produces periodic- ally a reducing atmosphere of greater or less duration in the kiln ; if this is succeeded each time by a clear oxidizing atmosphere, no danger of the reduction of even lead glazes need be feared. In fact, the alterna- ting of a short period of reducing influences followed by a longer one of oxidation is in many cases preferable to a continuously oxidizing fire throughout the burning. All fossil fuels, oil, coal, and the fuel-gas made from it, as well as natural gas, contain considerable sulfur, which, on burning, is absorbed in an oxidizing atmos- phere as sulfuric anhydride by glazed surfaces, causing unsightly separations of crystalline sulfates in the melted glass. Through the action of reducing gases, these sulfates formed on the surfaces of the glaze are decomposed, the BURNING THE WARE. 189 sulfur being carried away with the other gases of com- bustion as dioxid. Seger 1 found that the clay of Birkenwerder, rich in lime and iron, glowed for two hours in an atmosphere of sulfur dioxid, and air took up 13.6 per cent, of sulfuric anhydrid. By heating it again in a reducing atmos- phere this could be entirely expelled. Through this discovery he found an explanation of the phenomenon that clays of this type, which in the ordinary intermit- tent kilns fired with coal produce a yellow or cream- colored body from the formation of a lime-iron silicate, in continuous or gas-fired kilns, generally burned red or were flashed with red. In the latter the kiln-atmos- phere is generally continuously oxidizing. Under this condition the lime in the clay absorbing sulfuric acid from the fire-gases does not combine with the iron and leaves the latter to impart its red color to the product. It has been found that in following Seger 's suggestion of manipulating the draught in these kilns so that at regular intervals the kiln-atmosphere is strongly reduc- ing, these clays can be burned uniformly to a cream col- or as well as in the intermittent kilns fired with solid fuel. In the case of tin-enamels, the writer has observed that when burning in kilns fired with crude petroleum sprayed into the mouths with compressed air, so that the fire is continuously oxidizing, they are very liable to have a pink cast, while burned with an alternately re- ducing and oxidizing flame they become snowy white. These instances will suffice to show that the smoke 1 Thonindustrie-Zeitung, 1877, p. 22. IQO THE CHEMISTRY OF POTTERY. issuing at regular intervals from the ordinary kiln is not only harmless, but may play a very useful roll, provided that the burning be so conducted that too much fuel is not wasted and the kiln-temperature is not depressed in throwing too much coal on the fires at each baiting. The useful results of temporary and recurring reduc- ing periods in the burning, unwittingly attained in fir- ing with solid fuels can, of course, be just as easily ob- tained with oil and fuel-gas. But as these are often used, mainly, from a supposed advantage of the ease with which they are handled to give a smokeless com- bustion, and as to the average mind this is the perfec- tion of firing regardless of the results sought, it is dim- cult to make the potter understand that this may be as- sociated with practical objections. The phenomenon that potters commonly ascribe to the effect of sulfur in the coal is due to excessive reduction in glost-kilns, through careless firing. It is a reduction of the lead of the glaze so that the latter turns black and it comes when the fireman has been careless about clinkering his fires and has not kept the grate-bars clean. The phe- nomenon may in so far hang together with sulfur in the coal, that inasmuch as the sulfur is mainly present as iron pyrites, when this is high, the large amount of iron slag or clinker that is formed requires especial watchful- ness, to prevent its shutting off the air supply coming through