Parys Underground Group / Grwp Tanddaearol Parys

Smelting of copper

The knowledge of the process of smelting copper had been lost in Elizabethan England and it was necessary to bring a number of German experts into the country before copper smelting could begin.

These German experts understood that it was necessary to melt a variety of ores together to enable the best copper to be made. They also had a good idea of the effect of impurities in the copper ore had on the smelting process.

Various impurities were described at Keswick in 1581. The first of these was sulphur:- " a mynerall substance which verie quickly taketh fire,and wilbe consumed in smoke by blast ; it leaves the copper black and brittle,so that it wilbe broken w'th the hammer,in a manner like glass. Arsenic was also " consumed w'th fire into smoke w'ch is vere daungerous ayer or savor,and by his force maketh the copper white and brether than the sulphur does" Antinomy was "a great let and hinder to the copper smeltinge and is consumed by heat" Vitriall was removed by "stampering the copper ure into powder and by rostinge the same powder before it be smolten,and then letting water passe through the same roasted powder,the water doth not onlle carry the vitriall from the powder or ure,but carrieth w'th it the burnt powder or sinder of the sulphur,arsenicque and antimony,whereby it so clenseth the ure that when it cometh to the smeltinge the copper cometh forth easelie,w'thout such quantitie of slagges or drosse as otherwise would be,if the ure was not rosted and the vitriall in the manner taken from it;Thus is the vitriall of an enimye made a friende"

Other substances such as iron and black stone (Gangue) gave rise to much difficulty in the smelting,but the remaining materials "calcator" (Oxide of iron) alum and spar are mentioned as being "least hurtful to ye copper. It is also recorded that "8 or 9 seks of chare coles and thre horslod of sea coles" were required to smelt 24 hundredweights of ore.

Copper smelting furnace from Agricola De Re Mettalica 1556

In 1632 Edward Jorden obtained a patent ( Patent No 61 (1636)) for smelting minerals with pit coal,peat and turf. The type of furnace used was improved by Captain Thornese Francke in 1635.  This was followed in 1636 by Sir Phillibert Vernatts patent (Patent No91 1636) for " makeinge,melting or smelting,casting,founding,fineing,nealinge,beating and workeing of iron,steele,brasse,copper,tynne,leade ,a nd all other kind or oaremetals...and for casting of ordinance and doeing all the other workes...with fire os sea cole,pitt cole or stone cole ... without mixeing the same with charcole or charking the said sea cole.

A few months later George Danby patented a method of melting and casting copper into ingots.

In 1691 Sir Clemant Clerke ,JohnCoster, John Duckett and Gabriel Wayne discovered" a new method for the more easy and effectual melting down,refining and purifying of copper ore " This lead directly to the formation of the English Copper Company and the building of a smelting works in London and at Redbrook in The Forest of Dean. The later works were supplying almost 1000 tpa of copper by the end of the century.

The raw copper ore was obtained from mines at Alderly Edge in Cheshire ,Redbrook in Monmouthshire and also Devon,Cornwall and Cumberland. The work on the process of improving the reverberatory furnace to burn sea coals had been ongoing since 1688. In that year George Moor was acting as a merchant in London and was selling the "Rose copper" at ten pence a pound.

In 1722 George Moore himself patented a " new method of refining copper by air and blast with proper furnace and with sea coals,whereby the copper is purified and refined at one operation,with less charge of fire or loss of copper or the materials used as fluxes,than by any way hitherto known and whereby the worst copper is made fit for the best uses. ( Patent No 443 1722)

In 1728 Lund and Hawksbee were granted a patent ,no 495, which described a way in which the sulphur in the copper ores could be destroyed and the quality of brass improved. The ore was first calcined in a closed furnace or oven without the admission of air or naked fire until the volatile sulphur was discharged. The calcined ores were taken out of the furnace while they were hot and immersed in water. After drying the ore was again returned to the furnace and melted and then poured into a mould and so the fine parts would be separated from the course. The refining ,washing and melting was continued a number of times until the required quality of copper was obtained.

