Eyewitness Accounts of the Early Mining and Smelting of

Eyewitness Accounts of the

Early Mining and Smelting of Metals in

Mainland Southeast Asia

BENNET BRONSON AND PISIT CHAROENWONGSA

In Southeast Asia, as in many other

parts of the world, the student of

premodern technology is not likely to find

that much relevant data exists in

indigenous historical documents. He must

therefore adopt other approaches: studies

of historical and archaeological objects in

museums, excavations of early industrial

sites, inferences from general theory, and

comparisons with those traditional

techniques that survived long enough to

be recorded by outside observers.

4 SPAFA JOURNAL VOLUME FOUR NUMBER TWO

The last of these approaches,the use of what might be calledethnohistorical analogy, is ofparticular importance in the studyof ancient metal production, forhere the conceptual gap betweenmodern and early technologies isespecially wide. A specialist familiaronly with the metal making systemsof the present day, carried out on a vast scale and using methods basedon the developed theories of half a dozen sciences, is likely to havedifficulty in comprehending thesmall-scale, rule-of-thumb methodsof earlier times. He may have evenmore difficulty evaluating them. Ifhe knows what to look for,laboratory analyses will help him tounderstand what could and couldnot be accomplished by preliteratetechnologists working with simpleequipment. But no amount ofexcavation and laboratory work islikely to give him an idea of thesocial and economic parametersinvolved: the time and laborneeded to perform a process, itsefficiency when measured by thecost of labor and raw materials, theways in which production mighthave been organized, and the effectsthat all this might have had on thelarger societies within which alltechnological systems are necessarilyembedded.

Archaeometallurgists—students of ancient metallurgy—

therefore tend to be deeply interestedin first-hand historical descriptions oftraditional non-Western techniques forextracting and fabricating metals.They are particularly interested indescriptions by eyewitnesses whowere knowledgeable about technicalsubjects. The ideal eyewitness wouldbe a professional metal manufacturer,geologist or mining engineer with a command of the local language, a knowledge of the local culture, andthe ability to write unambiguousdescriptions of what he has seen.Such paragons did exist amongWestern and Western-trainedtravellers in various parts of theworld during the 18th and 19thcenturies. Unfortunately, they werevery scarce in Southeast Asia.

Premodern metal productionin East Asia, the Middle East andEurope has been extensivelyrecorded, by outside and localobservers and even by actualpractitioners. In South Asia andAfrica, most of the relevant datacome from outsiders during the 19thand early 20th centuries. These dataare abundant and (in the case ofSouth Asia, at least) reliable. Forboth regions, the number of recordsof indigenous mining and smeltingruns into the hundreds (see, e.g.,the bibliographies in Cline 1937 and

in LaTouche 1917-8). In SoutheastAsia, by contrast, the total does notexceed two-score, and many ofthese tell us no more than that a particular metal was extracted at a given place during the time of theauthor's visit.

The reasons for this scarcity ofpublished data cannot be gone intohere. It is sufficient to say that localchroniclers and other writers were notinterested in the subject, that mostforeign travellers in Southeast Asiawere neither interested nortechnologically literate, and thatindigenous metal-making there seemsto have already been in decline bythe time that the first interestedoutside observers came on the scene.One of the authors (Bronson n.d.)has argued elsewhere that SoutheastAsian miners and smelters had longsuffered from foreign competition,first from efficient Indian and Chineseproducers and then from theexplosively expanding technologies ofthe West. Tin was the only metalwhere Southeast Asia held its own,due to the impregnable advantage ofpossessing about two-thirds of theworld's reserves of that metal. Forlocal producers of iron, copper andthe rest, however it had beencenturies since commercial prospectswere especially encouraging. In theface of severe and long-continuedforeign competition, the mining andsmelting of metals within SoutheastAsia survived only where foreign

SPAFA JOURNAL VOLUME FOUR NUMBER TWO 5

technology, first Chinese and thenEuropean, could be easily adaptedor where producers were shelteredby geographical barriers.