the live coals. Our pottery kilns, like all the older apparatus using high temperatures that discharge their products of com- bustion directly into the air, without being compelled BURNING THE WARE. IQI to first give off their surplus heat, where it can again be utilized for the burning, are under the most favorable conditions of operation very wasteful of fuel. The most successful application of the regenerative principle, in which a greater effect is obtained from fuel, than perhaps, in any other apparatus, is the ring-brick kiln of Hoffmann and Licht. Few of these have as yet been built in the United States and few of the regenera- tive gas-kilns of Mendheim and of Dunnachie. Ignor- ance of the progress made in ceramic industries else- where, are in part responsible for this, though valid commercial reasons connected with the cheapness of our fuel and the generally higher price of labor, have made people slow to lock up the greater amounts of capital involved in these modern kilns. There is no reason, however, since the progress made in this direction by the iron and glass industries, why the pottery industries should not follow ; especially as in these the application is much more simple. Special ' ' stoves ' ' and ' ' regenerators ' ' for taking up the waste heat and imparting it to the air for the com- bustion being unnecessary, as a battery of kilns suita- bly connected, with their contained ware, act in turn in this capacity ; the kilns in advance of the one in fire taking up the waste heat from the products of combus- tion before these are turned into the stack, while the air for the combustion serves to cool those which have been fired and comes hot to the one burning. 1 1 Groves and Thorpe : Chemical Technology, Vol. i., Fuel and its Appli- cations ; Thonindustrie-Zeitung, 1877-1895. INDEX. ACETATES in white lead deleterious for preparation of glazes therefrom 125 Alloys of Prinsep 30 Analysis of clay. Necessity for accuracy in, 3 ; separation of silica and alumina, 4 ; alumina and ferric determination, 5 ; impurities in reagents, 5 ; impurities in distilled water, 6 ; de- termination of potash and soda, 6 ; prox- imate analysis, 6 ; " clay substance," 8 ; " rational analysis," 8 ; calculation of. to dried substance, 13 ; for white enameled brick, 140, 144 ; for saggars, fire brick, * etc 165,170 BANKO ware 78 " Barbotine" ware 73 Bischof's opinion of rational analysis 9 Biscuit, analysis of, 13 ; calculation of mass for 13 " Blue calx," composition of 120 Boron in glazes, 14 ; action of boron in glazes, 50 ; volatili- zation of 125 Brick, white enameled 139 Broginarts' classification of ceramics 41 Burning, 175 ; draught in the kiln, 176 ; analysis of kiln gases. 177; draught meter, 178; firing, 179; uni- form distribution of heat, placing of the ware in kilns, down and up draught kilns, 183 ; rapidity of firing, 186 ; 'reducing action of fresh fuel, 188 ; re- generative kilns 191 " C. C." ware 43, 117 Ceramics, classification of 41 " Checking" 21 Chemical ware 78 China-clay 96 194 INDEX. Chromic oxid use as coloring material 133 Clay analyses. Kaolin from Nelson Co., Va., 10 ; kaolin from Zettlitz, 37 ; from Western N. Caro- lina, 37 ; weathered shale, 60 ; red-colored clay, 61 ; of clay for yellow ware, 66 ; flint clay, 68 ; for stoneware, 80 ; Albany slip clay, 83 ; slip clay, 86 ; kaolin, 95, 96 ; from Lawrence Co., Ind., 98 ; from Inyo Co., Cal., 98; plastic kaolin from Florida, 100 ; New Jersey clay, 101 ; from Jefferson Co., Mo., and Calloway Co., Ky, 102 ; Cornwall stone, no, 112, 114; from Texas 115 Clay, importance of careful sampling for analysis of, i ; necessity for careful analysis of, 3 ; purification of, by washing and sifting, 3 ; action of sulfuric acid upon.. 9 Clay, physical properties of, 15 ; coloring of burned, 15 ; physical testing of, 16 ; mechanical analysis of, 17 ; plasticity of, 18 ; binding power