In 1767 William Champion showed how the copper could be made more suitable for making brass by refining the ore with wrought iron to remove the arsenic.

In 1778 Thomas Williams obtained another patent ,No 1191 ,with further improvements to the Champion process.

When smelting began at Amlwch is not certain.  However the system has been described by a number of eye witness account.

It is thought that the initial smelters were built along side the Afon Amlwch in an area now know as Craig y don.  The original building was 112 feet long by 50 feet wide. The river was diverted to make a small reservoir to be used in the works.  In 1797 Aikin describes the smelting works.

  Aikin 1797

The two smelting houses, of which one belongs to each company, contain 31 reverbatory furnaces, the chimneys of which are 41 feet high. They are charged every 5 hours with 12 cwt of ore which yields ½ cwt of rough  copper, containing 50% of the pure metal. The price of rough copper is £2-10-0 per cwt. The coals are procured from Liverpool and Swansea.

The location of the cooper smelters at Amlwch port were dependent of a steady stream of coal by sea which was the subject of a excise duty. Thomas Williams fought long and hard to get the tax reduced. However he also built smelters closer to the coalfields at Ravenhead in Liverpool. 

These were described by Lentin in 1800 who also said that only 20 smelters remained in the Amlwch area. 

…The ores that have been roasted in Anglesey are sent to Liverpool in 80 - 300 ton ships, where they are unloaded into barges which deliver them right to the door of the foundry, via the Liverpool-Wigan-Leeds canal. All these ships are the property of the shareholders of the company who raised the cost for them through shares, and earn them a considerable amount of profit because, in addition to the ore freight, they also conduct a very profitable shipping business.

The first operation, which is carried out with the ore in the smelting works consists of further divesting the ores of the Sulphur, which had not been removed during the first roasting. However, before I tell you about this process, allow me to say a few words about the structure of the English Calcination and smelting furnaces so that I can continue my explanation with fewer interruptions.

You know that in England one uses the wind furnace, which Schlüter called the cupel furnace, for all smelting, except that of iron ore.  In his very thorough treatise on metallurgical processes Schlüter has, such an excellent illustration of this furnace that I do not have to say much, because this valuable book is found in the possession of all metallurgists,   yourself included.

The interior of all furnaces of this type are made of an artificial fire-proof stone because they are exposed to a continuous and intense heat. These stones are called fire bricks and are made in a very spacious factory in Stourbridge where a particularly suitable clay is found.

The hearth of a smelting furnace is prepared from completely dry pure gravel, with a top layer a few inches high, which consisting of sand used by mirror manufacturers to polish glass plates. This has some pieces of glass in it which, when exposed to a sufficiently strong heat, begin to soften and so cement the sand granules to each other to a certain extent, and as a result hinder the metals from passing through. Under normal conditions such a hearth can be used for a whole year without requiring reconditioning.

The hearth in calcining furnaces, on the other hand, is completely horizontal and lined with fire-proof stones which are also made in Stourbridge.

The chimney or flue of such a furnace must be 55 - 60 feet high, rise vertically and be one square foot internally. This height is in no way superfluous, rather the intensity of the heat depends on it. There is no longer any doubt that the heat does not originate in the fuel but rather comes from the oxygen air that is a component of our atmosphere.  In this air the fuel is so bound through a substance which the analytical chemist calls oxygen, after Lavoisier, that it cannot express its propensity to heat.  This oxygen can, meanwhile, be removed at suitable temperatures by several bodies, in particularly carbon which is copiously contained in all combustible substances, and what results then, depending on how fast the separation occurs, is either gentle warmth or the most intense heat. Because all metallurgical smelting processes and, in particular, the copper smelting process, demand very high temperatures, one has to endeavour to construct a smelting furnace which allows for the decomposition of the largest possible volume of oxygen air in the shortest possible time.

