GLASS-MAKING

general principles of . Gla may be defined in technical phraseology, to be a transparent homogeneous compound formed by the fusion of silica with oxides of the alkaline, earthy, or common metals. It is usually colourle , and then resembles rock crystal, but is occasionally stained by accident or design with coloured metallic oxides. At common temperatures it is hard and brittle, in thick pieces; in thin plates or threads, flexible and elastic; sonorous when struck; fracture conchoidal, and of that peculiar lustre called vitreous; at a red heat, becoming soft, ductile and plastic. Besides gla properly so called, other bodies are capable of entering into vitreous fusion, as phosphoric acid, boracic acid, arsenic acid, as also certain metallic oxides, as of lead, and antimony, and several chlorides; some of which are denominated gla es. Impure and opaque vitriform ma es are called slags; such are the productions of blast iron furnaces and many metallurgic operations. Silica, formerly styled the earth of flints, which constitutes the basis of all commercial gla , is infusible by itself in the strongest fire of our furnaces; but its vitreous fusion is easily effected by a competent addition of potash or soda, either alone or mixed with lime or litharge. The silica, which may be regarded as belonging to the cla of acids, combines at the heat of fusion with these bases, into saline compounds; and hence gla may be viewed as a silicate of certain oxides, in which the acid and the bases exist in equivalent proportions. Were these proportions, or the quantities of the bases which silica requires for its saturation at the melting point, exactly ascertained, we might readily determine beforehand the best proportions of materials for the gla manufacture. But as this is far from being the case, and as it is, moreover, not improbable that the capacity of saturation of the silica varies with the temperature, and that the properties of gla also vary with the bases, we must, in the present state of our knowledge, regulate the proportions rather by practice than by theory, though the latter may throw an indirect light upon the subject. For example, a good colourle gla has been found by analysis to consist of 72 parts of silica, 13 parts of potash, and 10 parts of lime, in 95 parts. If we reduce these numbers to the equivalent ratios, we shall have the following results; taking the atomic weights as given by Berzelius. | 1 | atom | potash | = | 590 | 14·67 | | | 1 | lime | 356 | 8·84 | | 3 | silica | 1722 | 42·79 | | - | 71·49 | | 2 | silica | 1155 | 28·70 | | 3823 | 95·00 | | This gla would therefore have been probably better compounded with the just atomic proportions, to which it nearly approaches, viz. 71·49 silica, 14·67 potash, and 8·84 lime, instead of those given above as its actual constituents. The proportions in which silica unites with the alkaline and other oxides are modified by the temperature as above stated; the lower the heat, the le silica will enter into the gla , and the more of the base will in general be required. If a gla which contains an exce of alkali be exposed to a much higher temperature than that of its formation, a portion of the base will be set free to act upon the materials of the earthen pot, or to be di ipated in fumes, until such a silicate remains as to constitute a permanent gla corresponding to that temperature. Hence the same mixture of vitrifiable materials will yield very different results, according to the heats in which it is fused and worked in the gla -house; and therefore the composition should always be referrible to the going of the furnace. When a species of gla which at a high temperature formed a transparent combination with a considerable quantity of lime, is kept for some time in fusion at a lower temperature, a portion of the lime unites with the silica into another combination of a semi-vitreous or even of a stony aspect, so as to spoil the transparency of the gla altogether. There is probably a super silicate, and a sub silicate formed in such cases; the latter being much the more fusible of the two compounds. for 24 hours, is an example of this species of vitreous change in which new affinities are exercised at a lower temperature. An exce of silica, caused by the volatilization of alkaline matter with too strong firing, will bring on similar appearances. The specific gravity of gla varies from 2·3 to 3·6. That of least specific gravity consists of merely silica and potash fused together; that with lime is somewhat denser, and with oxide of lead denser still. Plate gla made from silica, soda, and lime, has a specific gravity which varies from 2·50 to 2·6; crystal or flint gla from 3·0 to 3·6. The power of gla to resist the action of water, alkalis, acids, air, and light, is in general the greater, the higher the temperature employed in its manufacture, the smaller the proportion of its fluxes, and the more exact the equivalent ratios of its constituents. When gla contains too much alkali, it is partially soluble in water. Most crystal gla is affected by having water boiled in it for a considerable time; but crown gla being poorer in alkali, and containing no lead, resists that action much longer, and is therefore better adapted to chemical operations. The affinity of gla for water, or its hygrometric attraction, is also proportional to the quantity of alkali which it contains. In general also potash gla is more apt to become damp than soda gla , agreeably to the respective hygrometric properties of these two alkalis, and also to the smaller proportion of soda than of potash requisite to form gla . Air and light operate upon gla probably by their oxidizing property. Bluish or greenish coloured gla es become by exposure colourle , in consequence undoubtedly of the peroxidize ment of the iron, to whose protoxide they owe their tint; other gla es become purple red from the peroxidize ment of the manganese. The gla es which contain lead, suffer another kind of change in the air, if sulphuretted hydrogen be present; the oxide of lead is converted into a sulphuret, with the effect of rendering the surface of the gla opaque and iridescent. The more lead is in the gla , the quicker does this iridescence supervene. By boiling concentrated sulphuric acid in a gla ve el, or upon gla , we can ascertain its power of resisting ordinary menstrua. Good gla will remain smooth and transparent; bad gla will become rough and dim. The brittlene of unannealed gla by change of temperature is sometimes very great. I have known a thick ve el to fly by vici itudes of the atmosphere alone. This defect may be corrected by slowly heating the ve el in salt-water or oil to the highest pitch consistent with the nature of these liquids, and letting it cool very slowly. Within the limits of that range of heat, it will, in consequence of this treatment, bear alternations of temperature without cracking as before. It has been said that gla made from silica and alkalis alone, will not resist the action of water, but that the addition of a little lime is nece ary for this effect. In general 100 parts of quartzose sand require 33 parts of dry carbonate of soda for their vitrification, and 45 parts of dry carbonate of potash. But to make unchangeable alkaline gla , especially with potash, a smaller quantity of this than the above should be used, with a very violent heat. A small proportion of lime increases the density, hardne , and lustre of gla ; and it aids in decomposing the alkaline sulphates and muriates always present in the pearl ash of commerce. From 7 to 20 parts of dry slaked lime have been added for 100 of silica, with advantage, it is said, in some German gla manufactories, where the alkaline matter is soda; for pota does not a imilate well with the calcareous earth. In many gla works on the Continent, sulphate of soda is the form under which alkaline matter is introduced into gla . This salt requires the addition of 8 per cent. of charcoal to decompose and di ipate its acid; a result which takes place at a high heat, without the addition of any lime. 88 pounds of quartz-sand, 44 pounds of dry glauber salt, and 3 pounds of charcoal, properly mixed and fused, afford a limpid, fluent, and workable gla ; with the addition of 17 pounds of lime, these materials fuse more readily into a plastic ma . If le carbon be added, the fusion becomes more tedious. The two following formulæ afford good glauber salt gla . | 1. | 2. | | Sand | 100 | 60·3 | | Calcined sulphate of soda | 50 | 26·8 | | Lime | 20 | 10·8 | | Charcoal | 2·65 | 2·1 | The first mixture has been proved in the looking-gla manufactory of Neuhaus near Vienna, and the second by the experiments of Kirn. The fusion of the first requires 