of, 20 ; burning test for, 21 ; for floor tile and paving brick, testing for porosity, 24 ; shrinkage 24 Cobalt oxid, composition of commercial 132 Copper oxid, use as coloring material 133 Cornish stone 118,119,120 " Crazing," 48 ; cause of 5 Cream colored ware, 117, glazes for 122 DISTILLED water, impurities in 6 Doulton ware 7 8 "Bunting" 21 Dust pressed articles 25 ENAMEL, description of the term, 46 ; see also glazes Enameled brick, 139 ; separation of glaze from, by freezing and thawing *47 Enameled tile, 127 ; glazes for 129 Engobe ware, 65 ; for white enameled brick 142, 145 JFAIENCE 45. 62 > 66 INDEX. 195 Fayal pottery 43 Feldspar, analysis of Rorstrand, 37 ; analysis of, from N. Y., 37 ; analyses of commercial, 108, 109 ; Cornwall stone, no, 112, 114; from Texas 114 Ferric oxid, use as a coloring material, 136; analysis of ' ' Crocus Martis, ' ' 136 Fire brick 158 Flint, [see quartz] Flint clays, analysis of 170 Flint glass, analysis of 52 Frit melting 125 GL,AZES. Preparations of samples for analysis, 13 ; con- taining boric acid, 14 ; necessary qualities of, 48 ; crazing and shivering, 50 ; action of boric acid, 50; silica, 51 ; alumina, 51 ; formula for, 51 ; distinction between "raw " and "fritted," 53 ; example of calculation of, 54 ; action of sulfates in, 55 ; for red ware, 58 ; for Rockwood pottery, 62 ; for yellow ware, 71 ; for Rocking- hatn ware, 72 ; salt glaze, 81 ; slip glaze for stoneware, 83, 91 ; for white granite and C. C. ware, 122 ; for majolica and enameled tile, 129; for white enameled brick, 141, 145, 146 ; action of frost upon glaze for enameled brick 147 Granite ware 117 HEINTZ'S glass mixtures 32 ' ' Hotel china " 44 KJL/NS, for testing clay, 23 ; dimensions of, 28 ; down draught, 183 ; [see also burning] " LIMOGES" ware..... 73 TWAJOLJCA, 127; glaz.es for 129 Majolica and faience, distinction between 45 Manganese, use as a coloring material 134 Marble, analysis of Carrara 37 196 INDEX. Mica, in clay, 3 ; clay containing products of the decompo- sition of IT INICKElv oxid, use and composition 133 PENNSYLVANIA Dutch pottery 62, 65 ' ' Plastic kaolins " 100 Porosity of floor tile and paving brick, test for 24 Prinsep, alloys of 30 Pyrometry 26 C^UARTZ, in clay, 9 ; analysis of Norwegian, 37 ; analysis of Illinois, 38 ; from N. J. and 111. analysis, 105 ; from Tenn., 106 ; from Ky 107 '* RED ware," 43, 58 ; glaze for, 58 ; articles made of, 61 ; examination of materials for 64 Refractory materials 158 " Rockingham " ware, 66 ; glaze for, 72; articles made of, 73 ; examinations of materials for 73 Rookwood pottery 62 SAGGARS, clay for 163, 165 Sampling of clay, directions for 2 Sand, [see quartz] Seger & Aron's rational analysis of clay 8 Seger's cones, 33 ; mixtures for preparing, 38 ; use of, 40 ; composition and use of 173 " Shivering," 48 ; cause of 50 Shrinkage in firing, test of 24 Silica, distinction between quartz and combined, 9 ; separa- tion from alumina and ferric oxid 4, 5 "Slips" 65 Stoneware, 77 ; articles made of, 77 ; use of, in chemical industries, 78 : clay for, 79 ; glazes, 81 ; exami- nation of material 92 Sulfates objectionable in glazes, 55 ; in floor tile and terra cotta, 153 ; action in the kiln 189 TEMPERATURE of the kiln, methods for determining. . . 26 INDEX. 197 Terra-cotta, 149 ; shrinkage in burning 152 Thenard's blue, composition of 120 Tile, enameled, 127; glazes for, 129; floor, 149 ; roofing, 156 ; encaustic 156 Tin oxid, use in enamels 137 UMBER, use as a coloring material 134 Uranium, use as a coloring material 135 VICAT'S needle 18 1A/ATER, impurities in distilled 6 " W. G." ware 43 White granite ware, 117 ; glazes for 122 White lead, deleterious influence on glazes of acetate con- tained in 125 White ware 93 Whiting, analysis of 38 ; composition of 116 YELLOW ware, 66 ; preparation of the clay, 67 ; glaze for, 71 ; articles made of, 73 ; examination of materials for 73 