One achieves this, as is well known, in two different ways.  First through blast furnaces whereby bellows constantly press fresh air into the furnace where its oxygen-air is broken down into its components by the coal present, thus producing the required degree of heat.  The second method uses the wind, or air furnace in which the required air flows through freely. This constant stream of air, without which, naturally, no fire could exist, is produced because the air in the interior of the furnace and in the flue is, through the fire, partly decomposed and partly greatly expanded; now air possesses a tendency to set itself in equilibrium everywhere and to force itself in again wherever this equilibrium is disturbed;  this penetrating air is continually expanded by the heat and in this way a draught develops.

The larger the volume of expanded air, the stronger the draught in the furnace will be, and the height of the chimney in such an air furnace is also important because it not only serves to take away the smoke, as very many people incorrectly believe, but helps to enlarge the thin air column that the influx of atmospheric air is supposed to carry in.

Because one cannot use the same degree of heat for all operations, one has to be able to increase and decrease its intensity, which here, where anthracite is the fuel, happens through the anthracite itself. If the heat is to be gentle, then one throws small, dampened pieces of anthracite on the fire, which soon start, because of the heat, to sweat out their bitumen, which then flows in a mass and forms a cover over the burning coals, which inhibits the draught and thereby decreases the heat.  If this covering is broken again, then the air can stream in again and the  heat will increase.

English smelters call such a mild heat a roasting heat and the other a melting heat - nomenclature which will often occur in my account, with which I will now continue.

I have already said above that calcining furnaces are built following the same principle as the smelting furnace and that they only differ themselves in that the hearth is horizontally bricked up with fire-proof artificial stones. Yet I have to note that instead of the openings which Schlüter presented in his drawing as the tap hearth and as the discharge opening, through which the slag is removed, there are two large openings affixed which can be closed with iron trapdoors and serve to introduce the ore, and following that to stir it.

In such a furnace two to three tons of ore 40 - 60 hundredweight are shovelled through both of  the openings just described, evenly spread about and initially subjected to a roasting heat, which, however, towards the end of the operation is greatly increased.  During this time the ore is frequently stirred with iron rods which are affixed to rollers on the front of the furnace in order to control them better.

The time which is required for sufficient Calcination of such a quantity of ore cannot be set because it depends on the degree of roasting which the ore has had; usually, though, 12 hours is sufficient. Actually it often does not require even that much time to separate off all the Sulphur present, which can be separated by this method.

At this point the calcined ore is brought to the foundry and smelted with the addition of flux in the following way. To 12 hundredweight of ore, which constitutes a load for this kind of furnace, 42 hundredweight of copper slag and two hundredweight of carbonized anthracite or coke are added.  The anthracite is carried in first, then the ore and over these comes the roughly broken up slag.  Everything is thrown into the outlet using a shovel; For the opening, found in Schlüters drawings through which the ore should be placed into furnace with a funnel, is not found on a copper smelting furnace, and the outlet is stopped up with a door of clay and cemented shut.  A roasting heat is applied for about one to one and a half hours so that, if there were still Sulphur present in the ore, this could completely volatilise.  The temperature of the fire is raised to the highest degree and maintained at this level, uninterrupted, for two to three and a half hours, which is why the anthracite is always added in small portions, so that no noticeable decrease in temperature takes place through adding too great a quantity. Then the door of the outlet is taken away to check the state of the smelting, which should be so far along that the mass is nearly fluid but the slag is still semi-solid.

The smelter then throws a few shovels full of anthracite over the melting mass, closes the outlet again, increases the heat and keeps it at the highest temperature until the smelting is complete. This takes about six hours from the beginning of the operation.

During the smelting the tap hearth which consists of a frame of iron plates six feet long, three feet wide and 18 feet high, is put in order.  The forms into which the whole molten mass will be poured are prepared in these using damp sand. The shape of the forms is arbitrary; however, one usually gives the shape of an inverted pyramid which has an upper surface eight inches wide. There are as many of these shapes cut into the sand as space allows, with a space of four inches between each one, in which a communicating gutter is cut so that the excess molten material from the first can flow into the second and that from the second into the third, etc. The form which is connected directly with the tap is cut out a little wider and deeper than the others because it mainly takes the metallic part, while the others, in contrast, are filled only with slag.