The few known eyewitnessaccounts of traditional metal makingin the region are therefore precious.Those relating to Indonesia and EastMalaysia have been collected andsummarized by Marschall (1968)and by Harrisson and O'Connor(1969); between them, these writersdescribe about ten cases for whichuseful details exist, all of them relatingto iron smelting in Borneo, Sulawesiand the Minangkabau area of Sumatra.We know of only one case that thosewriters missed: Abendenon's (1917-8:1713) description of iron making by a Bugis-ToBada group at Soroako ineastern Sulawesi. We also know of twocases from the Philippines: Santos (Jagor1875: 179-85) and Hernandez (Eveland1905: 16-18) on copper smelting by theKankanay of Mancayan in northernLuzon, and McCasky (1903) on ironsmelting and casting by Chineseworkers at Bulacan in centralLuzon. These three examples plusthose cited by Marschall and byHarrisson and O'Connor may closethe list as far as insular SoutheastAsia is concerned. We would besurprised if more than a handful ofother cases were to be found in theliterature.

The purpose of this article isto present the results of a firstattempt at collecting as many casesas possible from the Southeast

Asian mainland. The elevenaccounts summarized here do notrepresent closure. We believe that a number of other descriptions existin the geological l i terature onBurma and Vietnam. Moreover, wehave deliberately left out Yunnan insouthwestern China, which is notonly in many senses a part ofSoutheast Asia but also the home of a highly developed and well recordedtradition of metal production: indeed,there are more good accounts ofmining and smelting operations forYunnan than for all other parts ofmainland Southeast Asia combined.The availability of comprehensivesummaries of these accounts (e.g.,Tegengren 1923-4: 347-64; Brown1920) makes it unnecessary for us toconsider Yunnan in any detail here,but the proximity of the vast andsophisticated Yunnanese metalsindustry must constantly be kept inmind when considering premodernmetallurgy in the rest of the region.

We do not claim that thedata presented here relate toeconomic activities that wereimportant to the region at the timethey were practiced. They are in a sense no more than relics, obscureand small-scale technologicalsystems that managed to survive,often precariously, down to the timewhen some outsider happened to

notice them. Their importance lies inwhat they tell us about much earlierperiods, when Southeast Asia seemsto have been the home of a prosperous and even advancedmetals industry. The remains of thatindustry litter the landscape of thoseparts of the region that were rich inores and fuels, forming the subjectof an increasing number ofarchaeological investigations. It isthese investigations that the presentarticle is meant to assist. Althoughthe data are sparse, they show thata surprising variety of metals wasmade and that the processes usedwere by no means unsophisticated.Many elements of these processesare likely to turn up in futureexcavations.

What follows is thereforeintended in par t as anethnotechnological handbook forarchaeologists . For the sake ofconvenience, it treats the evidencerelating to different metals separately,arranging the cases by date within thesections that relate to the appropriatemetal. Only cases that involve actualdescriptions are included. These are oftwo kinds: accounts where the writeris himself an eyewitness and accountswhere the writer managed tointerview former participants,checking their memories as far aspossible by whatever furnaces andother metal-producing apparatusstill remained at the time of hisvisit.


A certain number of interestingdetails will be underlined in the"comments" paragraph that followseach description. However, we havenot attempted a full technical analysis,nor to follow up every trail thatopened along the way. The data arepresented as straightforwardly aspossible, leaving it to others toexplore their implications.

IRON MINING AND SMELTING1860s

Thanbo and other villages, Mount Popaor Puppadaung, ca. 20 kilometers eastof Pagan, central Burma. Personalobservation by W. Blanford (quotedby Percy 1864: 270-3) andreconstruction using interviews andon-site study by H. L. Chhibber(1926: 219-32).

The mines are open quarrieson the mountainside, yielding ironconcretions from a clay matrix. Themineral is hematite with somelimonite. Two samples tested byChhibber showed low iron contentsof 26.3 and 22.8 percent.

Smelting is done at variousvillages in the immediate neighborhood.The one visited by Chhibber still had 12furnaces that were visible; Blanfordimplies that he saw similar numbers atone locality. Chhibber reconstructs thesefurnaces as "circular or oval pits 3-4feet in diameter, dug in compactearth," as in a river bank, and"connected with a circular hole, a

little more than a foot in diameter,"which served for feeding thefurnace with charcoal. He goes onto say "it is remarkable that theyhad no opening to provide a blast."Blanford describes quite a differenttype of apparatus: shaft furnacesabout 3 meters high, with a rectangular cross-section and (likeChhibber's furnaces) dug into a bank that surrounds the shaft onthree sides; the fourth or front sideis walled in by a thin clay screen.There is a large opening to providea blast, for—like many reportedfurnaces in Africa but unlike anyothers in Asia except for a singlecase reported from Yunnan byMoore-Bennett (Tegengren 1923-4:352)—air was supplied by naturaldraft, not bellows.