Chemical Books. NOTES ON METALLURGICAL ANALYSIS. By Prof. N. W. Lord, Columbus, Ohio. pp. 101, bound in cloth. Price $1.25. A GUIDE TO STEREOCHEMISTRY. By Arnold Eiloart. Price $1.00. THE CHEMISTRY OF THE POTTERY INDUSTRY. By Karl L/angen- beck. Bound in flexible cloth. Price $2.00. PRINCIPLES AND PRACTICE OF AGRICULTURAL ANALYSIS. By Harvey W. Wiley. Published in monthly parts, 25 cents each, $3.00 a year. Vol. I, bound in cloth, price $3.75. Vol.11, bound in cloth, price $2.00. Send for prospectus. ON THE DENSITIES OF OXYGEN AND HYDROGEN AND ON THE RATIO OF THEIR ATOMIC WEIGHTS. By Edward W. Morley, Ph.D. Quarto, paper, 117 pp. Price $i .00. THE SCIENTIFIC FOUNDATIONS OF ANALYTICAL CHEMISTRY. By Wilhelm Ostwald, Ph.D. Translated by George M'Gowan, Ph.D. viii.207 pp. Price $1.60. A HANDBOOK OF INDUSTRIAL ORGANIC CHEMISTRY. By Samuel P. Sadtler, Ph.D., F.C.S. Second revised and enlarged edi- tion. Price $5. oo. MANUAL OF ORGANIC CHEMISTRY. By Dr. L,assar-Cohn. Trans- lated by Alexander Smith. 403 pp. Price $2.25. Alembic Club Reprints. Crown 8vo. Cloth, Price 50 cts. each. No. i. EXPERIMENTS UPON MAGNESIA ALBA. By Joseph Black, M.D., (1755) 47 PP- No. 2. FOUNDATIONS OF THE ATOMIC THEORY. Papers by Dalton, Wallaston and Thomson, (1802-1808) 48 pp. No. 3. EXPERIMENTS ON AIR. Papers by Hon. Henry Caven- dish, F. R. S., (1784-1785) 52 pp., illustrated. No. 4. FOUNDATIONS OF THE MOLECULAR THEORY. Papers by Dalton, Gay-L,ussac and Avogadro, (1808-1811) 51 pp. No 5. EXTRACTS FROM MICROGRAPHIA. By Robert Hooke, F.R.S., (1665) 52 pp. No. 6. THE DECOMPOSITION OF THE FIXED ALKALIES AND AL- KALINE EARTHS. By Humphrey Davy, (1807-1808) 51 pp. No. 7. THE DISCOVERY OF OXYGEN. Part I, by Joseph Priest- ley, LL.D., (1775) 35 PP- No. 8. THE DISCOVERY OF OXYGEN. Part II, by Carl Wilhelm SCHEELE, (1777) 46 pp. No. 9. THE ELEMENTARY NATURE OF CHLORINE. By Hum- phrey Davy, (1809-1818) 77 pp. No. 10. RESEARCHES ON THE ARSENIATES, PHOSPHATES AND MODIFICATIONS OF PHOSPHORIC ACID. By Thomas Graham, F. R. S., (1833) 45 pp. No. ii. THE INCREASE IN WEIGHT OF TIN AND LEAD ON CAL- CINATION. By Jean Rey, (1630) 54 pp. OTHER NUMBERS WILL FOLLOW. Journals. THE JOURNAL OF ANALYTICAL AND APPLIED CHEMISTRY, 1887- 1893. Seven volumes, $3.00 a volume. Send for table of contents. THE CHEMICAL NEWS. Renewals and subscriptions to this pe- riodical may be sent to us at the rate of $4.87 per annum. THE JOURNAL OF THE AMERICAN CHEMICAL SOCIETY. Monthly. $5.00 a year. We Buy and Sell Odd Numbers of ^^^ B^B^ American Chemical Journal. The Journal of the Society of Chemical Industry. The Chemical News. Journal of the Chemical Society (London). The Analyst. Berichte der deutschen chemischen Gesellschaft. Zeitschrift fiir analytische Chemie. Zeitschrift fiir angewandte Chemie. Pamphlets. HINTS TO BEGINNERS IN IRON ANALYSIS. By David H. Brown. Price 15 cents. THE PIERCE PROCESS FOR THE PRODUCTION OF CHARCOAL, WOOD ALCOHOL AND ACETIC ACID. By William L. Dudley. Price 10 cents. Chemical Publishing Co., EASTON, PA. PATENT PORTABLE KILNS! -FOR- FIRING DECORATED CHINA, GLASS, BISQUE, &c. In several sizes, with Charcoal, Manufactured Gas, and Natu- ral Gas. Sixteen years in market. Are not an experiment but an assured success. Send for circular. STEARNS, FITCH & CO., SPRINGFIELD, OHIO. HYDROCHLORIC, NITRIC, \ ACIDS, SULFURIC, Ammonia free from Pyridin and Anilin, ZincOxjd, Hanganese Dioxid, Calcium Chlorid, Potassium Permanganate And all Chemicals Used by the Analyst. Foral1 proses. Lowest in ASH DADPD 'rMr Lrl. ofanymade. Write for Price-List. BAKER & ADAMSON, Easton, Pa. LIBRARY USE RETURN TO DESK FROM WHICH BORROWED MAIN LIBRARY CIRCULATION DEPARTMENT THIS BOOK IS DUE BEFORE CLOSING TIME ON LAST DATE STAMPED BELOW INTtRLI&RAftY LOAN FEB 1 1 1986 UNIV. OF CALIF., BERK. LD62A-30m-7/73 (R2278lO)9412-A-32 General Library University of California Berkeley YB 15424 THE UNIVERSITY OF CALIFORNIA LIBRARY-