When the ore is sufficiently melted, then the furnace is opened, which is achieved by cleaning away the clay which has sealed the tap hole or mouth of the furnace with a scraper after which the molten mass runs into the forms. When all the molten material has run out of the furnace, the furnace is charged anew by throwing little, moistened pieces of anthracite into the hearth so that the heat drops, whereupon one puts in the hard coal and other ingredients. And so the smelting operation continues, uninterrupted, and one puts 40 hundredweight  through the furnace in 24 hours.

When the mass in the tap hearth is suitably cooled down, the pieces next to the tap hole are removed and the slag is separated out from the metallic part. This is achieved very easily with a few blows of the hammer. The slag is thereafter brought to a special place where one breaks it into bits in order to get out the stone granules which might still be there, although this does not often occur because the smelting proceeds very completely, and the mass remained in the fluid state in the furnace long enough so that the metallic portion could separate out from the slag. These metallic parts have a blackish green appearance, are granular and compact like basalt when broken and sound like it if one hits them with a hammer.

The reguline or metallic portion, which is in the same state as that which the people in our copper works call Sporstein, is broken into pieces the size of a fist, and 60 hundredweight of it is brought to a special furnace, which has the same construction as the smelting furnace, with the exception that, instead of a tap hearth, a water container is brought near the furnace and is set up so that fresh water can flow continuously.

In this furnace the stone is roasted for 12 - 16 hours; the heat is then gradually increased in intensity so that the mass melts, and in this state about a hundredweight of anthracite is thrown onto the melting mass. When everything is quite fluid, the slag which has been produced on the surface is removed.  A lead gutter lined with clay reaches from the tap hole to the water container. The fluid stone flows along the gutter into the water where it is granulated into pieces the size of beans through the constant movement of the water.

The purpose of this operation is first to fluidise a part of the Sulphur through the long roasting and then, through the subsequent granulation, to produce a greater surface area in order to be able to dislodge the Sulphur more effectively.

After the granules have dried somewhat, they are put into a calcining furnace and calcined with very active stirring for 26 hours at a temperature as intense as one can use without melting the mass. Then the granules are melted in a smelting furnace, which does not have a tap hole and whose hearth is so constructed that the greatest depth is in front of the outlet. 

When everything is sufficiently melted, the slag is removed, and the molten mass is emptied into iron boxes washed out with clay using large ladles.  The boxes are 16 inches long, 11 inches wide and 11 inches deep, and one places a suitable number of these in a half-circle in front of the furnace.  Two workers, who have pulled moistened sleeves of rough linen over their arms to keep off the heat, alternate in taking a shovelful of the material from the furnace and fill the boxes one after the other, in that each one pours a ladleful into a box.   While the last box is being filled, the metal in the first has solidified to such an extent that a new layer can be poured on to it without binding with the first.  So the boxes are filled, one after the other, producing a number of thin cakes which are approximately 1/4 inch thick and which can be separated when they are properly cooled.

The purpose of this procedure is to give the mass, which is now in the state of a very impure black copper and which one can no longer break into bits, a large surface area so that the pieces of Sulphur still present can be more easily volatilised.

For this purpose the cakes of black copper are put in another, but similarly constructed furnace, where they will be roasted for 16 hours with the heat increasing in measured increments, so that the metal melts very slowly.  After everything is fluid, one allows the metal to simmer for a few hours so that the calcined parts can scorify, and one removes the slag through the outlet.  Then one lets the fire die down a bit and pours the fluid black copper into iron boxes as before.

The metal is now in a state where it can be purified. This, however, does not take place in the same way as in our own copper works but rather in a way which distinguishes it, in that the copper at these works is made completely pure immediately so that it can be worked by coppersmiths without further refining.

