British Manufacturing Industries Part 4

When it is desired to apply to any portion of white gla.s.s some yellow silver stain, this can be done either in the first firing, by floating it on to the places to be stained, and allowing it to run in a sort of stream from the brush, so that it will evenly cover the surface and cause the heavier portions of the stain, namely, the mixed metallic silver and antimony, to sink regularly to the bottom, and come fairly in contact with the gla.s.s. Not very long ago, it was mentioned to me by a gla.s.s painter of note, that the workmen much prefer using the old stain made with silver and antimony, to that which is produced by using nitrate of silver. This really is a mistake on their part, for, when properly managed (and the knowledge of how to manage this stain can be acquired with very little trouble), the nitrate of silver stain is by far the best, and produces much better tints, with less chance of what the men call sulphuring when the gla.s.s is fired. This sulphuring is simply the result of opacity, obtained by heating the gla.s.s to too high a temperature. If the staining is to be performed in the same firing as that by which the painting is to be fixed, it is quite clear that the outlines of the part to be stained must be painted in, with tar-oil, or with some such substance which is not affected by the moisture of the stain. However, in general, the staining operation is performed after the first firing, that is to say, those pieces of gla.s.s to which the silver is to be applied are stained in the method above described after the first firing, and are then fired again, because the heat required to produce a good stain from silver is of a somewhat different character from that which is required simply to fuse the flux that binds the pigment to the gla.s.s.

A longer and less intense heat, technically called a "soaking," is the best for producing an even and pure yellow tint. If the temperature be allowed to rise too high, the oxide of silver, which alone can stain the gla.s.s, gets reduced wholly or in part, and when this happens to only a slight extent, it destroys the transparency of the stain; and when it happens to a great extent, it destroys its colour altogether, making the gla.s.s opaque.

It is a matter of astonishment to me that gla.s.s painters do not use a ruby stain, which, with a little practice, can be managed quite as successfully as the yellow silver one. It is true that it would be impossible to fire the ruby and the silver stains together, and it would not be at all convenient to fire the ruby stain at the first firing of the painted gla.s.s. The method of staining ruby is as follows: grind up carefully some black oxide of copper, mix it with water (or with a small quant.i.ty of gum added), float it on the parts to be coloured, place it in a kiln and heat it. Black oxide of copper, when mixed with gla.s.s and melted in a gla.s.s-pot, makes the gla.s.s green; suboxide of copper, which contains less oxygen than the black oxide, when treated in the same way, makes it red. Now, if it can be reduced to the lower oxide of copper, while the black oxide of copper on the surface of the gla.s.s is heated, it will then colour the gla.s.s red. The best way of reducing the black oxide, is to connect the m.u.f.fle with a gas-supply pipe, and allow coal gas to pa.s.s during the whole time that the heating process goes on. The action of the gas, which contains hydrogen and carbon, is to take away oxygen from the black oxide of copper, when it is at a high temperature; and, as soon as sufficient is taken away by the hydrogen to reduce the black oxide to the state of suboxide, it stains the gla.s.s red. It does not matter if the reducing action be continued longer, so that the oxide of copper be reduced to the metallic state; for at that temperature, the stain produced by the red oxide of copper is not removed by the continued action of hydrogen gas. The employment of this process would certainly enable artists who paint in the later styles of gla.s.s painting, to very much enrich their draperies, and to produce, more easily, effects which now can only be obtained by a complicated system of lead-work.

When the pieces of gla.s.s which have been fired are perfectly cold, the next process is to unite them altogether by peculiarly shaped strips of lead, which are of various kinds, according to the character of the subject required. The lead has a thick part or core, and at right angles to the top and bottom of this are thin plates called the "leaves." The core is milled with little ridges running at right angles to them, so as to enable the workman to bend the lead about with facility. The edges of the piece of gla.s.s to be leaded are placed between the leaves and resting upon the core, and the lead is thus arranged all round the gla.s.s, and is then laid in its proper situation upon another cartoon, prepared from the one from which the figure was painted, and indicating simply, by lines, where the lead-work is to come. The first piece is fixed by means of nails temporarily placed through the lead. Those pieces which touch it in the design are put in their proper positions, so that the edge touching the next piece will be underneath the opposite leaves to those which confine the first.