18, of the second 21 hours. The bluish-green tinge which these otherwise beautiful and brilliant gla es po e , is not removable by the ordinary means, such as manganese or arsenic, which decolour alkaline gla . When the sulphate of soda and charcoal are used in smaller proportions, the gla becomes more colourle . The tinge is no doubt owing to the sulphur combining with the oxide of sodium, in some such way as in the pigment ultramarine . By a proper addition of galena (the native sulphuret of lead), to glauber salt and quartz sand, without charcoal, it is said a tolerably good crystal gla may be formed. The sulphuric acid of the salt is probably converted by the reaction of the sulphuret of lead into sulphurous acid gas, which is disengaged. One atom of sulphuret of lead = 1495·67, is requisite to decompose 3 atoms of sulphate of soda = 2676. It is stated, on good authority, that a good colourle gla may be obtained by using glauber salt without charcoal, as by the following formula. | Quartz-sand | 100 | pounds | | Calcined glauber salt | 24 | | Lime | 20 | | Cullet of soda gla | 12 | The melting heat must be continued for 26 1 ⁄ 2 hours. A small quantity of the sand is reserved to be thrown in towards the conclusion of the proce , in order to facilitate the expulsion of air bubbles. The above mixture will bear to be blanched by the addition of manganese and arsenic. The decomposition of the salt is in this case effected by the lime, with which the sulphuric acid first combines, is then converted into sulphurous acid, and di ipated. Gla made in this way was found by analysis to consist of 79 parts of silica, 12 lime, and 9·6 soda, without any trace of gypsum or sulphuric acid. Glauber salt is partially volatilized by the heat of the furnace, and acts upon the arch of the oven and the tops of the pots. This is best prevented by introducing at first into the pots the whole of the salt mixed with the charcoal, the lime, and one fourth part of the sand; fusing this mixture at a moderate heat, and adding gradually afterwards the remainder of the sand, increasing the temperature at the same time. If we put in the whole ingredients together, as is done with potash gla , the sand and lime soon fall to the bottom, while the salt rises to the surface, and the combination becomes difficult and unequal. Sulphate of potash acts in the same way as sulphate of soda. Muriate of soda also, according to Kirn, may be used as a gla flux with advantage. The most suitable proportions are 4 parts of potash, 2 of common salt, and 3 of lime, agreeably to the following compositions :— | 1. | 2. | | Quartz-sand | 60·0 | 75·1 | | Calcined carbonate of potash | 17·8 | 19·1 | | Common salt | 8·9 | 9·5 | | Lime | 13·3 | 14·3 | For No. 1., the melting heat must be 10 hours, which turns out a very pure, solid, good gla ; for No. 2., 23 hours of the furnace are required. Instead of the potash, glauber salt may be substituted; the proportions being then 19·1 glauber salt, 9·5 muriate of soda, 14·3 lime, 75·1 sand, and 1·3 charcoal. The oxide of lead is an e ential constituent of the denser gla es, and may be regarded as replacing the lime, so as to form with the quartz-sand a silicate of lead. It a imilates best with purified pearl ash, on account of the freedom of this alkali from iron, which is present in most sodas. Its atomic constitution may be represented as follows :— | Computation. | Analysis. | | Silicic acid | 5 | atoms | = | 2877· | | 59·19 | 59·20 | | Oxide of lead | 1 | = | 1394· | 5 | 28·68 | 28·20 | | Potash | 1 | = | 590· | 0 | 12·13 | 9·00 | | Oxides of iron and manganese | | — | — | 1·40 | | 4861· | 5 | 100·00 | 100·00 | The above analysis by Berthier relates to a specimen of the best English crystal gla , perfectly colourle and free from air-bubbles. This kind of gla may however take several different proportions of potash and silica to the oxide of lead. The composition of mirror plate, as made on the Continent, is as follows :— | White quartz-sand | 300 | pounds | | Dry carbonate of soda | 100 | | Lime slaked in the air | 43 | | Cullet, or old gla | 300 | The manganese should not exceed one half per cent. of the weight of the soda. Optical gla requires to be made with very peculiar care. It is of two different kinds; namely, crown gla and flint gla . The latter contains a considerable proportion of lead, in order to give it an increased dispersive power upon the rays of light, in proportion to its mean refractive power. Optical crown gla should be perfectly limpid, and have so little colour, that a pretty thick piece of it may give no appreciable tinge to the rays of light. It should be exempt from str iæ or veins as well as air-bubbles, and have not the slightest degree of milkine . It should moreover preserve these qualities when worked in considerable quantities. Potash is preferable to soda for making optical crown gla , because the latter alkali is apt to make a gla which devitrifies and becomes opalescent, by long exposure to heat in the annealing proce . A simple potash silicate would be free from this defect, but it would be too attractive of moisture, and apt to decompose eventually by the humidity of the atmosphere. It should therefore contain a small quantity of lime, and as little potash as suffices for making a perfect gla at a pretty high temperature. It is probably owing to the high heats used in the English crown gla works, and the moderate quantity of alkali (soda) which is employed, that our crown gla has been found to answer so well for optical purposes. Practical details of the Manufacture of Gla . The Venetians were the first in modern times who attained to any degree of excellence in the art of working gla , but the French became eventually so zealous of rivalling them, particularly in the construction of mirrors, that a decree was i ued by the court of France, declaring not only that the manufacture of gla should not derogate from the dignity of a nobleman, but that nobles alone should be masters of gla -works. Within the last 30 or 40 years, Great Britain has made rapid advances in this important art, and at the present day her pre-eminence in every department hardly admits of dispute. There are five different species of gla , each requiring a peculiar mode of fabrication, and peculiar materials: 1. The coarsest and simplest form of this manufacture is bottle gla . 2. Next to it in cheapne of material maybe ranked broad or spread window gla . An improved article of this kind is now made near Birmingham, under the name of British or German plate. 3. Crown gla comes next, or window gla , formed in large circular plates or discs. This gla is peculiar to Great Britain. 4. Flint gla , crystal gla , or gla of lead. 5. Plate or fine mirror gla . The materials of every kind of gla are vitrified in pots made of a pure refractory clay; the best kind of which is a species of shale or slate clay dug out of the coal-formation near Stour bridge. It contains hardly any lime or iron, and consists of silica and alumina in nearly equal proportions. The ma es are carefully picked, brushed, and ground under edge iron wheels of considerable weight, and sifted through sieves having 20 meshes in the square inch. This powder is moistened with water (best hot), and kneaded by the feet or a loam-mill into an uniform smooth paste. A large body of this dough should be made up at a time, and laid by in a damp cellar to ripen. Previously to working it into shapes, it should be mixed with about a fourth of its weight of cement of old pots, ground to powder. This mixture is sufficiently plastic, and being le contractile by heat, forms more solid and durable ve els. Gla -house pots have the figure of a truncated cone, with the narrow end undermost; those for bottle and window-gla , being open at top, about 30 inches diameter at bottom, 40 inches at the mouth, and 40 inches deep; but the flint-gla pots are covered in at top with a dome-cap, having a mouth at the side, by which the materials are introduced, and the gla is extracted. Bottle and crown-house pots are from 3 to 4 inches thick; those for flint-houses are an inch thinner, and of proportionally smaller capacity. The well-mixed and kneaded dough is first worked upon a board into a cake for the bottom; over this the sides are raised, by laying on its edges rolls of clay above each other with much manual labour, and careful condensation. The clay is made into lumps, is equalized, and slapped much in the same way as for making Pottery . The pots thus fashioned must be dried very prudently, first in the atmospheric temperature, and finally in a stove floor, which usually borrows its heat directly from the gla -house. Before setting the pots in the furnace, they are annealed during 4 or 5 days, at a red heat in a small reverberatory vault, made on purpose. When completely annealed, they are transferred with the utmost expedition into their seat in the fire, by means of powerful tongs supported on the axle of an iron-wheel carriage frame, and terminating in a long lever for raising them and swinging them round. The pot-setting is a desperate service, and when unskilfully conducted without due mechanical aids, is the forlorn hope of the gla -founder.