To achieve this one places the cakes in the refining furnace, which is just like the previous furnace, and roasts it for 16 hours, gradually increasing the temperature so that the copper melts slowly.  When this has happened, 25 to 30 pounds of lead is put onto the three tons of copper which one will refine all at the same time. Then one maintains a strong heat for several hours so that a proper slag can form and the impurities still found in the copper can dissolve out.  The slag is removed from time to time so that the calcination of the metal pieces being scorified can proceed even better with the  introduction of fresh air.

When the scorification is coming to an end, enough coal dust is thrown on to sufficiently coat the surface of the metal. Then the refiner shoves a tree in the furnace and  presses it into the metal as strongly as he can. Because one uses elms or young oaks from four to six inches in diameter which have been freshly cut or left to lie in water to keep them damp, a violent bubbling occurs in the copper when one pushes the tree into it. This is allowed to continue for several minutes. Then one throws coal dust over the copper again and the bubbling recurs. This process continues until the copper has reached the requisite degree of purity.  One ascertains this by removing a little copper with a small test spoon. As soon as the copper solidifies, it is quenched in water and breaks into bits.  If the colour is a bright brick red, and the break shows that it is fine-fibred and as glassy as satin, then it is of the required purity.  If, on the other hand, it is blistered inside and dark red, then one has to carry out the previous procedure, which one calls phlogistics for a longer period of time. One must even add some lead, if the metal is still not ready, whereby, however, the phlogistics must always be repeated.

When the copper has reached the desired purity, it is either poured out as cakes in the previously described manner, granulated or poured out in small bars, which are formed like bars of sealing wax and are usually called Japanese copper in Germany but should actually be called lacquered copper because the English word japanned copper signifies this. Also it is generally known in England that all the copper produced by this firm is produced in the country and is used in business with the East Indies.

The cakes, most of which are treated in the rolling mills of the company, are usually one and an half inches thick.  In order to maintain this thickness, it is necessary that both of the workers who do the pouring shovel as quickly and at as regular intervals as possible so that when the first one has poured out his ladleful, the other one can follow on immediately so that the metal does not solidify before the next ladleful arrives.

The boxes are washed out with clay beforehand and the bottom cakes are only poured thinly because they have to be melted again.  When the pouring is finished, the boxes are cooled in a tank with water.  Then the cakes are taken out, trimmed at the edges, stamped and numbered and sent off from the works.

Granulation occurs in the following way. A cylindrical container of copper which can be brought up again afterwards with the help of a crane winch is let down into a round 12 foot deep tank, through which water continuously flows,  A copperplate dome is place over the tank, with an opening on one side for the ladle carrying the metal. This hole must, for this occasion, be much smaller than the hole used in pouring of the plates because the metal can only be poured into the water in small portions, on account of the violent explosion, which would occur if too much metal hits the water at once, and because too much copper ladled out at once would also solidify in the ladle. For this reason this process is carried out by only one worker. He scoops up some copper, puts the spoon through the opening of the dome over the tank and gradually shakes the copper into the water where it sinks to the bottom producing crackling sounds and frequently, blue flames.  The copper falls two to three feet through a column of air before it reaches the water and by doing this takes on a form which has a distant resemblance to the texture of a feather.  One therefore calls these granules feather-shot to distinguish them from the rounder granules which one gets when the copper is poured directly into the water.  Because the granulation takes a lot of time, one has to ensure that the metal in the furnace does not get too cold; to protect it against calcining, the surface is dusted with coal dust.

Granulation was used everywhere in England in the preparation of brass, for which it is also undoubtedly more suitable than refined copper because the granules consists of very thin plates which can mix with the calamine better and offer a larger surface area for the developing zinc vapour to penetrate. However, more of this when I discuss brass production.

The so-called lacquered copper is poured in the form of ingots. Several such forms are fastened to an iron wheel which is positioned over a tank filled with water, which is likewise positioned very close to the front of the furnace.  When the metal reaches the necessary refined state, a young lad, using a paint brush, paints the form with oil, and the pouring master, using a spoon which holds just about what the form will hold, pours the metal in the forms.

