The process of manufacturing the best qualities of optical glass may be briefly described as consisting in obtaining a crucible full of the purest and most homogeneous glass, and then allowing it to cool slowly and to solidify in situ. From the resulting mass of glass the best pieces are picked and moulded into the desired shape for optical use. It will be seen at once that in this process there is an essential difference from all others that have been described in this book—viz., that the glass is never removed from the melting-pot while molten, and that none of the operations of gathering, pouring, rolling, pressing, or blowing are applied to it. The reason for this apparently irrational mode of procedure lies in the fact that the perfect homogeneity essential for optical purposes can only be attained by laborious means, and can then only be retained if the glass is left to solidify undisturbed; any movement by the introduction of pipes or ladles would result in the contamination of the glass by striÆ and other objectionable defects.

The choice and proportion of raw materials used in the production of any given quality of optical glass is governed by the chemical composition which experiment has shown to be necessary to yield the desired optical properties. The composition of optical glass mixtures cannot therefore be varied to suit the conditions of the furnace or to facilitate ready melting and fining, so that many of the usual resources of the glass-maker cease to be available in the very case where their aid would be most welcome to facilitate the production of technically perfect glass. On the other hand, the manufacturer has a certain amount of choice as to the precise form in which the various chemical ingredients are to be introduced into the mixture, and he makes his choice among oxides, carbonates, nitrates, and hydrates, according to the behaviour that it is desired to impart to the mass during the earlier stages of fusion. The state of purity in which the various substances are commercially obtainable also enters largely into the question, since the greatest possible degree of purity in the raw materials is essential to the production of glass of good colour, or rather freedom from colour.

Since homogeneity is so essential in the finished product, very thorough mixing of the raw materials is necessary in the case of optical glass, and the ingredients are for this purpose generally used in a state of finer division than is necessary with other varieties of glass. As a rule the quantities of mixture of any one kind that are required are not large enough to justify the use of mechanical appliances, and very careful hand-mixing is carried out.

Although it is quite possible to obtain successful meltings from raw materials alone, it is preferable to mix with these a certain proportion of “cullet” or broken glass derived from a previous melting of the same sort. The broken glass used for this purpose is first carefully picked over for the purpose of rejecting pieces that contain visible impurities, although pieces showing striÆ are not usually rejected. The greater part of this cullet is generally mixed as evenly as possible with the raw materials, but a certain proportion is reserved for another purpose, as explained below.

The furnaces used for the production of optical glass vary very much in type in different works. In some the old-fashioned conical coal furnaces are still used, the disadvantages attached to their employment being outweighed—in the opinion of the manufacturers—by their simplicity and ease of regulation. In other works gas-fired regenerative furnaces of the most recent type are installed, and in these also optical glass of the highest quality can be produced. As a rule, however, optical glass furnaces differ from other pot-furnaces found in glass-works in this respect—that the former are usually constructed to receive one pot or crucible only, while in other glass furnaces from four to twelve or even twenty pots are heated at the same time. The reason for this restriction in the capacity of the furnaces lies in the fact that since the mixtures used for optical glass cannot be adjusted to suit the furnace, the latter must be worked as far as possible in such a way as to suit the mixture to be melted in it, and this implies that every pot will require its own adjustment of times and temperatures, and this it would be difficult, if not impossible, to secure if more than one pot were heated in the same furnace. It is further to be remembered that the amount of care and attention required during the melting of a pot of optical glass is out of all proportion to that needed with other varieties, so that little would be gained by having a number of pots in one furnace, since several sets of men would be required to tend them.

