S5003141

22

density) becomes a possible feature for discrimination. Many of the coloured minerals have a veined appearance, sonie give strikingly facetted surfacee when broken and many are of a very shlny character, reflactlng the aun. We colltct these, perhaps to use ground up for colouring plcturea as a slurry, perhaps for personal adornment (for example, galena could be used as eye-ahadowl) or just to keep as attractive possesslons ln their own right as a mat ter of aesthetlc fancy. We know that many minerals were collected for their own sake ln Neolithic times.

There were one or two very atrange materials that we would occaslonally find. If we llved ln certaln areas there could be local concentratlons of a shlny yellow materiał, perhaps shinlng on a stream bed but of very different texture to the rock-llke minerals, being duet ile such that you could even bend it back and forth if the piece were thln enough; and it only deformed when hammered and did not shatter as all other ¹ stones' did, being ao different ln texture and behaviour. We know that such native gold, as we would cali lt, was collected and used by the late Neolithic people, e.g. at Varna ln Bułgaria. Additionally, assoclated with the strongly coloured copper minerals would be natlve copper, exhibiting similar properties of ductility: you could bend blti that stuck out of the host rock.

Such native copper, where it was found ln large enough pleces, could be worked to a shape directly. It may be that they found that native copper was best freed from assoclated rock by 'flresetting’, a technique already applied for breaking stones. The associatlon of heat wlth the handling of copper would then be found to ease the hammering process, sińce the materiał is sof ter when hot, and copper which had hardened under the hammer when worked as it cooled, or when cold, could be resoftened by heating again ln the fire, the cold-worked copper structure being re-crystallized into new grains, i.e. annealing would have been discovered. Perhaps, even, a copper rod hammered from natlve copper was found to be a useful tool for manipulating the flre -as a poker - and the softening effect was noted and utilised. Such annealing permltted the working of the natlve metals to thln sheet or section without cracking.

Once established, the associatlon of minerals and treatment by flre would inevitably lead to the first smeltlng, i.e. the reduction of compounds to the metal. For example lead would have been collected as the strlking and attrac-tive sulphide galena, heavy and easily identlfied by its facetted, shiny tex-ture. When heated on the top of the fire it would largely convert to lead oxide. Both copper and lead oxides are readily reduced to the metal at tem- i peratures achieved ln a fire burning charcoal or dry wood, but it is likely that lead was the first metal to be recovered in this way, sińce separation of the lead as molten metal from the reduction zonę and its collection in the hearth or the separate melting of furnace prill reąuires only temperatures in excess of 327*C. Copper, on the other hand, could only be agglomerated to a usable form at much higher temperatures in excess of 1083°C, achlevable only by a directed draught, whether forced or natural. Eventually, sufficiently hot conditions wlth draughting were developed to produce melting and, from this, the agglomeratlon and shaping of copper metal. The distinction has to be madę here with gold, which could be agglomerated simply by hammering pieces into close contact, sińce there are no oxide or other fllms present which prevent metal bonding. At the beginning of the Chalcolithic, c. 4000 BC, we find cast objects of copper, such as the axes from the Vlnca culture in the Balkans. It is difflcult to say whether the materiał melted is native copper or the product of the early smeltlng of rich copper ores. As already pointed out, the reduction of copper oxide is an easy matter at temperatures in excess of 700°C, above which temperaturę the CO/CO2 ratio inereases rapidly in equilibrium with carbon, givlng efflclent gaseous reduction. In pure carbon monoxide supplled from an external source malachlte can be reduced to

copper at beIow red heat, sińce the reactlon is no longer dependent on the direct presence ot carbon at 700*C plus to produce the carbon monoxid«.

How would anclent nan have learnt that the highly colonred green and blue minerals were a likely source of copper? Ia the first place aatlee copper ia often assoclated directly in juxtapoeltion with osclde minerals, and ha may hava found that his yield to metal did r.ot suffer, but lmcreseed if ha did not bother to aeparate the two before melting for aggloaeratlom. Secondly, he would have noted that heating the coloured minerals alone gave the same atrong green flamę coloratlon aa produced when he heated metailic copper. Attentlon would gradually change from the r.eed to find the increas— ingly rare native copper for melting to the strong heating of strongly coloured 'stones' which gave the green flamę. In this reapect I do not belleve that the discovery of smeltlng haa to be considered aa a aeparate lnnovatlve event.

Flame colour haa always been used by mineraloglsts as a slmple distin-gulahlng test for the compounda of certain metals, notably copper and lead. Lead minerale give a pale azure-blue flame. Purther slmple teata relate to the colour and form of encruatationa produced by heating and the degree of sublimation of the oocldes produced and to the odour produced on Heating - the pungent smell of the sulphur dloxlde when heating sulphldes and the garllc odour produced from arsenie.

It Is elear that early smeltlng was based on the oxidised minerals availabie at weathered surface outerops, beIow the malnly Iron oxlde surface gossan or Iron hat, green malachite (C1JCO3 .Cu(OH >2 ). blue azurite (2 CUCO3 (CuOH)2> and red cuprlte (Cu₂°)*

In the Balkans, certainly, we have heard of the first induatrlal depres-sion at the end of the Chalcollthlc period, partly assoclated wlth the worklng out of the oxldised minerals there. But wlth continued excavation the minera would have eventually encountered che minerals of che secondary enrlchaemt zonę, contalnlng the copper-rich sulphide minerals, chalcocite (Cu₂S)» covel-lite (CuS) and the mixed copper/lron sulphldes chalcopyrlte (Co23.Pi2S3) and bornlte 5 Cu2S.Fe₂S3, on which the deposlt was origlnally based, together wlth Iron pyrites. These attractive gllnting minerals would be trled ln the flre, and chalcopyrlte and bornlte would stlll have glven the strong green flame and all a pungent odour. On roasting they would glve a blach ozide, just as malachite does, but separatlon from the Iron component requlres a somewhat morę complex operation, nowadays achieved by the oxldation of a separated molten sulphide or matce, then by some selectlue reduction at che CO/O2 ratlos employed and the separatlon of metailic iron from the copper on remelting. It is probably unlikely that the ose of the morę complex copper-Iron sulphide minerals was assoclated wlth early copper smeltlng, although developing later in che fuli Bronze Age.

The introduction of flux

An lmportant dlstlnctlon between the mere melting of natlve copper and che reduction of even hand-plcked rich copper ores was the need to separata the mixture of flne copper droplets and high melting point solid oxldes - limę, silica, alumina, etc - to glve efficient collection either es coarse prlll or run-down onto the hearth. It would not necessarily folIow that the meltlag point of the oxlde system would need to be lowered to that of copper at 1063*C, but to a temperatura achlevable ln the draughted charcoal furnace. so that a degree of lląuld separatlon could be achleved, the two phases being lmmlsclble. The best flux, easily avallable, is Iron oxlde, partlcularly limonite avallable from the gossan or Iron hat,assoclated wlth the copper deposlt. The gossan lm