In the meantime another worker holds ready a container with water which he pours on the copper in the form; As soon as the copper solidifies, he turns the wheel so that the bar cools itself completely in the water in the tank.

The copper is poured into the form next to the first form following the same procedure; however, before one paints the form with oil it is first carefully dried out with a cloth. Otherwise the water mixing with the molten copper could cause a dangerous explosion for one of the workers. One uses the forms in rotation so that there is enough time for them to cool again before introducing molten copper; otherwise the copper would stick to them.

This is the simple process which produces a quality of copper which one values so much abroad, which, however, the English smelters who produce it regard as less pure than cake copper. The oil with which the forms are painted before each casting serves to hinder the effect of oxygen from the air as well as the cooling water and keeps the outer surface completely metallic. The outer surface also becomes shiny through the sudden pulling together of the particles in the same way that any metal takes on a higher shine when its surface particles are brought closer together through pressure, under the coin press, for example.

The working up of the precipitated copper is less difficult than the procedure for the ore.  At the first smelting one places eight hundredweight of precipitate with four hundredweight of slag and half a hundredweight of anthracite in the furnace, takes away the slag when everything is well melted and pours the metallic part, which is very impure black copper, in iron boxes to make cakes. These undergo a 24 hour roasting and are then slowly melted and poured out.  The same operation is repeated again, after which the metal is sent for refining.

These then are the English smelting and refining processes, which I cannot leave without first explaining a few operations which are not common for us and without making a brief comparison between these and the normal smelting methods used in Germany....


  The  decline in the output from the  mines after 1800 lead to the reduction of the need to smelt ore both at Amlwch and at Liverpool. By 1814 the Liverpool smelter had closed down although around 8 survived and were operational at Amlwch being run by the Vivian Family from Swansea.  As the deep mining techniques of the Cornish miners like Treweek began to increase the output from the mine further smelters were brought on line until by 1818 the smelter was in full operation again.

  In 1819 Faraday visited the smelters.  These consisted of row of reverbatory furnaces exactly similar to thosed used at Swansea works in which the metal is roasted over and over again. The ore does not come in the crude state to them but is roasted up the mountain in kilns by which a great part of the sulphur is separated and then comes to these works.  The metal is not tapped into water here as in Swansea but runs into pigs. It is however granulated after the last step of the refinery process for the use of the makers of brass. There are two ranges of refineries in this place. Between the town and the port. One belongs to Parys and the other the Mona mine.

In 1847 O.Jones of Beaumaris also described the smelting process.

In conjuntion with the Mona mines , smelting of the ore is carried out on a very extensive scale in the Smelting works in Amlwch town.  Twenty furnaces are in full operation built of the Beecher’s  reverbatory  furnace design. These Furnaces are so contrived that the ore is melted not through immeadiate contact with the fuel,but by the reverberations of the flame upon it.  Each furnace is charged with 14 cwt of ore which smelts four four hours and yeields about 40% pure copper.

These furnaces are divided into 6 roasters , 6 ore furnaces , 3 calciners , 3 precipitates and 2 refiners.  The process are conducted in the following manner.

1    The ores are calcined.

2    The calcined ore is melted.

3    The metallic mixture from process 2 is calcined.

4    The calcined course metal is melted

5    The material from process 4 is calcined.

6    The metal calcined in process 5 is melted.

7    The copper from process 5 is roasted.

8    The course or blister copper is refined.

The charging of calcining in the first process is 3 to 3.5 tones of ore, and the cacining lasts 12 hours.  In the second process the melted matter is let out at a hole opened in the side of the furnace into adjoining sand pits,where it become granulated and cools in the form of coarse pigs.