In addition to the single-pot melting furnace, a very important part of the equipment of the optical glass works is formed by a number of kilns or ovens which are used for the preliminary heating, and sometimes for the final cooling of the various crucibles or pots. Similar kilns are used in other branches of the industry, but in those cases the pots, once introduced into the furnace, are expected to last for a number of weeks, or even months. In optical glass manufacture, on the other hand, a pot is used once only, so that fresh pots are required for every new melting. The kilns in which these pots are heated up before being placed in the melting furnace are thus in very frequent use. As a rule they are simply fire-brick chambers provided with sufficient grate-room and flue-space to be gradually raised to a red heat in the course of four or five days, while for the purpose of gradual cooling they can be sealed up like the annealing kilns used for polished plate-glass.

The pots or crucibles in which optical glass is melted are usually of the same shape as the covered pots used for flint-glass as illustrated in Fig. 2. The optical glass pots, however, are made considerably thinner in the wall, since they are not required to withstand the prolonged action of molten glass in the same way as pots used for flint-glass manufacture. On the other hand, the fire-clays used for this purpose must be chosen with special care so as to avoid any contamination of the glass by iron or other impurities which might reach the glass from the pot. For the production of certain special glasses, in fact, pots made of special materials are required, since these glasses, when molten, produce a rapid chemical attack upon ordinary fire-clays. A certain amount of the aluminiferous material of the pot is, in fact, always introduced into the glass by the gradual dissolving action of glass on fire-clay which we have already described. The glass contaminated with these aluminiferous substances is generally more viscous than the rest of the contents of the pot, and therefore ordinarily remains more or less adherent to the walls of the crucible, but the inevitable disturbances which accompany the processes of melting and fining lead to the dissemination of some of this viscous glass through the entire pot in the form of veins or striÆ, which are only removed during the stirring process. On the other hand, more of this viscous glass is constantly being formed so long as the glass remains molten, and if disturbances are not sufficiently avoided during the later stages of the process fresh veins may easily be formed.

The actual operations of producing a melting of optical glass begin by the gradual heating-up of the pot in the kiln just described. When the pot has reached a full red heat the doors of the kiln are opened and the pot drawn out by means of a long heavy iron fork running on wheels; this implement is run into the mouth of the kiln and the tines of the fork are pushed under the pot, and the latter is then readily lifted up and withdrawn from the kiln. Meanwhile the temperature of the furnace has been regulated in such a manner as to be approximately equal to that attained by the heating kiln, so that the pot, when transferred as rapidly as possible from the kiln to the furnace, is not subjected to any very sudden heating; were it attempted to place the new pot in a furnace at full melting heat the fire-clay would shrink rapidly and the entire vessel would fall to pieces. Even under the best conditions it is not possible to avoid the occasional failure of a pot by cracking either at this or a slightly later stage of the process. The latter occurrence is apt to be particularly disastrous, as the pot may then be full of molten glass, which runs out and is lost.

As soon as the empty pot has been put into place, the melting furnace is carefully sealed up by means of temporary work built of large fire-bricks, the whole being so arranged that the mouth of the hood of the pot is left accessible by means of an aperture in the temporary furnace wall. This aperture can be closed by one or more slabs of fire-clay, and when these are removed an opening is left by which the raw materials are introduced, and through which the other manipulations are carried out.

When this stage of the process is reached, the wagons containing the mixed raw materials are usually wheeled into place in front of the furnace, but the introduction of the materials themselves into the pot is not begun until several hours later, when the furnace has been vigorously heated and an approach to the melting heat has been attained.

When the furnace and pot have attained the necessary temperature, but before the raw materials are introduced, a small quantity of the cullet, which has been reserved for this purpose, is thrown into the pot and allowed time to melt, and then only is the first charge of mixture put into the pot. The object of this proceeding is to coat the bottom and part of the walls of the pot with a layer of molten glass which serves to protect it from the chemical and physical attack of the raw materials during the violent action which takes place when they are first exposed to the furnace heat.

The gradual filling of the pot with molten glass is now carried out by the introduction of successive charges of raw material; as the mixture not only occupies more space than the glass it forms, but also froths up a good deal during melting, the quantities introduced each time must be carefully adjusted so as to avoid an overflow of half-melted glass through the mouth of the pot. As the pot is more and more nearly filled, the space left for the raw materials is proportionately diminished, and the later charges are therefore much smaller than the first few.