This operation is repeated, till all the parts of the design are surrounded by lead, and by it united to one another; the joints being secured by solder, generally applied by gas. Nothing now remains but to fill in the interstices between the lead and the gla.s.s, so as to make the window firm, solid, and water-tight; and this is done by rubbing into them with a scrubbing brush a cement, usually made of white lead, oil, and plaster of Paris. This composition varies in different stained gla.s.s works, nor is it material, provided that the substance hardens, does not crack, and is waterproof.

From this description it will be seen, that the various colours in the different parts of the window are put in as pieces, and that no colours, properly so called, are applied by the brush to the surface.

There are, however, certain tints of the "tracing brown," which can be obtained by the addition of black oxide of manganese, or by a different method of preparation of the oxide of iron, to give it its body. Sulphate of iron, when heated, loses its sulphuric acid, and the oxide, which was, as sulphate, in the state of protoxide, becomes, by heating, the red or peroxide of iron; its tint, when made in this way, being generally lighter than the tint of that form of oxide which is employed as ordinary tracing brown. It is sometimes called flesh tint, though this is decidedly an objectionable name for it.

It has been suggested to me, that I should give some receipts for the manufacture of the enamel colours used in mediaeval gla.s.s painting; I have therefore added a few which are easily prepared. Others of a more complicated nature had much better be obtained from the makers of the enamel used in porcelain painting. And here again, let me remark, that in ordering fluxes from these manufacturers, it should be stated especially that a flux is required which does not contain borax, nor should the painters in any establishment be allowed to use these softer fluxes, which they are almost certain to do, unless forbidden; for though they are easier to work with, they will infallibly lead to calamitous results.

YELLOW.

Oxide of tin 2 parts.

Oxide of antimony 2 "

Red Lead 16 "

ORANGE.

Red Lead 12 "

Oxide of antimony 4 "

Persulphate of iron 1 "

Flint powder 3 "

BROWN.

Black oxide of manganese 225 "

Flint slate (powdered) 40 "

Red lead 85 "

BROWN RED.

Crocus (oxide of iron) 3 "

Green sulphate of iron (calcined) 1 "

mixed with six parts of these No. 2.

LIGHT RED FOR FLESH TINTS.

Carbonate of lead 15 "

Persulphate of iron (calcined) 1 "

Flint gla.s.s 3 "

The use of enamels--that is, substances which impart various colours to the gla.s.s, when placed on its surface by their fusion--is not admissible in windows which pretend to belong to any of the earlier styles of gla.s.s painting; though enamel painting is used for the decoration of houses, and sometimes, as I consider very improperly, for the decoration of church windows. One sheet of gla.s.s, colourless and transparent, or it may have its surface ground, is usually employed. A subject is painted on it with enamel colours, much as subjects are painted upon porcelain. When the work is completed, the gla.s.s plate is fired, and thus the colours become semi-transparent, and perfectly adherent to the plate; but they are not clear and bright, and transparent, as are the colours of gla.s.s which is coloured in the pot, and therefore have not the same brilliancy, nor do they allow of the same bold and effective treatment.

It is much to be desired that amateurs who can draw, and who have a feeling for this particular style of art, should devote a portion of their time to its execution. They will find it to be extremely agreeable and pleasant, and the few difficulties which they meet with in their first attempts will be readily overcome by perseverance, or by applying for a.s.sistance and advice to gentlemen engaged in the pursuit of this interesting profession.

_Moulded and Cut Gla.s.s._--Flint gla.s.s is now very commonly blown in moulds, and this art has been brought to such perfection that moulded decanters and tumblers have an appearance very similar to that of cut gla.s.s. The moulds are always made of metal, and so constructed, that they open out into two or more pieces, which are generally hinged to the bottom of the mould. The workman places it on the ground, and fixes it by standing on projections from its side. He then gathers a suitable quant.i.ty of gla.s.s on the end of his blowpipe, which he places in the mould, and the side of the gla.s.s touching it will thus have impressed upon it whatever form is engraved on it. After the gla.s.s has become hard, the mould is opened, and the gla.s.s vessel is removed and annealed.