This granulated metal is subjected to calcinations and fusions alternately, as above until it comes to the 7th process of roasting. At this stage the material contains about 85% pure metal. In this sate the bars aor pigs are put into the refining furnace and gradually melted.  The surface of the metal is covered with charcoal and a birchwood pole is then held in the liquid metal, which causes considerable ebulution , owing to the evolution of gaseous mater, and this operation of poleing is continued until the refiner ascertains ,by various trails,that the copper is in the proper state of purity and malleability.

As the produce of our mines requires fluxes for melting,ore from all parts of the world are extensivly brought to assist the fusion of our native production. At the present time the smelters at Amlwch consume 30,000 tons of coal annually.


Copper sulfide ores have a complex chemistry, and the copper smelted from them tends to contain a variety of impurities, some of which "improve" their properties in terms of tool, weapon, or even ornamental function. Technology evolved from  as miners, smelters, and smiths learned to use these new and more complex materials, especially to form deliberate alloys. 

As the kiln heats up, the sulphides begin to break down to release water vapor and sulphur dioxide, and in doing so convert to a copper oxide,

                                 CuS(OH)2 = CuO + SO2 + H2O

Copper oxide is red to black, rather than the blue and green of the original powders, and the completeness of the change would vary, depending on the temperature of the kiln and the length of the firing. If copper oxide is fired to high temperature in an oxygen-starved kiln, large quantities of hot carbon monoxide are produced in the kiln as charcoal burns to CO rather than CO2. Carbon monoxide attacks oxides, ripping off an oxygen atom to form CO2.At about 1100° C, copper oxide forms copper metal:

                                       CuO + CO = Cu + CO2

Once malachite and azurite were identified as significant sources of the well-known (and valuable) metal copper, potters may have deliberately built kilns dedicated to smelting copper ores, and it would not be long before the basic design of the kiln was modified to make an effective smelting furnace, operated by a new kind of craftsman who was a specialized smelter.  The smelter would pack charcoal in intimate contact with crushed ore, to ensure more complete chemical reactions in more reliably oxygen-starved conditions immediately round the ore. The working temperature was raised by blowing air into the furnace through specially designed pipes called tuyères, with or without bellows. Once the copper metal had melted, it would flow downwards to be collected in a ceramic container or crucible. Such furnaces could reliably reach temperatures that would smelt high-quality copper ores.

The first smelters would have used the first, best surface ore. As smelting became more widely used, as demand for lead and copper increased, smelters would have had to use whatever ore they could, even it it was clearly not the best. If a batch of ore is not pure, smelting may not work well. The reactions inside the furnace might produce copper, but if there was too much iron ore or quartz in the ore, it would remain in the copper mass and make it useless for further working. The furnaces were not hot enough to melt these impurities so that they would separate from the copper. The ancient smelters solved the problem in practical terms, but the thermodynamics that explains it was not understood until the late 19th century.

At some point, a smelter discovered that adding some extra ingredient (a flux) would help the smelting process. The flux interferes with the smelting reactions in a beneficial way, forcing the impurities to react to form more complex compounds that would melt inside the furnace. Once they melted, the new compounds would separate from the denser molten copper and float upward to form a layer of waste (slag) on top of it. The smelter would have set clay plugs into the walls of the furnace, and by breaking them open could tap off the slag while it was still molten, running it into a waste clay-lined pit at the side of the furnace.

Fluxes would have been discovered fairly quickly and easily. By great good fortune, it turns out that one of the best fluxes (even today) is fayalite, a mineral which has the chemical formula Fe2SiO4. Fayalite is extremely rare in the rocks of the Earth's crust, but it turns out that it is chemically just a 1:1 combination of quartz (SiO2) and  hematite (or ochre), Fe2O3. Given that quartz and iron oxide are very common minerals, smelters would have found that certain ores came with impurities that made the copper ores "self-fluxing". Once someone had noticed that, the smelter would quickly learn to add ochre to ores with silica impurities in them, and to add crushed quartz or sand to iron-rich smelter loads, to achieve just the right mixture to remove both of them from the copper as slag. At 1120° C, the right mixture of fluxes would form a fayalite liquid much lighter than molten copper, and the problem would be solved.