— Quæque ipse miserrima vid i. The celebrated chemist, Dr. Irvine, caught his last illne by a isting imprudently at this formidable operation. The working breast of the hot furnace must be laid bare so as to open a breach for the extraction of the faulty pot, and the insertion of the fresh one, both in a state of bright incandescence. It is frightful to witne the eyes and fuming visages of the workmen, with the blackening and smoking of their scorched woollen clothes, exposed so long to the direct radiations of the flame. A light mask and sack dre coated with tinfoil, would protect both their faces and persons from any annoyance, at a very cheap rate. The gla -houses are usually built in the form of a cone, from 60 to 100 feet high, and from 50 to 80 feet in diameter at the base. The furnace is constructed in the centre of the area, above an arched or groined gallery which extends acro the whole space, and terminates without the walls, in large folding doors. This cavern must be sufficiently high to allow labourers to wheel out the cinders in their barrows. The middle of the vaulted top is left open in the building, and is covered over with the grate-bars of the furnace. 1. Bottle gla . —The bottle-house and its furnace resemble nearly fig. 505. The furnace is usually an oblong square chamber, built of large fire-bricks, and arched over with fire-stone, a siliceous grit of excellent quality extracted from the coal measures of Newcastle. This furnace stands in the middle of the area; and has its base divided into three compartments. The central space is occupied by the grate-bars; and on either side is the platform or fire-brick siege , (seat,) raised about 12 inches above the level of the ribs upon which the pots rest. Each siege is about 3 feet broad. In the sides of the furnace, semi-circular holes of about a foot diameter are left opposite to, and a little above the top of, each pot, called working holes, by which the workmen shovel in the materials, and take out the plastic gla . At each angle of the furnace there is likewise a hole of about the same size, which communicates with the calcining furnace of a cylindrical form, dome-shaped at top. The flame that escapes from the founding or pot-furnace is thus economically brought to reverberate on the raw materials of the bottle gla , so as to di ipate their carbonaceous or volatile impurities, and convert them into a frit. A bottle-house has generally eight other furnaces or fire-arches; of which six are used for annealing the bottles after they are blown, and two for annealing the pots, before setting them in the furnace. The laws of this country till lately prohibited the use for making common bottles of any fine materials. Nothing but the common river sand, and soap-boilers’ waste, was allowed. About 3 parts of waste, consisting of the insoluble residuum of kelp, mixed with lime and a little saline substance, were used for 1 part of sand. This waste was first of all calcined in two of the fire arches or reverberatories reserved for that purpose, called the coarse arches, where it was kept at a red heat, with occasional stirring, from 24 to 30 hours, being the period of a journey or journée , in which the materials could be melted and worked into bottles. The roasted soap-waste was then withdrawn, under the name of ashes, from its arch, coarsely ground, and mixed with its proper proportion of sand. This mixture was now put into the fine arch, and calcined during the working journey, which extended to 10 or 12 hours. Whenever the pots were worked out, that frit was immediately transferred into them in its ignited state, and the founding proce proceeded with such despatch that this first charge of materials was completely melted down in 6 hours, so that the pots might admit to be filled up again with the second charge of frit, which was founded in 4 hours more. The heat was briskly continued, and in the course of from 12 to 18 hours, according to the size of the pots, the quality of the fuel, and the draught of the furnace, the vitrification was complete. Before blowing the bottles, however, the gla must be left to settle, and to cool down to the blowing consistency, by shutting the cave doors and feeding holes, so as to exclude the air from the fire-grate and the bottom of the hearth. The gla or metal becomes more dense, and by its subsidence throws up the foreign lighter earthy and saline matters in the form of a scum on the surface, which is removed with skimming irons. The furnace is now charged with coal, to enable it to afford a working heat for 4 or 5 hours, at the end of which time more fuel is cautiously added, to preserve adequate heat for finishing the journey . It is hardly po ible to convey in words alone a correct idea of the manipulations nece ary to the formation of a wine bottle; but as the manufacturers make no mystery of this matter, any person may have an opportunity of inspecting the operation. Six people are employed at this task; one, called a gatherer, dips the end of an iron tube, about five feet long, previously made red-hot, into the pot of melted metal , turns the rod round so as to surround it with gla , lifts it out to cool a little, and then dips and turns it round again; and so in succe ion till a ball is formed on its end sufficient to make the required bottle. He then hands it to the blower, who rolls the plastic lump of gla on a smooth stone or cast-iron plate, till he brings it to the very end of the tube; he next introduces the pear-shaped ball into an open bra or cast-iron mould, shuts this together by pre ing a pedal with his foot, and holding his tube vertically, blows through it, so as to expand the cooling gla into the form of the mould. Whenever he takes his foot from the pedal-lever, the mould spontaneously opens out into two halves, and falls asunder by its bottom hinge. He then lifts the bottle up at the end of the rod, and transfers it to the finisher, who, touching the gla -tube at the end of the pipe with a cold iron, cracks off the bottle smoothly at its mouth-ring. The finished bottles are immediately piled up in the hot annealing arch, where they are afterwards allowed to cool slowly for 24 hours at least. See Bottle Mould . 2. Broad or spread window gla . —This kind of gla is called inferior window gla , in this country, because coarse in texture, of a wavy wrinkled surface, and very cheap, but on the Continent spread window gla , being made with more care, is much better than ours, though still far inferior in transparency and polish to crown gla , which has, therefore, nearly superseded its use among us. But Me rs. Chance and Hartley, of West Bromwich near Birmingham, have of late years mounted a spread-gla work, where they make British sheet gla , upon the best principles, and turn out an article quite equal, if not superior to any thing of the kind made either in France or Belgium. Their materials are those used in the crown-gla manufacture. The vitrifying mixture is fritted for 20 or 30 hours in a reverberatory arch, with considerable stirring and puddling with long-handled shovels and rakes; and the frit is then transferred by shovels while red hot, to the melting pots to be founded. When the gla is rightly vitrified, settled, and brought to a working heat, it is lifted out by iron tubes, as will be described under the article Crown Gla , blown into pears, which being elongated into cylinders, are cracked up along one side, parallel to the axis, by touching them with a cold iron dipped in water, and are then opened out into sheets. Gla cylinders are spread in France, and at West Bromwich, on a bed of smooth stone Paris-plaster, or laid on the bottom of a reverberatory arch; the cylinder being placed on its side horizontally, with the cracked line uppermost, gradually opens out, and flattens on the hearth. At one time, thick plates were thus prepared for subsequent polishing into mirrors; but the gla was never of very good quality; and this mode of making mirror-plate has accordingly been generally abandoned. The spreading furnace or oven is that in which cylinders are expanded into tables or plates. It ought to be maintained at a brisk red heat, to facilitate the softening of the gla . The oven is placed in immediate connection with