Both Blanford and Chhibberagree about the general course ofsmelting: the furnace is filled withcharcoal and ore and then lit;charcoal but not more ore is addedduring the smelting run. A furnacerun continues until "the molten metalflows out by tapping the lower partof the furnace" (Chhibber) or until a solid bloom has formed on thebottom of the furnace, which takesabout 24 hours (Blanford). No flux isused (Chhibber and Blanford). Thebloom weighs about 20 kilograms(Blanford).

Although the smelters arepart-time farmers, in their role asironworkers they are specialists inthe production of raw metal that isfabricated elsewhere. Their iron wasformerly famous, being shippedthroughout the lowland area ofcentral and upper Burma. Productionceased in the late 1860s, shortly afterthe British annexation of the area.

Comment:

Blanford's description, thoughoften quoted in archaeometallurgicalwritings, was evidently not knownto Chhibber. Both were geologists inthe employ of the Geological Surveyof India. Chhibber was an oldBurma hand; Blanford was not, buthad seen and written about a number of traditional smeltingoperations in India. That he knewmore about practical non-Westernmetallurgy is shown by the fact thathe does not bother to say that fluxwas not used (it is almost neverused in simple bloomery furnaces)and does not suggest that the ironmight have flowed out as a liquid(which is highly improbable in a furnace of the low-temperature typehe describes). Because his visit tookplace during the wet season,Blanford did not see the furnaces inoperation either, but his account iscircumstantial enough to make itclear that he had good informants.Chhibber's curious statement thatthe furnaces "had no opening to


provide a blast" (a clear impossibility,as pointed out by Suchitta 1983) mustrepresent a misunderstanding of whathis elderly informants told him.They may have said that smeltingwas carried out, not without an airblast, but without bellows.

IRON MINING AND SMELTING1860s

Near Loei, Northeast Thailand.Personal observation by H. Mouhot(1864: 133).

Mouhot does not describemining directly but implies that theore comes from "immense beds ofmagnetic ore of a remarkably goodquality," found on a mountain inthe Loei area.

The furnaces are holes "abouta yard and a half square hollowedout close to the mountain." Themineral is piled up and smeltedwith charcoal. The "liquified iron"collects in a hollow in the furnacefloor, whence it is withdrawn whenthe smelting session is completed.The iron, presumably in the form ofa raw bloom, is then carried back tothe village where it is worked up ina hearth described as a "cavity,"supplied with air by a pair ofupright piston bellows operated by

a child.

The same people do thesmelting and smithing. They are notfull-time ironworkers, being "half-agriculturalists, half-artisans." Theirspades and cutlasses are shipped toall the surrounding provinces, "evenbeyond Korat."

Comment:

An explorer rather than anacademically trained scientist,Mouhot was not knowledgeableabout mining or metallurgy.Moreover, the published accountwas compiled without hissupervision from the rough notesthat survived his death in Laos, a few weeks after his visit to Loei.This may be why the account isincomplete. It says nothing aboutany bellows used with the smeltingfurnace, or about yields, smeltingtime or many other details thatwould be of interest. It is not in factabsolutely clear that Mouhotactually saw a smelting furnace inoperation. In spite of its deficiencies,however, the report is important. Itis one of the very few publisheddescriptions of smelting in an areawhich on archaeological andgeological grounds appears to havebeen one of the key metal-producingzones of ancient Southeast Asia.

IRON MINING AND SMELTINGCA. 1880

Villages of the Kui, Cuoi, Khouy,etc. ethnic group, Phnom Deckrange, Kompong Soai Province,Kampuchea. Personally observed byJ. Moura (1883: 43-6) and by E.Fuchs and E. Saladin (summarizedat length in Beck 1891: 1009-11).

Two kinds of ore are used,one called "heavy" and the other"light." The latter is preferredbecause of the hardness of the ironproduced, which makes it suitablefor weapons manufacture. Mouraseems to say that the "light" ore is"very rich with a 65-66% metalcontent." Fuchs and Saladin cite a less impressive 35-40% metalcontent without mentioning that theKui recognize two types of ore. Theminers apparently do their owncharcoal-making and smelting. Themines are in the Phnom Deck;smelting is done in the villages atthe edge of the range.