When it is desired to cut a design on the outside of a tumbler or wine-gla.s.s, the vessel is, in the first instance, blown of a thicker substance than if it is to be left uncut. The necessary shapes, which are usually in facets, are cut upon it by the action of sand and water, a lathe of a very simple construction being used to give a rotary motion to cutting discs, made of stone and kept continually moist by water dripping on them, so that when the gla.s.s is pressed against them, the required portion of its surface is worn away. The usual diameter of these stones is about 10 inches. After the rougher stone has been used, a finer kind of sandstone disc is employed, or a disc of slate, upon which sand and water are allowed to drop, and the already roughly cut surface is, by their action, partly polished.

Copper discs with flattened circ.u.mference are used for polishing the gla.s.s, and for this purpose, emery mixed with oil, is applied to the edges of their circ.u.mference.

_Ground Gla.s.s_ is made by rubbing the surface of gla.s.s with sand and water, just as in the first operation of plate gla.s.s polishing.

But a very ingenious method is now generally adopted for grinding gla.s.s, by placing it in a cradle, so that it can swing from side to side; sand and water are placed upon the gla.s.s, and it grinds itself, so to speak, by this operation.

_Annealing and Devitrification._--As the word "annealing" has been often used in this article, it will be well to explain what is its action. If a piece of molten gla.s.s be dropped into water, it will a.s.sume an oblong shape, the lower end of which will be round, while the other will taper off into a fine point. These drops, which have received the name of Prince Rupert's drops, look like pieces of ordinary gla.s.s, and if the small end of one of them be broken off, a sort of explosion takes place, and the whole ma.s.s flies into a thousand minute pieces, some of which will be found to have been driven to a considerable distance. Here then it appears, that when the skin, which is perfect and entire in the Rupert drop, is broken, the bond which held together the const.i.tuent particles is broken also, and so they are acted on by a repellent force, and fly away from one another. If hot water be poured into a thick common tumbler, it very generally cracks it: but if the tumbler be thin and of better manufacture, it will bear almost boiling water without cracking. In the first case it has been badly annealed; and besides this, gla.s.s being a bad conductor of heat, from its thickness, the heat imparted by the hot water expands the inner surface, while the outer coating, not being warmed, does not expand, and, retaining its original form, is burst. If, however, a tumbler be thick and properly annealed, there is not so much danger of its breaking, when a portion of it is exposed to a considerable rise of temperature. In the case of the Rupert drops, they are not annealed at all, and so there is no cohesive bond between the particles, such as there would be if they were properly annealed, that is, if, instead of being cooled suddenly from the molten state, they were allowed to cool in a heated chamber very slowly. After gla.s.s has been heated, the particles of which it is composed take a long time to rearrange themselves, so that in the manufacture of thermometers, it is necessary, after sealing up the bulb and tube which contain the mercury, to allow them to remain for a long time; otherwise the pressure of the air on the outside of the bulb, not being supported by any air on the inside, causes the particles of gla.s.s to become more compact, and thus renders the capacity of the thermometer bulb and tube smaller than it was, when the thermometer was first sealed. It seems that the process of annealing gla.s.s gives time for the particles to arrange themselves in such a way, that when the gla.s.s is cold, it will not be so liable to fracture from sudden changes of temperature.

Considerable curiosity has been excited of late by a new invention, which has resulted from the investigations of a Frenchman. We have been told that tumblers and wine-gla.s.ses, and other gla.s.s utensils, could be so treated that they would never break; and experiments performed upon many samples of these gla.s.ses led one to suppose, that the object had been attained. There is no doubt whatever, that some who have had experience of what is termed toughened gla.s.s know, that in many cases very uncertain results are obtained in the resisting power of the gla.s.s to the action of a violent blow. Before, however, entering into some researches which I have made on the subject, it will be well to state what is the nature of the change which the toughening process produces in the gla.s.s, and this seems to be a fit place for this consideration, as the method of making, and the behaviour, of Prince Rupert's drops, have just been discussed.

