74162 S5003133

12

Table: Results of cupellatlon experiments

Charge    Product

(melted)    (calculated assuming

		no losses; %)
Pb	395 g	Pb	91.88
Ag	25	Ag	5.80
Bi	5	Bi	1.16
Cu	5	Cu	1.16
Total	430 g	Total	402

This was cupelled to give:-

21.88 g Ag plus some slag Analysed composition of the silver bullion:-

Sample	Ag	Cu	Pb	Bit
1	97.8	0.3	1.0	0.9
2	97.9	0.4	0.9	0.8

<d) Tin

There are so few flnds of any type of tin-smelting furnace that lt is not surprising that llttle or no work has been done on thls subject. Pure tln-stone can be reduced relatlvely easlly ln a crucible, albelt at high temperaturę and thls ls the basls of many an assay process for tln ore.

But the remains of early tln smeltlng furnaces have been reported fron South Africa (Frlede & Steel 1976). Even here the evldence is meagre and only the foundations of saucer-shaped hearth furnaces were found near Roolberg. On the basls of these flnds and some ln Nigeria an experimental furnace was constructed as shown ln Fig. 11. Thls consisted of a shaft 60 cm high and 13 cm ln diameter, wlth one tuyere. Ore with a tln content of 6.3X was used whlch ls much below the quallty used ln Brltain ln the Medleval and later perlods. It was elear that tln oxlde was reduced but that the gangue kept the tln partlcles apart so that no masslve amount of tln was found. A better grade of ore (30.88% tln) gave tln prllls and larger nasses but no plano-convex Ingots and lt was elear that these could only be madę by remelting the prllls ln a crucible and pouring into a mould.

The purity of the tln was ln the rangę 98.8 to 99.15%.

In Brltaln, however, it was possible co obtaln pcbblcs of pure cassiterite whlch contains 79% tln, and the probleas found ln South Afrlca would not arlse. Even so, temperatures as high aa 1000*C ate needed and cara muat ba taken to see that all the charcoal la burnt away at the end ao that lt doea not keep the tln prllls fron coalesclng.

Concluslons and recommendations

Aa the precedlng shows, a great deal of experimental work has been done on Iron and copper and llttle on the other aetals. Korę people today are fanlllar with the early and now obsolete techniąues of Iron-aa king than they are with copper and there is stlll roon for further work on thls metal partlcularly the sulphlde ores.

On Iron, recent work by Clough (1985) has shown that the rlch sulphlde nodules - partlcularly the limonitized ones - can be smelted so thls minerał is now presented as a new source of Iron for the south of England. On the whole the probleas of Iron are well understood. But there ls scope for a study of the true bowl furnace and Its aanlpulation.

As far as lead and tln are concerned there is conslderable need for morę work. No work has been carried out on zlnc and antimony.

It ls only recently that traces of Roman and Hedieval zlnc smelting have co me to light and investigation and examlnation of the sites nust precede simulation experiments. But it will be of great value to have details of cementation and retorting techniąues. One exaaple of the latter ls stlll belng used ln Chlna and ls under lnvestlgation by the archaeo-metallurgy group ln Beljlng (Tylecote 1983).

How far antimony was recognised as a metal ln its own right in classi-cal and medieval times is stlll ln doubt. We know that antimony was being extracted in Southwest Scotland in the elghteentb century (Sinclair 1794) and it would be useful to slmulate thls as it ls a simple example of a liąuation process.

Precious metals such as gold and platinum are not thought to need much work done on them. Gold is mainly a matter of minerał dressing g a subject that comes lnto our archaeometallurgical field but whlch we will not be considering at thls conference. Even so we might mentlon Its importance and express the hope that morę workers will undertake work on these processes.

Platinum is mainly a South American problem and one of minerał dressing and physlcal metallurgy rather than smelting. While gold could be melted, platinum could not and could only be Consolidated by powder metallurgy or 'brazing' with gold and copper. Conslderable work is being done on thls subject at the Institute of Archaeology, London by Scott and at the Massachusetts Institute of Technology (Lechtman 1976).

The universities seen the right place for these experiments. They are simple to engineer and fairly cheap to instrument, and teach the prln-clples of extractive metallurgy. They can be mtegrated as short-term projects into the finał year of flrst degrees or ln the morę detailed form lnto post-graduate years. There is only one problem; a good deal of the space once available in universitles has been turned over to physl-