Because of the expense,furnaces tend to be built by groupsof villagers or even by severalvillages. Each is a long rectangle inplan, 9. meters wide, 2.5 m. longand about. 4 centimeters deep,made of clay mixed with whitesand. The top of this shallow basinis open: its walls taper from 20 cm.in thickness at the base to about 5


cm. at the lip. The furnace properrests on a solid stamped-claypedestal that raises the bottom ofthe hearth 80 cm. above groundlevel. Both ends of the furnace basinare pierced with holes for tappingslag. Each of the sides has 26 holeswith 26 tuyeres which project somedistance into the furnace. Alltuyeres on each side are connectedby bamboo tubes to a commonbellows. There are thus two bellowsin all. Each has the form of a drummade of stamped clay about 50centimeters in diameter, with a leather diaphragm attached at itscenter to a cord and a spring pole.The feet of three operators (to eachbellows) press the diaphragm down,forcing air into the furnace; thespring pole pulls the diaphragmback upward when the operatorsstep off, letting the bellows refillwith air. Two high bamboo screensprotect the operators from the heatof the open hearth.

About 250 kilograms of oreand 5-6 hectoliters of charcoal,arranged in layers, constitute thecharge. Once lit, the fire is blowngently at a rate of 8 strokes perminute for the first eight hours,then blown at a more vigorouspumping rate for four or so hoursmore. One heat is finished per day,yielding 10-15 kilograms of iron.

The bloom then is worked up

into half-kilogram bars with pointedends (illustrated in Colquhoun 1885:fig 10). This is done in a separatehearth, blown with a pair of uprightsingle-acting piston bellows (Moura)or a single horizontal double-actingwindbox bellows (Fuchs andSaladin). The iron bars contain"mixed particles" of iron and steelbut are otherwise to a high purity.They are highly prized by manypeoples of the surrounding area.The bars are traded all overCambodia, Annam and Siam. TheLao use them as money.

Comment:

This is the best description oftraditional smelting we possess fromSoutheast Asia. Interestingly, theprocess is one of the most unusual.The furnace, clearly described byboth sources, is unique amongrecorded bloomeries. The closestanalogue is found in the tatara furnace of Japan, which is also long,narrow and furnished with rows oftuyeres along both sides, although itis much larger and produces a different kind of iron. The Kuibellows is unique in Southeast Asia(nothing like them is reported inMarschall's comprehensive 1968survey) but is of a type common inBihar and Orissa in eastern India,where clay drum bellows withspring poles and often with severaloperators per bellows were widelyused by small-and medium-scaleironmakers of the preindustrialperiod.

The reason for the furnacedesign is not clear; one imaginesthat with a somewhat deeper basinand an excess of fuel, such a devicemight be able, like the tatara hearth,to produce a natural steel or even a mixture of steel and cast iron. Thebellows is without question animport from India. The fact thatthe Kui were not in close contactwith overseas areas at the time ofMoura's, Fuchs' and Saladin's visitsuggests that the importation mayhave occurred at a relatively earlydate.

IRON MINING AND SMELTING1880s-1920s

Bo Luang (or Baw) Plateau, ChiangMai Province, northern Thailand.Personal observation by A. R.Colquhoun (1885: 51); reconstructionusing interviews and on-site studyby E. W. Hutchinson (1934: 164-5)and by P. Suchitta (1983: 35-41).

The ore is mined at a depositabout 35 kilometers away. None ofthe sources describe the mines ormining procedures. The ore is "ared oxide" according to Colquhoun,but Suchitta was shown a samplethat resembled magnetite. Womendo much (or, according toColquhoun, all) of the mining work.The mines are considered theexclusive property of Bo Luang


village. The ore is carried there forsmelting on elephants (Colquhoun)or on the backs of women and men(Hutchinson and Suchitta).

The smelting furnaces arecubical clay structures measuringtwo and one-half feet on a side(Hutchinson) or about one meter high(Suchitta). They are supported withwooden boards (Hutchinson) or a clay-covered bamboo framework(Suchitta). There are two largeopenings in the furnace, a 10-20centimeter hole in the top and a 10-15 cm. hole in one side that servedfor extracting the bloom and,perhaps, for tapping the slag. A third opening admits the tuyere,which is loosely connected to thebamboo tubes that leads to pair ofupright single-acting piston bellows.Suchitta suggests that the loose fitbetween the tuyere and the tubesmay have made it easier to drawair into the bellows.