the annealing arch, so that the tables may be readily and safely transferred from the former to the latter. Sometimes the cylinders are spread in a large muffle furnace, in order to protect them from being tarnished by sulphureous and carbonaceous fumes. IMG:4147767755307473660_illo0579a.png:Gla furnace Fig. 500. represents a ground plan of both the spreading and annealing furnace; fig. 501. is an oblong profile in the direction of the dotted line X X , fig. 500. a is the fire-place; b b the canals or flues through which the flame rises into both furnaces; c the spreading furnace, upon whose sole is the spreading slab. d is the cooling and annealing oven; e e iron bars which extend obliquely acro the annealing arch, and serve for resting the gla tables against, during the cooling. f f the channel along which the previously cracked cylinders are slid, so as to be gradually warmed; g the opening in the spreading furnace, for enabling the workmen to regulate the proce ; h a door in the annealing arch, for introducing the tools requisite for raising up and removing the tables. IMG:4147767755307473660_illo0579b.png:Gla transformations In forming gla -plates by the extension of a cylinder into a plane, the workman first blows the lump of gla into the shape of an oblong pear, the length of which must be nearly equal to the length of the intended plate, and its diameter such, that the circumference when developed, will be equal to the breadth of the plate. He now rests the blowing-iron on a stool or iron bar, while an a istant with a pointed iron, pierces a hole into the extreme end of the pear, in the line of the blowing-pipe. This opening is then enlarged, by introducing the blade of a pair of spring-tongs, while the gla is turned round; and by skilful management, the end of the pear is eventually opened out into a cylindrical mouth. The workman next mounts upon a stool, and holds the blowing-iron perpendicularly. The blown cylinder is now cracked off, a punto rod of iron having been previously stuck to its one end, to form a spindle for working the other by. This rod has a flat disc on its end, or three prongs, which being dipped in melted gla , are applied to the mouth of the cylinder. By this as a handle, the gla cone is carried to the fire, and the narrow end being heated, is next opened by spring tongs, and formed into a cylinder of the same size as the other end. The cylinder thus equalized, is next cracked or slit down in its side with a pair of shears, laid on a smooth copper plate, detached from the iron rod, spread out by heat into a plane surface, and finally annealed. This series of transformations, is represented in fig. 502. , at A , B , C , D , E , F , G , H . IMG:4147767755307473660_illo0579c.png:Bohemian furnace Fig. 503. and 504. represent a Bohemian furnace in which excellent white window gla is founded. Fig. 503. is a longitudinal section of the gla and annealing furnace. Fig. 504. is the ground plan. a is the ash pit vaulted under the sole of the furnace; the fireplace itself is divided into three compartments; with a middle slab at d , which is hollowed in the centre, for collecting any spilt gla , and two hearth tiles or slabs b b . c c are the draught or air holes; e e are arches upon which the bearing slabs f f partly rest. In the middle between these arches, the flame strikes upwards upon the pots g g , placed as closely together as po ible, for economy of room. h is the breast wall of the furnace; i , fig. 504. , the opening through which the pots are introduced; it is bricked up as soon as they are set. k k , is the base of the cone or dome of the furnace; l l l , the working orifices, which are made larger or smaller according to the size of the gla articles to be made. m is the flue which leads to the annealing stove n , with an arched door. Exterior to this, there is usually a drying kiln not shown in the figure; and there are adjoining stoves called arches , for drying and annealing the new pots before they are set. The cooling or annealing arch, or leer, is often built independent of the gla -house furnace, is then heated by a separate fire-place, and constructed like a very long reverberatory furnace. See Copper . The leer pans or trays of sheet iron, are laid upon its bottom in an oblong series, and hooked to each other. IMG:4147767755307473660_illo0580a.png:Crown-gla furnace 3. Crown-gla . —The crown-gla house with its furnace is represented in fig. 505. , where the blowing operation is shewn on the one side of the figure, and the flashing on the other. The furnace is usually constructed to receive 4 or 6 pots, of such dimensions as to make about a ton of gla each at a time. There are, however, several subsidiary furnaces to a crown-house. 1. A reverberatory furnace or calcar , for calcining or fritting the materials; 2. a blowing furnace, for blowing the pear-shaped balls made at the pot-holes, into large globes; 3. a flashing furnace, and bottoming hole for communicating a softening heat, in expanding the globe into a circular plate; 4. the annealing arch for the finished tables; 5. the reverberatory oven for annealing the pots prior to their being set upon the founding siege . The materials of crown gla used to be, fine sand, by measure 5 parts, or by weight 10; ground kelp by measure 11 parts, or by weight 16 1 ⁄ 2 ; but instead of kelp, soda ash is now generally employed. From 6 to 8 cwt. of sand, lime, and soda-ash, mixed together in wooden boxes with a shovel, are thrown on the sole of a large reverberatory, such as is represented in the article Copper . Here the mixture is well worked together, with iron paddles, flat shovels, and rakes with long handles; the area of this furnace being about 6 feet square, and the height 2 feet. The heat soon brings the materials to a pasty consistence, when they must be diligently turned over, to favour the di ipation of the carbon, sulphur, and other volatile matters of the kelp or soda ash, and to incorporate the fixed ingredients uniformly with the sand. Towards the end of 3 hours, the fire is considerably raised, and when the fourth hour has expired, the fritting operation is finished. The ma is now shovelled or raked out into shallow cast-iron square cases, smoothed down, and divided before it hardens by cooling, into square lumps, by cro sections with the spade. These frit-bricks are afterwards piled up in a large apartment for use; and have been supposed to improve with age, by the efflorescence of their saline constituents into carbonate of soda on their surface. The founding-pots are filled up with these blocks of frit, and the furnace is powerfully urged by opening all the subterranean pa ages to its grate, and closing all the doors and windows of the gla -house itself. After 8 or 10 hours the vitrification has made such progre , and the blocks first introduced are so far melted down, that another charge of frit can be thrown in, and thus the pot is fed with frit till the proper quantity is used. In about 16 hours the vitrification of the frit has taken place, and a considerable quantity, amounting often to the cwt. of liquid saline matter floats over the gla . This salt is carefully skimmed off into iron pots with long ladles. It is called Sandiver or Gla -gall, and consists usually of muriate of soda, with a little sulphate. The pot is now ready for receiving the topping of cullet , which is broken pieces of window gla , to the amount of 3 or 4 cwt. This is shovelled in at short intervals; and as its pre ure forces up the residuary saline matter, this is removed; for were it allowed to remain, the body of the gla would be materially deteriorated. The heat is still continued for several hours till the gla is perfect, and the extrication of gas called the boil , which accompanies the fusion of crown gla , has nearly terminated, when the fire is abated, by shutting up the lower vault doors and every avenue to the grate, in order that the gla may settle fine. At the end of about 40 hours altogether, the fire being slightly raised by adding some coals, and opening the doors, the gla is carefully skimmed, and the working of the pots commences. IMG:4147767755307473660_illo0580b.png:Base of crown-gla house cone Before describing it, however, we may state that the marginal figure 506. shews the base of the crown-house cone, with the four open pots in two ranges on opposite sides of the furnace, sitting on their raised sieges , at each side of the grate. At one side of the base the door of the vault is shewn, and its course is marked by the dotted lines. IMG:4147767755307473660_illo0581.png:Crown-gla furnace Detailed description of the crown-gla furnace , figs. 507. 