When, finally, sufficient material has been introduced to fill the pot completely, the next stage of the process commences. When the last charge of raw materials has melted, the glass in the pot is left in the state of a more or less viscous liquid full of bubbles of all sizes; it is essential that these bubbles should escape and leave the glass pure and “fine,” and this result can only be achieved by raising the temperature of the furnace and allowing the glass to become more fluid, while the rise of temperature also causes the bubbles to expand owing to the expansion of the gas contained in them. In both ways, rise of temperature facilitates the escape of the bubbles, and the furnace is therefore heated to the full, and this extreme heat is maintained until the glass is free from bubbles. In the case of the more fusible glasses the temperature required for this purpose is not excessively high, and, indeed, in the case of these glasses care is taken to avoid too high a temperature, as it entails other disadvantages. In the case of the harder crown glasses, however, the difficulty lies in producing an adequately high temperature without at the same time endangering the life of furnace and crucible. The difficulty of freeing the molten glass from bubbles constitutes one of the causes that limit the range of our optical glasses in one direction—still harder glasses could be melted, but it would not be feasible to maintain a temperature high enough to render them fluid enough to “fine.”

In the case of other kinds of glass, again, it becomes impossible to entirely remove the bubbles from the molten mass even when very hot and very fluid. The exact cause is not known, but in some kinds of glass the bubbles formed are so minute that even when the glass is perfectly mobile the bubbles show no tendency to escape, while in other kinds of glass there appears to be a steady evolution of minute bubbles as soon as the temperature is raised with a view to removing those already in the glass. As this property attaches to some of the most valuable of the newer varieties of optical glass, opticians and the public have learnt to put up with the presence of minute bubbles in the lenses and prisms made of these glasses. These bubbles are, however, very minute and do not interfere with the optical performance of the lenses, &c., except to the extent of arresting and scattering the very small proportion of light that falls upon them; their presence is therefore to be regarded as a small but unavoidable drawback to the use of glasses which offer advantages that completely outweigh this defect.

Returning to the melting process, we find that the extreme heating required for the purpose of “fining” the glass is continued for a considerable period of time, as long as thirty hours in some cases, the glass being examined from time to time to test its condition as regards freedom from bubbles. This is done by taking a small sample of glass out of the pot and examining it to see if it still contains bubbles. In some works this test is made by taking up a very small gathering of glass on the end of a small pipe and blowing it into a spherical flask; on looking at such a flask in a suitable light the presence of even minute bubbles is readily detected. In other works a simpler process is adopted, a small quantity of glass being ladled out of the pot on the surface of a flat iron rod. It is allowed to cool on the rod, and when pushed off forms a small bar of glass some eight or ten inches long and about an inch wide; in this also the presence of bubbles is easily detected. These test pieces are known among glass-makers as “proofs.”

When proofs, taken as just described, have shown that the glass is free from bubbles, the extreme heat of the furnace is allowed to abate, and the fire-clay slabs in front of the mouth of the pot are removed. The next step is that of skimming the surface of the glass. Since most of the materials liable to contaminate the contents of a pot are specifically lighter than the molten glass, they will be found floating on the surface, and the surface glass is therefore removed with a view to ridding the glass of anything that may have been accidentally introduced and that has not melted and become incorporated with the molten mass.