The size of the charge andsmelting time are not given.Hutchinson notes that the "irondroppings from the ore are collectedfour times and returned to thefurnace." Suchitta had a piece of oldslag analyzed and found that it hada very low percentage of ironoxides (about 8% FeO and Fe203)which may confirm some recyclingof slag, although Colquhounmaintains that the smelting methodused is so "rough" that it "yields

only fifty percent of the metal." Thelump of iron that is extracted at theend resembles a small orange(Colquhoun). One such bloom thathad been preserved as a keepsakewas weighed and analyzed bySuchitta: it weighed 0.94 kilogramsand contained 93.49% iron, 0.28%carbon, 0.02% manganese, 0.052%phosphorus and 0.004% sulfur; therest was presumably slag.

The bloom seems to havebeen reheated and forged in thesame furnace as that used forsmelting, and the same individualsdid at least some of the work.Although women did much of themining and smelting, and althoughmen must have done a good deal ofthe smithing, none of the sourcessuggest a rigid sexually-basedseparation of different parts of theprocess. The knives, bells, tongs, etc.thus produced were traded to theChiang Mai area. Colquhoun saysthat in his day the village paid ironarticles as tribute to the "Zimmechief," the governor of Chiang Mai.The tribute articles were elephantchains, spear heads, cooking potsand other ironware.

The inhabitants of Bo Luangare Lawas, a tribal people speakinga Mon-Khmer language. Theircustomers were northern Thais.Suchitta and Hutchinson imply thatthey were not full-time ironmakers.Colquhoun, however, says flatly thatthey also were not "agriculturists."

All sources agree that the Lawawere large-scale raisers of livestock:cattle and, according to Colquhoun,elephants. They were rich in thelatter animal.

Comment:

Between them, the threesources give a clear but somewhatundetailed picture. One is left witha number of questions concerningmining methods, ores, inputs, yieldsand costs. It would also beinteresting to know more about thefurnace used: in spite of its boxlikedescription it may just be anordinary short shaft furnace withthick walls. One of the first writersto connect the Lawa with smelting(of lead, not iron) was Crawfurd(1828: 419); their propensity formetallurgy was widely known inthe late 19th and early 20thcenturies (see, e.g., McCarthy 1902:120). The pattern of tribal peoplesmaking iron for sale to state-organized peoples also appears inCambodia (see the description of theKui in this paper), in Borneo andSulawesi, and in the Khasi Hills ofnortheastern India.

IRON MINING AND SMELTING1890

Loi Nam Lin (Doi Nam Lin?), LaiHka Circle, N. Pang Long State,


eastern Burma. Personal observationby G. B. Stirling, quoted in full byScott and Hardiman (1900: 299-301).

The ore was mined fromthree deep, narrow shafts and froma number of open workings on a hillside, apparently not far from LoiNam Lin village. Formerly severalvillages worked at mining the ore,carrying it to be sold to the smeltersat Loi Nam Lin. At the time ofStirling's visit in 1890, smelting wascarried out only on a smaller scale.The smelters seem to have donetheir own mining by then.

The furnace is made of earthand has two large openings, a lowerone for loading in charcoal at thestart and an upper one for feedingin charcoal and ore, a handful at a time, while smelting is under way.The fact that (upright piston?) bellowsof bamboo are used suggests thatthere is a third opening for thetuyeres. Nothing is said abouttapping slag. Two men are needed totend the furnace. Each smelting run,done only once each day starting at2 am, lasts about four hours. A bloom made experimentally "in animprovised furnace" weighed about5 kilograms.

Four blooms can be made infive days, the fifth being reservedfor marketing. The iron is boughtby smiths in neighboring villages in

Pang Long but some goes to MongNai (Muang Nai) State. Theproducts of the Pang Long smithiesare well known. Some are shippedas far as Mae Hong Son andChiang Mai in Thailand.