508. —It is an oblong square, built in the centre of a brick cone, large enough to contain within it, two or three pots at each side of the grate room, which is either divided as shown in the plan, or runs the whole length of the furnace, as the manufacturer chooses. Fig. 507. is a ground plan, and fig. 508. a front elevation, of a six-pot furnace. 1, 2, 3, fig. 507. , are the working holes for the purposes of ventilation, of putting in the materials, and of taking out the metal to be wrought. 4, 5, 6, 7, are pipe holes for warming the pipes before beginning to work with them. 8, 9, 10, are foot holes for mending the pots and sieges. 11 is a bar of iron for binding the furnace, and keeping it from swelling. The arch is of an elliptic form; though a barrel arch, that is, an arch shaped like the half of a barrel cut longwise through the centre, is sometimes used. But this soon gives way when used in the manufacture of crown gla , although it does very well in the clay-furnace used for bottle houses. The best stone for building furnaces is fire-stone, from Cox green in the neighbourhood of Newcastle. Its quality is a close grit, and it contains a greater quantity of talc than the common fire-stone, which seems to be the chief reason of its resisting the fire better. The great danger in building furnaces is, lest the cement at the top should give way with the exce ive heat, and by dropping into the pots, spoil the metal. The top should therefore be built with stones only, as loose as they can hold together after the centres are removed, and without any cement whatever. The stones expand and come quite close together when annealing; an operation which takes from eight to fourteen days at most. There is thus le risk of any thing dropping from the roof of the furnace. The inside of the square of the furnace is built either of Stour bridge fire-clay annealed, or the Newcastle fire-stone, to the thickne of sixteen inches. The outside is built of common brick about nine inches in thickne . The furnace is thrown over an ash-pit, or cave as it is called, which admits the atmospheric air, and promotes the combustion of the furnace. This cave is built of stone until it comes beneath the grate room, when it is formed of fire-brick. The abutments are useful for binding and keeping the furnace together, and are built of masonry. The furnaces are stoutly clasped with iron all round, to keep them tight. In four-pot furnaces this is unnece ary, provided there be four good abutments. IMG:4147767755307473660_illo0582b.png:Flashing furnace Fig. 509. is an elevation of the flashing furnace. The outside is built of common brick, the inside of fire-brick, and the mouth or nose of Stour bridge fire-clay. IMG:4147767755307473660_illo0582a.png:Annealing kiln Fig. 510. is the annealing kiln. It is built of common brick, except round the grate room, where fire-brick is used. IMG:4147767755307473660_illo0582c.png:Blowing and flashing tools Few tools are needed for blowing and flashing crown-gla . The requisite ball of plastic gla is gathered, in succe ive layers as for bottles, on the end of an iron tube, and rolled into a pear-shape, on a cast-iron plate; the workman taking care that the air blown into its cavity is surrounded with an equal body of gla , and if he perceives any side to be thicker than another, he corrects the inequality by rolling it on the sloping iron table called marver, (marbre). He now heats the bulb in the fire, and rolls it so as to form the gla upon the end of the tube, and by a dexterous swing or two he lengthens it, as shewn in I , fig. 511. To extend the neck of that pear, he next rolls it over a smooth iron rod, turned round in a horizontal direction, into the shape K , fig. 511. By further expansion at the blowing-furnace, he now brings it to the shape L , represented in fig. 511. This spheroid having become cool and somewhat stiff, is next carried to the bottoming hole (like fig. 509. ), to be exposed to the action of flame. A slight wall erected before one half of this hole, screens the workman from the heat, but leaves room for the globe to pa between it and the posterior wall. The blowing-pipe is made to rest a little way from the neck of the globe, on a hook fixed in the front wall; and thus may be made easily to revolve on its axis, and by giving centrifugal force to the globe, while the bottom of it, or part opposite to the pipe, is softened by the heat, it soon a umes the form exhibited in M , fig. 511. In this state the flattened globe is removed from the fire, and its rod being rested on the casher box covered with coal cinders, another workman now applies the end of a solid iron rod tipped with melted gla , called a punto , to the nipple or prominence in the middle; and thus attaches it to the centre of the globe, while the first workman cracks off the globe by touching its tubular neck with an iron chisel dipped in cold water. The workman having thereby taken po e ion of the globe by its bottom or knobbed pole attached to his punty rod, he now carries it to another circular opening, where he exposes it to the action of moderate flame with regular rotation, and thus slowly heats the thick projecting remains of the former neck, and opens it slightly out, as shewn at N , in fig. 511. He next hands it to the flasher , who resting the iron rod in a hook placed near the side of the orifice A , fig. 509. , wheels it rapidly round opposite to a powerful flame, till it a umes first the figure O , and finally that of a flat circular table. The flasher then walks off with the table, keeping up a slight rotation as he moves along, and when it is sufficiently cool, he turns down his rod into a vertical position, and lays the table flat on a dry block of fire-clay, or bed of sand, when an a istant nips it off from the punto with a pair of long iron shears, or cracks it off with a touch of cold iron. The loose table or plate is lastly lifted up horizontally on a double pronged iron fork, introduced into the annealing arch fig. 510. and raised on edge; an a istant with a long-kneed fork preventing it from falling too rapidly backwards. In this arch a great many tables of gla are piled up in iron frames, and slowly cooled from a heat of about 600° to 100° F., which takes about 24 hours; when they are removed. A circular plate or table of about 5 feet diameter weighs on an average 9 pounds. 4. Flint gla . —This kind of gla is so called because originally made with calcined flints, as the siliceous ingredient. The materials at present employed in this country for the finest flint gla or crystal, are first, Lynn sand, calcined, sifted, and washed; second, an oxide of lead, either red lead or litharge; and third, pearl ash. The pearl ash of commerce must however be purified by digesting it in a very little hot water, which di olves the carbonate of potash, and leaves the foreign salts, chiefly sulphate of potash, muriate of potash, and muriate of soda. The solution of the carbonate being allowed to cool and become clear in lead pans, is then run off into a shallow iron boiler, and evaporated to dryne . Nitre is generally added as a fourth ingredient of the body of the gla ; and it serves to correct any imperfections which might arise from accidental combustible particles, or from the lead being not duly oxidized. The above four substances constitute the main articles; to which we may add arsenic and manganese, introduced in very small quantities, to purify the colour and clear up the transparency of the gla . The black oxide of manganese, when used in such quantity only as to peroxidize the iron of the sand, simply removes the green tinge caused by the iron; but if more manganese be added than accomplishes that purpose, it will give a purple tinge to the gla ; and in fact, most manufacturers prefer to have an exce rather than a defect of manganese, since cut gla has its brilliancy increased by a faint lilac hue. The arsenic is supposed to counteract the injury arising from exce of manganese, but is itself very apt on the other hand to communicate some degree of opalescence, or at least, to impair the lustre of the gla . When too much manganese has been added, the purple tinge may indeed be removed by any carbonaceous matter, as by thrusting a wooden rod down into the liquid gla ; but this cannot be done with good effect in practice, since the final purple tinge is not decided till the gla is perfectly formed, and then the introduction of charcoal would destroy the uniformity of the whole contents of the pot. The raw materials of flint gla , are always mixed with about a third or a fourth of their weight of broken crystal of like quality; this mixture is thrown into the pot with a shovel; and more is added whenever the preceding portions by melting subside; the object being to obtain a pot