The next steps in the process are those of stirring the molten glass with a view to rendering it homogeneous and free from striÆ. The stirrer used for this purpose is usually a cylinder of fire-clay, previously burnt and heated. This is provided with a deep square hole in one end, and it is held at first by means of a small iron bar passed into this hole. By this means the red-hot cylinder of fire-clay is introduced into the open mouth of the pot, and when it has attained approximately the temperature of the molten glass it is dipped into the glass itself, in which it ultimately floats. When stirring is to begin, the square, down-turned end of a long iron bar is introduced into the corresponding square hole in the upper end of the stirrer, and by this means the fire-clay cylinder is held in a vertical position in the glass and given the steady rotatory movement which constitutes the stirring process. For this purpose the long iron bar just mentioned is made to pass over a swivel-wheel, while a workman moves it steadily by the aid of a large wooden handle. This operation is always laborious and trying; the workman is necessarily exposed to the intense heat radiated from the open mouth of the crucible, so that men have to relieve each other at frequent intervals.

During the earlier stages of the stirring process the glass is very hot and mobile, but the stirring is continued, with short intervals, until the glass is so cold and stiff that the stirrer can scarcely be moved in it at all, so that the work of moving the stirrer becomes heavy towards the end of the operation. The actual amount of stirring required varies according to the nature of the glass, and the size of the pot or crucible in question. Some meltings are found to be satisfactory after as little as four hours’ stirring, while for others as much as 20 hours are required.

When the glass has stiffened to such an extent that it is no longer possible to continue the stirring, preparations are made for the final cooling-down of the pot of glass. The fire-clay stirrer is sometimes withdrawn from the glass, but this is laborious, and entails dragging a considerable quantity of glass out of the pot with the clay cylinder; more usually, therefore, the stirrer is simply left embedded in the glass.

The next object to be accomplished is that of cooling the glass as rapidly as safety will permit until it has become definitely “set”—the purpose being to prevent the recrudescence of striÆ as a result of convection currents or other causes which might disturb the homogeneity of the glass. This rapid cooling is obtained in various ways; in one mode of procedure the furnace is so arranged that by opening a number of apertures provided for the purpose cold air is drawn in and the pot and its contents chilled thereby without being moved. This method has the advantage that the pot containing the viscous glass is never moved or disturbed in any way, but on the other hand the cooling which can be effected within the furnace itself is never very rapid, and the furnace as well as the pot is chilled. Further when the glass has been chilled down to a certain point this rapid rate of cooling must be arrested, as otherwise the whole contents of the pot would crack and splinter into minute fragments. Where the pot has been left in the furnace this can only be done by sealing up the whole furnace with temporary brickwork and lutings of fire-clay, leaving it to act as an annealing kiln until the glass has cooled down approximately to the ordinary temperature, a process that occupies a period of from one to two weeks according to the size of the melting. Such enforced idleness of a melting furnace is of course very undesirable from an economical point of view, and it is generally avoided by adopting the alternative method of drawing the pot bodily out of the furnace as soon as the stirring operation is ended. For this purpose the temporary brickwork forming the front of the furnace is broken down, and with the aid of a long crow-bar the bottom of the pot is levered up from the bed or siege of the furnace to which it adheres strongly, being bound down by the sticky viscous mass of molten glass and half-molten fire-clay which always accumulates on the bed of the furnace. The pot being temporarily held up by the insertion of a piece of fire-brick, the tines of a long and heavy iron fork running on a massive iron truck are introduced beneath the pot; an iron band provided with long handles is then passed around the pot, and the latter is then drawn forward by the aid of suitable pulley blocks. The tines of the fork are then raised, and the pot is wheeled out of the furnace and deposited upon a suitable support. Here it is allowed to cool to the requisite extent, when it is again picked up on the tines of the fork and deposited in an annealing kiln which has been previously warmed to a suitable temperature. It will be seen that this handling of a heavy mass of intensely hot material involves much labour, while there is also a risk of losing the glass if the pot should break before the glass has set sufficiently. Every care is taken to prevent such an accident, the pot being wrapped round with chains or otherwise supported in such a way that a small crack could not readily develop into a large gap.