The physical properties of these Rupert's drops have been examined with great care by M. Victor de Luynes, and the results of his experiments have been communicated to the _Societe de Secours des Amis des Sciences_. For the purposes of this article, many of his experiments have been repeated, confirming in general his observations, and others have also been inst.i.tuted. The toughness and hardness of these drops are remarkable; the thick pear-shaped portion will bear a sharp stroke with a hammer without breaking; nor can it be scratched with a diamond. To break the tapering thread or tail, as it may be conveniently called, requires considerable force. To find out what weight was required to do this, a series of experiments was performed, the results of which are given in the table following. The tail of a drop was placed over a small hole bored in the top of a table; a hook was then adjusted round a part of the tail which measured 19 on a Birmingham wire gauge; below the table and attached to this hook, a scale-pan was hung. This pan was then carefully loaded, all shock being avoided, until the thread was ruptured and the weight required to effect this was then noted:

White Gla.s.s Rupert's Drops.

Gauge. Strain.

19 16 lb. 0 oz.

19 15-1/2 " 0 "

19 16 " 0 "

19 (poor) 9-1/4 " 0 "

Green Gla.s.s.

Gauge. Strain.

19 18-3/4 " 0 "

19 (poor) 9 " 0 "

19 28 " 6 "

16 26-1/4 " 0 "

It will be observed that the drops made from green bottle gla.s.s withstood a greater strain than those made from crown gla.s.s; the latter, in fact, did not break throughout their ma.s.s, but left a portion of the bulb unbroken, showing some fault in the tempering. It was with difficulty that the workmen could be induced to make drops out of this kind of gla.s.s, as they knew by experience that they usually failed to break perfectly, and they stated that it was quite impossible to make them with lead gla.s.s. To ascertain what force was required to fracture a thread of like dimensions that had not been tempered, one of the drops was heated to redness, and annealed by allowing it to cool very gradually. When subjected to the same trial, it was fractured by a weight of 12 ozs., and the drop did not break into small fragments, but behaved exactly like ordinary gla.s.s, thus showing that the gla.s.s had been _un_tempered by the heating process. A piece of gla.s.s rod, drawn out into a thread in a gas flame, when subjected to the same conditions, bore a strain of 10 oz. A sewing-needle of the same thickness was broken by a weight of 3 lb. 14 oz., thus showing that the tail of the Rupert's drop was very much tougher than tempered steel. By suspending a Rupert's drop in such a manner as to allow the tail to dip into hydrofluoric acid, it is found, that when the surface or skin is eaten away to a certain depth, rupture takes place exactly in the same manner as when the tail is broken. In whatever way fractured, the particles, when examined by the microscope, show a crystalline structure, and do not at all resemble pieces of ordinary gla.s.s; when rubbed between the palms of the hands, they do not cut, nor scratch, nor penetrate the cuticle. If a drop be enclosed in plaster of Paris so as to leave a portion of the tail exposed, it may then be broken and all the particles will remain _in situ_. On removing the plaster, it will be found that the drop has been broken up into thousands of minute needle-shaped particles arranged in cones, the apices being in the direction of the tail. It would appear then from these experiments, and from observations with polarized light, that the gla.s.s in the interior of a Rupert's drop exists under enormous tension, and that it is only prevented from bursting into fragments by the outer skin; on its being broken in any part, the bond which holds together the const.i.tuent particles is broken also, and so, being acted upon by a repellent force, they fly away from one another. There is another kind of toy resembling in some respects the Rupert's drop, known as the Bologna bottle or philosopher's flask. It has the form of a soda-water bottle with the neck cut off, the bottom being rounded off and very much thicker than the walls. These flasks are sometimes formed accidentally in gla.s.s-works by the workman, who, in order to examine the quality of the gla.s.s, takes out a portion from the pot on the end of his blowpipe, and blows a small quant.i.ty of air into the ma.s.s, manipulating it in the usual manner. Whilst still at a very high temperature, it is detached from the blowpipe, and is probably allowed to fall on the ground in a place where there is a current of cold air, the exterior thus becoming suddenly chilled. When cold, these flasks will bear very rough handling, and will withstand the blow of a hammer on the outside, it being almost impossible to break them by striking the bottom; the interior will also bear the blow of a leaden bullet falling into it from a considerable height, but if a few grains of sand be allowed to fall into it, or if the inside skin be slightly scratched, the ma.s.s splits into fragments in the same manner as a Rupert's drop. The examination of these curious phenomena leads us to the subject of "toughened gla.s.s," as it has been termed. The invention of rendering articles of gla.s.s less fragile, which has given rise to so much public attention during the last year, is due to M. Alfred de la Bastie, a French engineer. His process consists in heating the gla.s.s to be toughened to a temperature close upon its softening point, and then plunging it into a bath of oil, or into a mixture of oleaginous substances kept at a much lower temperature. When this operation is successfully performed, the gla.s.s acquires properties very similar to those of Rupert's drops; it becomes much less fragile than ordinary gla.s.s, but when sufficient force is employed to fracture it, the whole flies into small pieces. It cannot be cut with a diamond, but is immediately disintegrated when the outer skin is scratched to a certain depth.