full of gla , to facilitate the skimming off the impurities, and sandiver. The mouth of the pot is now shut, by applying clay-lute round the stopper, with the exception of a small orifice below, for the escape of the liquid saline matter. Flint gla requires about 48 hours for its complete vitrification, though the materials be more fusible than those of crown gla ; in consequence of the contents of the pot being partially screened by its cover from the action of the fire, as also from the lower intensity of the heat. IMG:4147767755307473660_illo0583.png:Flint gla making Fig. 512. represents a flint gla house for 6 pots, with the arch or leer on one side for annealing the crystal ware. In fig. 513. , the base of the cone is seen, and the gla pots in situ on their platform ranged round the central fire grate. The dotted line denotes the contour of the furnace, fig. 512. Whenever the gla appears fine, and is freed from its air bubbles, which it usually is in about 36 hours, the heat is suffered to fall a little by closing the bottom valves, ., that the pot may settle; but prior to working the metal, the heat is somewhat raised again. It would be usele to describe the manual operations of fashioning the various articles of the flint-gla manufacture, because they are indefinitely varied to suit the conveniences and caprices of human society. Every different flint-house has a peculiar proportion of gla materials. The following have been offered as good practical mixtures. | 1. | Fine white sand | 300 | parts. | | Red lead or litharge | 200 | | Refined pearl ashes | 80 | | Nitre | 20 | | Arsenic and manganese, a minute quantity. | In my opinion, the proportion of lead is too great in the above recipe, which is given on the authority of Mr. James Geddes, of Leith. The gla made with it would be probably yellowish, and dull. | 2. | Fine sand | 50·5 | | Litharge | 27·2 | | Refined pearl ashes (carbonate of potash, with 5 per cent. of water) | 17·5 | | Nitre | 4·8 | | 100·0 | To these quantities from 30 to 50 parts of broken gla or cullet are added; with about a two-thousandth part of manganese, and a three-thousandth part of arsenic. But manganese varies so extremely in its purity, and contains often so much oxide of iron, that nothing can be predicated as to its quantity previously to trial. M. Payen, an eminent manufacturing chemist in France, says that the composition of crystal does not deviate much from the following proportions :— | Wood fire. | Coal fire. | | Siliceous sand | 3 | | 3 | | | Minium | 2 | | 2 | 1⁄4 | | Carbonate of potash | 1 | 1⁄2 | 1 | 2⁄3 | I conceive that this gla contains too much lead and potash. Such a mixture will produce a dull metal, very attractive of moisture: defects to which the French crown-gla also is subject. The flint-gla leer for annealing gla , is an arched gallery or large flue, about 36 feet long, 3 feet high, 4 wide; having its floor raised above 2 feet above the ground of the gla -house. The hot air and smoke of a fire-place at one end pa along this gallery, and are discharged by a chimney 8 or 10 feet short of the other end. On the floor of the vault, large iron trays are laid and hooked to each other in a series, which are drawn from the fire end towards the other by a chain, wound about a cylinder by a winch-handle projecting through the side. The flint-gla articles are placed in their hot state into the tray next the fire, which is moved onwards to a cooler station whenever it is filled, and an empty tray is set in its place. Thus, in the course of about 20 hours, the gla advances to the cool end thoroughly annealed. Besides colourle transparent gla , which forms the most important part of this manufacture, various coloured gla es are made to suit the taste of the public. The taste at Paris was lately for opaline crystal; which may be prepared by adding to the above composition (No. 2.) phosphate of lime, or well burnt bone-ash in fine powder, washed, and dried. The article must be as uniform in thickne as po ible, and speedily worked into shape, with a moderate heat. Oxide of tin, putty , was formerly used for making opalescent gla , but the lustre of the body was always impaired by its means. Crystal ve els have been made recently of which the inner surface is colourle , and all the external facets coloured. Such works are easily executed. The end of the blowing-rod must be dipped first in the pot containing colourle gla , to form a bulb of a certain size, which being cooled a little is then dipped for an instant into the pot of coloured gla . The two layers are a ociated without intermixture; and when the article is finished in its form, it is white within and coloured without. Fluted lines somewhat deeply cut, pa through the coloured coat, and enter the colourle one; so that when they cro , their ends alone are coloured. For some time past, likewise, various crystal articles have been exhibited in the market with coloured enamel-figures on their surface, or with white incrustations of a silvery lustre in their interior. The former are prepared by placing the enamel object in the bra mould, at the place where it is sought to be attached. The bulb of gla being put into the mould, and blown while very hot, the small plate of enamel gets cemented to the surface. For making the white argentine incrustations, small figures are prepared with an impalpable powder of dry porcelain paste, cemented into a solid by means of a little gypsum plaster. When these pieces are thoroughly dried, they are laid on the gla while it is red hot, and a large patch of very liquid gla is placed above it, so as to encase it and form one body with the whole. In this way the incrustation is completely enclosed; and the polished surface of the crystal which scarcely touches it, gives a brilliant aspect, pleasing to the eye. An uniform flint-gla , free from str iæ, or wreath , is much in demand for the optician. It would appear that such an article was much more commonly made by the English manufacturers many years ago, than at present; and that in improving the brilliancy of crystal-gla they have injured its fitne for constructing optical lenses, which depends not so much on its whitene and lustre as on the layers of different densities being parallel to each other. The oxide of lead existing in certain parts of a potful of gla in greater proportion than in other parts, increases the density unequally in the same ma , so that the adjoining strata are often very different in this respect. Even a potful of pretty uniform gla , when it stands some time liquid, becomes eventually unequable by the subsidence of the denser portions; so that str iæ and gelatinous appearances begin to manifest themselves, and the gla becomes of little value. Gla allowed to cool slowly in ma in the pot is particularly full of wreath; and if quickly refrigerated, that is in two or three hours, it is apt to split into a multitude of minute splinters, of which no use can be made. For optical purposes, the gla must be taken out in its liquid state, being gathered on the end of the iron rod from the central portion of a recently skimmed pot, after the upper layers have been worked off in general articles. M. Guinand, of Brennets near Geneva, appears to have hit upon proce es that furnished almost certainly pieces of flint-gla capable of forming good lenses of remarkable dimensions, even of 11 inches diameter; of adequate density and transparency, and nearly free from str iæ . M. Cauchoix, the eminent French optician, says, that out of ten object gla es, 4 inches in diameter, made with M. Guinand’s flint-gla , eight or nine turned out very good, while out of an equal number of object gla es made of the flint-gla of the English and French manufactories, only one, or two at most, were found serviceable. The means by which M. Guinand arrived at these results have not been published. He has lately died, and it is not known whether his son be in po e ion of his secret. An achromatic object gla for telescopes and microscopes consists of at least two lenses; the one made with gla of lead, or flint gla , and the other with crown gla ; the former po e ing a power of dispersing the coloured rays relatively to its mean refractive power, much greater than the latter; upon which principle, the achromatism of the image is produced, by re-uniting the different coloured rays into one focus. Flint gla to be fit for this delicate purpose must be perfectly homogeneous, or of uniform density throughout its substance, and free from wavy veins or wreathes; for every such inequality would occasion a corresponding inequality in the refraction and dispersion of the light; like what is perceived in looking through a thick and thin solution of gum Arabic