When such a melting of glass has cooled sufficiently, either in the furnace or in the annealing kiln, to be safely handled, the whole pot is drawn out, and the fire-clay shell, which is generally found cracked into many pieces, is broken away by the aid of a hammer. Under favourable circumstances the whole of the glass may have cooled intact as one solid lump sometimes weighing over half a ton. Unless special care is taken, however, it is more usual to find the glass more or less fissured, a number of large lumps being accompanied by a great mass of small fragments. These are now picked over, and all those which are free from visible imperfections or which can be readily detached from such imperfections by the aid of a chipping hammer are put upon one side for further treatment.

The next step of this treatment consists in moulding the rough broken lump into the shape of plates, blocks, or discs according to the purpose for which the glass may be required by the optician. The plant used for the moulding process varies widely, but in all cases the operation consists in gradually heating the glass in a suitable kiln until it is soft enough to adapt itself to the shape of the mould provided for the purpose. In some cases these moulds are made of fire-clay, and the glass is simply allowed to settle into them by its own weight; in other cases iron moulds are used, and the glass is worked into them by the aid of gentle pressure from wood or metal moulding tools. In yet other cases, particularly where the glass is required in the form of small thin discs or where it is to be formed into the approximate shape of concave or convex lenses, the aid of a press is sometimes invoked.

In all cases the moulding process is followed by the final annealing, which consists in cooling the glass very gradually from the red heat at which it has been moulded, down to the ordinary temperature. The length of time occupied by such cooling depends very much upon the size of the object and also upon the degree of refinement to which it is necessary to carry the removal of small internal strains in the glass. For many purposes it is sufficient to allow it to cool down naturally in a large kiln in the course of six or eight days. For special purposes, however, where perfect freedom from double refraction is demanded, much greater refinements are required, and special annealing kilns, whose temperature can be accurately regulated and maintained, are employed. In these the annealing operation can be carried out so gradually that a rate of cooling in which a fall of 1° C. occupies several hours can be maintained, so that very perfectly annealed glass can be produced even in discs or blocks of large size.

When removed from the annealing kiln the plates or discs of optical glass are taken to a grinding or polishing workshop, where certain of their faces or edges are ground and polished in such a way as to permit of the examination of the glass for bubbles, striÆ and other defects in the manner indicated in the previous chapter. As the amount of sorting that can be done while the glass is still in rough fragments is necessarily very limited, it follows that a considerable proportion of the glass which has been moulded and annealed must be rejected as useless when thus finally examined. A yield of perfect optical glass, amounting to 10 or at most 20 per cent. of the total contents of each pot, is therefore all that can be expected, and smaller yields are by no means infrequent—a consideration that will serve to explain the relatively high price of optical as compared with other varieties of glass.

A consideration of the various factors that are involved in the production of a piece of perfect optical glass will make it apparent that the cost and difficulty of its production increases rapidly with the weight of the piece to be produced, so that it is not surprising to find that the price of very large discs of perfect optical glass such as those required for large astronomical telescopes, reaches figures which become prohibitive when very large sizes are considered. Thus, while it is quite possible to obtain say 100 pounds of good glass from a single melting if the glass is to be used in the form of pieces not weighing more than five or six pounds each, it is only rarely that a single block of perfect glass can be found weighing 100 pounds. In the former case the best pieces can be picked, the worst defects can be eliminated by chipping the rough fragments, and at a later stage other defective pieces can be cut off or ground away; not so where a large single block is required. A single fine vein, perhaps too small to be visible to the unaided eye, may be found to run through a whole block in such a way that it cannot be removed without breaking or cutting up the whole piece, and it will be seen that the frequency with which this is liable to occur increases with the volume of the piece required. The difficulties of re-heating and moulding are also increased enormously with the size of the individual pieces of glass that have to be dealt with, and where very large pieces have to be heated and cooled accidental breakage becomes a serious risk. In view of these difficulties it is not surprising to find that the dimensions of our astronomical refractors appear to have approached their limit, but rather are we led to admiration of the skill and enterprise that has pushed this limit so far as to produce discs of optical glass measuring as much as one metre in diameter.