It is to be observed, however, that in particular cases it is possible both to saw and pierce the toughened gla.s.s. M. de Luynes reports, that when a square of St. Gobain plate gla.s.s that had been submitted to the process of tempering was examined by polarized light, it showed the appearance of a black cross, the arms of which were parallel to the sides of the square. The gla.s.s was sawed in two, along the line of the stem of the cross, without causing fracture. On examining the divided gla.s.s with polarized light, black bands and fringes of colour were observed, which, by their position, proved that the molecular condition of the gla.s.s had changed; on placing one half of the divided gla.s.s on the other half, the fringes and black bands disappeared--on folding one half on to the other, the black bands presented the appearance that would have been produced by gla.s.s of double the thickness. These facts show, that the molecular forces on the gla.s.s were arranged symmetrically in reference to the line of parting: and we may conclude that toughened gla.s.s being in a state of tension, similar to that of the Rupert drop, may be divided or pierced, provided that the molecules of the pieces produced are able to rearrange themselves into a stable equilibrium. Polarized light shows the directions on which the division can be made with safety.

M. de Luynes, in his communication referred to above, gives an account of some experiments performed on plates of gla.s.s of the same quality, tempered by this process, and untempered; one or two examples will suffice. A tempered plate measuring about[1] 6-1/2 inches by 5 inches, and 2/10 inch thick, was placed between two wooden frames, and a weight of over 3-1/2 ounces (100 grammes[2]) was allowed to drop upon it from a height of more than 13 feet (4 metres[3]) without breaking it. It only broke, when double the weight was employed from the same height. A piece of ordinary gla.s.s under the same conditions broke, with the weight of 3-1/2 oz. dropped upon it from a height 16 inches (040 metre). Plates of toughened gla.s.s were allowed to fall on the floor from a height, or were thrown to a distance, without breaking. A rectangular piece of ordinary window gla.s.s, about 1/10 inch in thickness, was bent into the form of a bridge, and then subjected to the tempering process; placed upon the ground; it bore the weight of a man easily without breaking. A commission, inst.i.tuted by the French naval authorities, to inquire into this process of M. de la Bastie, has reported at some length on the subject. The following series of experiments were tried with a view of ascertaining the comparative power of resistance of tempered and ordinary gla.s.s. The plates experimented upon were placed loosely in wooden frames constructed for the purpose.

[1] These numbers are approximate translations of the numbers given in the communication: no object could be gained in giving complex fractions.

[2] 1 ounce avoirdupois weighs 28349 grammes.

[3] 1 metre equals 3937 English inches.

_Rectangular plates about 21 inches_ (0525 m.) _by 10 inches_ (0248 m.) _and 1/6 inch_ (0004 m.) _thick_.

The frame with the gla.s.s inserted was laid on the ground, and in the middle of the plate a weight of more than 10 lbs. (5 kilogrammes[4]) was placed, and upon it as a base, other weights were placed, care being taken to avoid all shock.

[4] 1 kilogramme = 22 lbs. avoirdupois.

1 _Ordinary gla.s.s_, broke with a weight of about 70 lb. (35 kilos.) having resisted weights of from 30 to 50 lb.

2 _Toughened gla.s.s_ resisted fracture until a weight of more than 510 lb. (255 kilos.) had been added, and then was not broken. The experiment was not carried to its limit for want of weights.