imperfectly mixed. Three plans have been prescribed for obtaining homogeneous pieces of optical gla : 1. to lift a ma of it in large ladles, and let it cool in them; 2. to pour it out from the pots into moulds; 3. to allow it to cool in the pots, and afterwards to cut it off in horizontal strata. The last method, which is the most plausible, seldom affords pieces of uniform density, unle peculiar precautions have been adopted to settle the flint gla in uniform strata; because its materials are of such unequal density, the oxide of lead having a specific gravity of 8, and silica of 2·7, that they are apt to stand at irregular heights in the pots. One main cause of these inequalities lies in the construction of the furnace, whereby the bottom of the pot is usually much le heated than the upper part. In a plate gla furnace the temperature of the top of the pot has been found to be 130° Wedgew., while that of the bottom was only 110°, constituting a difference of no le than 2610° F. The nece ary consequence is that the denser particles which subside to the bottom, during the fusion of the materials, and after the first extrication of the gases, must remain there, not being duly agitated by the expansive force of caloric, acting from below upwards. The preparation of the best optical gla is now made a great mystery by one or two proficients. The following suggestions, deduced from a consideration of principles, may probably lead to some improvements, if judiciously applied. The great object is to counteract the tendency of the gla of lead to distribute itself into strata of different densities; which may be effected either by mechanical agitation or by applying the greatest heat to the bottom of the pot. But however homogeneous the gla may be thereby made, its subsequent separation into strata of different densities must be prevented by rapid cooling and solidification. As the deeper the pots, the greater is the chance of unequal specific gravity in their contents, it would be advisable to make them wider and shallower than those in use for making ordinary gla . The intermixture may be effected either by lading the gla out of one pot into another in the furnace, and back again, with copper ladles, or by stirring it up with a rouser, then allowing it to settle for a short time, till it becomes clear and free from air bubbles. The pot may now be removed from the furnace, in order to solidify its contents in their homogeneous state; after which the gla may be broken in pieces, and be perfected by subjecting it to a second fusion; or what is easier and quicker, we may form suitable discs of gla without breaking down the potful, by lifting it out in flat copper ladles with iron shanks, and transferring the lumps after a little while into the annealing leer. To render a potful of gla homogeneous by agitation, is a more difficult task, as an iron rod would discolour it, and a copper rod would be apt to melt. An iron rod sheathed in laminated platinum would answer well, but for its expense. A stone-ware tube supported within by a rod of iron, might also be employed for the purpose in careful hands; the stirring being repeated several times, till at last the gla is suffered to stiffen a little by decrease of temperature. It must be then allowed to settle and cool, after which the pot, being of small dimensions, may be drawn out of the fire. IMG:4147767755307473660_illo0585.png:Gla furnace 2. The second method of producing the desired uniformity of mixture, consists in applying a greater heat to the bottom than to the upper part of the melting pot. Fig. 514. represents in section a furnace contrived to effect this object. It is cylindrical, and of a diameter no greater than to allow the flames to play round the pot, containing from three to four cwts. of vitreous materials. A is the pot, resting upon the arched grid b a , built of fire-bricks, whose apertures are wide enough to let the flames rise freely, and strike the bottom and sides of the ve el. From 1 1 ⁄ 2 to 2 feet under that arch, the fuel grate c d is placed. B C are the two working openings for introducing the materials, and inspecting the progre of the fusion; they must be closed with fire-tiles and luted with fire-clay at the beginning of the proce . At the back of the furnace, opposite the mouth of the fire-place, there is a door-way, which is bricked up, except upon occasion of putting in and taking out the pot. The draught is regulated by means of a slide-plate upon the mouth of the ash-pit f . The pot being heated to the proper pitch, some purified pearl ash, mixed with fully twice its weight of colourle quartz sand, is to be thrown into it, and after the complete fusion of this mixture, the remaining part of the sand along with the oxide of lead (fine litharge) is to be strown upon the surface. These siliceous particles in their descent serve to extricate the air from the ma . Whenever the whole is fused, the heat must be strongly urged to ensure a complete uniformity of combination by the internal motions of the particles. As soon as the gla has been found by making test phials to be perfectly fine, the fire must be withdrawn, the two working holes must be opened, as well as the mouths of the fire-place and ash-pit, to admit free ingre to cooling currents of air, so as to congeal the liquid ma as quickly as po ible; a condition e ential to the uniformity of the gla . It may be worth while to stir it a little with the pottery rod at the commencement of the cooling proce . The solidified gla may be afterwards detached by a hammer in conchoidal discs, which after chipping off their edges, are to be placed in proper porcelain or stone-ware dishes, and exposed to a softening heat, in order to give them a lenticular shape. Great care must be taken that the heat thus applied by the muffle furnace be very equable, for otherwise wreathes might be very readily re-produced in the discs. A small oven upon the plan of a baker’s, is best fitted for this purpose, which being heated to dull redne , and then extinguished, is ready to soften and afterwards anneal the conchoidal pieces. Guinand’s dense optical flint gla , of specific gravity 3·616, consists by analysis, of oxide of lead 43·05; silica 44·3; and potash 11·75; but requires for its formation the following ingredients: 100 pounds of ground quartz; 100 pounds of fine red lead; 35 pounds of purified potash; and from 2 to 4 pounds of saltpetre. As this species of gla is injured by an exce of potash, it should be compounded with rather a defect of it, and melted by a proportionally higher or longer heat. A good optical gla has been made in Germany with 7 parts of pure red lead, 3 parts of finely ground quartz, and 2 parts of calcined borax. 5. Plate gla . This, like English crown-gla , has a soda flux, whereas flint-gla requires potash, and is never of good quality when made with soda. We shall distribute our account of this manufacture under two heads. 1. The different furnaces and principal machines, without whose knowledge it would be impo ible to understand the several proce es of a plate-gla factory. 2. The materials which enter into the composition of this kind of gla , and the series of operations which they undergo; devoting our chief attention to the changes and improvements which long experience, enlightened by modern chemistry, has introduced into the great manufactory of Saint-Gobin in France, under the direction of M. Ta aert. It may however be remarked that the English plate-gla manufacture derives peculiar advantages from the excellence of its grinding and polishing machinery. The clay for making the bricks and pots should be free from lime and iron, and very refractory. It is mixed with the powder of old pots pa ed through a silk sieve. If the clay be very plastic it will bear its own weight of the powder, but if shorter in quality, it will take only three-fifths. But before mingling it with the cement of old pots, it must be dried, bruised, then picked, ground, and finally elutriated by agitation with water, decantation through a hair sieve, and subsidence. The clay fluid after pa ing the sieve is called slip (coulis.) The furnace is built of dry bricks, cemented with slip, and has at each of its four angles a peculiar annealing arch, which communicates with the furnace interiorly, and thence derives sufficient heat to effect in part, if not wholly, the annealing of the pots, which are always deposited there a long time before they are used. Three of these arches exclusively appropriated to this purpose, are called pot-arches. The fourth is called the arch of the materials , because it serves for drying them before they are founded. Each arch has, moreover, a principal opening called the throat, another called bonnard , by the French workmen, through which fire may be kindled in the arch itself, when it was thought to be nece ary for the annealing of the pots; a practice now abandoned. The duration of a furnace is commonly a year, or at most 14 months; that of the arches is 30 years or upwards, as they are not exposed to so strong a heat. In the manufacture of plate-gla two sorts of crucibles are employed, called the pots and the basins, ( cuvettes ). The first serve for containing the materials to be founded, and for keeping them a long time in the melted state. The cuvettes receive the melted gla after it is refined, and decant it out on the table to be rolled into a plate. Three pots hold liquid gla for six small basins, or for three large ones, the latter being employed for making mirrors of great dimensions, that is, 100 inches long and upwards. Furnaces have been lately constructed with 6 pots, and 12 cuvettes, 8 of which are small, and 4 large; and cuvettes of three sizes are made, called small , middling , and large . The small are perfect cubes, the middling and the large ones are oblong parallelopipeds. Towards the middle of their height, a notch or groove, two or three inches broad, and an inch deep, is left, called the girdle of the cuvette, by which part they are grasped with the tongs, or rather are clamped in the iron frame. This frame goes round the four sides of the small cuvettes, and may be placed indifferently upon all their sides; in the other cuvettes, the girdle extends only over the two large sides, because they cannot be turned up. See m T , fig. 515. , p. 590 . The pot is an inverted truncated cone, like a crown gla pot. It is about 30 inches high, and from 30 to 32 inches wide, including its thickne . There is only a few inches of difference between the diameter of the top and that of the bottom. The bottom is 3 inches thick, and the body turns gradually thinner till it is an inch at the mouth of the pot. The large building or factory, of which the melting furnace occupies the middle space, is called the halle in French. At Raven head in Lancashire it is called the foundry, and is of magnificent dimensions, being probably the largest apartment under one roof in Great Britain, since its length is 339 feet, and its breadth 155. The famous halle of St. Gobin is 174 feet by 120. Along the two side walls of the halle , which are solidly constructed of hewn stone, there are openings like those of common ovens. These ovens, destined for the annealing of the newly cast plates, bear the name of carquaises . Their soles are raised two feet and a half above the level of the ground, in order to bring them into the same horizontal plane with the casting tables. Their length, amounting sometimes to 30 feet, and their breadth to 20, are required in order to accommodate 6, 8, or even 10 plates of gla , alongside of each other. The front aperture is called the throat, and the back door the little throat ( gueulette ). The carquaise is heated by means of a fire-place of a square form called a tisar , which extends along its side. The founding or melting furnace is a square brick building laid on solid foundations, being from 8 to 10 feet in each of its fronts, and rising inside into a vault or crown about 10 feet high. At each angle of this square, a small oven or arch is constructed, likewise vaulted within, and communicating with the melting furnace by square flues, called lunettes , through which it receives a powerful heat, though much inferior to that round the pots. The arches are so distributed as that two of the exterior sides of the furnace stand wholly free, while the two other sides, on which the arches encroach, offer a free space of only 3 feet. In this interjacent space, two principal openings of the furnace, of equal size in each side, are left in the building. These are called tunnels. They are destined for the introduction of the pots and the fuel. On looking through the tunnels into the inside of the furnace, we perceive to the right hand and the left, along the two free sides, two low platforms or sieges , at least 30 inches in height and breadth. See figs. 506. 508. These sieges (seats) being intended to support the pots and the cuvettes filled with heavy materials, are terminated by a slope, which ensures the solidity of the fire-clay mound. The slopes of the two sieges extend towards the middle of the furnace so near as to leave a space of only from 6 to 10 inches between them for the hearth. The end of this is perforated with a hole sufficiently large to give pa age to the liquid gla of a broken pot, while the rest is preserved by lading it from the mouth into the adjoining cuvette. In the two large parallel sides of the furnace, other apertures are left much smaller than the tunnels, which are called ouvreaux (peep holes). The lower ones, or the ouvreaux en bas , called cuvette openings, because being allotted to the admi ion of these ve els, they are exactly on a level with the surface of the sieges , and with the floor of the halle. Plates of cast iron form the thresholds of these openings, and facilitate the ingre and egre of the cuvettes. The apertures are arched at top, with hewn stone like the tunnels, and are 18 inches wide when the cuvettes are 16 inches broad. The upper and smaller apertures, or the higher ouvreaux called the lading holes, because they serve for trans vas ing the liquid gla , are three in number, and are placed 31 or 32 inches above the surface of the sieges . As the pots are only 30 inches high, it becomes easy to work through these openings either in the pots or the cuvettes . The pots stand opposite to the two pillars which separate the openings, so that a space is left between them for one or more cuvettes according to the size of the latter. It is obvious that if the tunnels and ouvreaux were left open, the furnace would not draw or take the requisite founding heat. Hence the openings are shut by means of fire-tiles. These are put in their places, and removed by means of two holes left in them, in correspondence with the two prongs of a large iron fork supported by an axle and two iron wheels, and terminated by two handles which the workmen lay hold of when they wish to move the tile. The closing of the tunnel is more complex. When it is shut or ready for the firing, the aperture appears built up with bricks and mortar from the top of the arch to the middle of the tunnel. The remainder of the door-way is closed; 1. on the two sides down to the bottom, by a small upright wall, likewise of bricks, and 8 inches broad, called walls of the glaye ; 2. by an a emblage of pieces called pieces of the glaye , because the whole of the closure of the tunnel bears the name of glaye . The upper hole, 4 inches square, is called the tisar , through which billets of wood are to ed into the fire. Fuel is also introduced into the posterior openings. The fire is always kept up on the hearth of the tunnel, which is, on this account, 4 inches higher than the furnace-hearth, in order that the gla which may accidentally fall down on it, and which does not flow off by the bottom hole, may not impede the combustion. Should a body of gla , however, at any time obstruct the grate, it must be removed with rakes, by opening the tunnel and dismounting the fire-tile stoppers of the glaye . Formerly wood fuel alone was employed for heating the melting-furnaces of the mirror-plate manufactory of Saint-Gobin; but within these few years, the Director of the works makes use with nearly equal advantage of pit-coal. In the same establishment, two melting furnaces may be seen, one of which is fixed with wood, and the other with coals, without any difference being perceptible in the quality of the gla furnished by either. It is not true, as has been stated, that the introduction of pit-coal has made it nece ary to work with covered pots in order to avoid the discoloration of the materials, or that more alkali was required to compensate for the diminished heat in the covered pots. They are not now covered when pit-coal is used, and the same succe is obtained as heretofore by leaving the materials two or three hours longer in the pots and the cuvettes. The construction of the furnaces in which coal is burned, is the same as that with wood, with slight modifications. Instead of the close bottomed hearth of the wood furnace, there is an iron grate in the coal-hearth through which the air enters, and the waste ashes descend. When billets of wood were used as fuel, they were well dried beforehand, by being placed a few days on a frame work of wood called the wheel, placed two feet above the furnace and its arches, and supported on four pillars at some distance from the angles of the building.