Comment:

No data are given about the natureof the ore, the size and form of thefurnace, or many other details thatare of interest. However, it is worthnoting that ore and charcoal are fedin continuously. This implies thatthe furnace is of the shaft type; a yield of 5 kilograms in four hourssuggests that is not very small: a height of 1.5-2 meters seemsreasonable on analogy with Indianand African furnaces. The smeltersappear to have been natives of thearea; that Scott does not bother toidentify them means that they areprobably Shans. Scott and Stirlingboth regard the Loi Nam Linsystem as a mere relic of the larger-scale system of production thatexisted before the British conquestand the introduction of cheapBritish iron and steel. Their opinionis in part confirmed by the evidentcapacity of the furnace. It they haddone several runs per day, the twofurnace workers would each haveproduced about 10 kilograms/day—a quite respectable level of labor

productivity by preindustrialstandards.

IRON MINING AND SMELTINGCA. 1910

Uang Sa Pong (Wang Sapung), 26kilometers S.E. of Loei, N.E.Thailand. Personal observation byHogboom (1913-4: 78).

The mines, about 5 kilometersnorthwest of the village, are just"shallow diggings to get the loosepieces." The excellent hematite orewas analyzed at 68.5% iron, 0.042%phosphorus and 0.006% sulfur.

This was smelted at the villagein small furnaces approximately one-half meter high, about which nofurther details are given. Althoughsmelting had almost died out at thetime of Hogboom's visit, the villagestill housed an important smithingindustry, using imported Englishiron bought from Chinese traders.

Comment:

This very brief description issignificant mainly because itrepresents one of the few caseswhere iron smelting in SoutheastAsia survived into the 20th century.It is interesting to note that thevillage visited by Mouhot in the


Loei area also did both smeltingand smithing. Judging fromMouhot's description of his route,the two villages are not the same.

IRON-ARSENIC MINING ANDSMELTING

CA. 1910

Ban Hoei Tat, 40 kilometers north ofLoei, N.E. Thailand. Personalobservation by B. Hogboom (1913—4:79-81).

The mine is on top of a mountain to the northeast of thevillage. The two diggings visiblethere are several meters deep andmay be following a weathered veinof minerals. The ore was analyzedat 34.5% iron oxide (Fe203), 47.5%arsenic oxide (As205) and 16.2%water, with a little pyrite and tracesof nickel and cobalt.

Smelting is done, apparentlyright at the mine, in "small furnacessunk in the ground." The localpeople are aware of the poisonouscharacter of the furnace fumes. Noother details of the smelting processare given, but it is clear that themetal produced has a lower meltingpoint than ordinary iron. Hogboomhad a piece of the alloy analyzed.He reports the result as "Cu —93%,As —7%." The "Cu" here isevidently a printer's error; the alloymust actually contain 93% of ironand 7% of arsenic.

The local people "did notthemselves know of any other usefor the metal than casting bullets,"but formerly traders from the Southhad bought it. Hogboom speculatesthat this may have been for makingBuddha images.

Comment:

An iron-arsenic alloy like theone described would be easy to castbecause of its low melting point butwould be too brittle for most uses.The idea that it was used forcasting religious objects is plausible;antimony alloys are known to havebeen used by northern Thais forsimilar purposes (see, eg, Gardener1967: 1-2). What equipment wouldbe needed to smelt such an ore isnot clear. Deposits of iron andarsenic oxides are not common, andvery few cases of smelting them arereported in the literature.

LEAD, SILVER (AND ZINC?)MINING AND SMELTING

CA. 1400-1850 AD

Bawdwin or Bawdwin-gyi, TawngPeng-Loi, N. Shan States, N.E.Burma. Reconstruction usinghistorical data and study of the site,abandoned 40 years earlier, by T. D.LaTouche and J. C Brown (1908:235-63).

The gigantic deposit consistsof galena and anglesite (lead sulfate)in more or less intimate associationwith zinc blende, along with lesseramounts of pyrite, chalcopyrite andsilver. The silver content is variablebut high. A sample picked up in1900, 400 years after the richestdeposits began to be cleaned out,showed 88 ounces of silver per tonof ore; run-of-mine ore in the 1920sand 1930s still averaged 18 ounces/ton (Bender 1983: 178). Certain oreswith a very high zinc contentappear to have been avoided by theearly miners. LaTouche and Brownsuggest that it was not possible tosmelt these by the methods thenavailable.

The mines consist of largenumbers of shafts, galleries andopen cuts that honeycomb the sidesof a valley for a distance of aboutfive kilometers. At least one suchgallery cuts through solid rock forseveral kilometers. A number arewide and high enough to admitpack animals (mules, according toScott 1900: 303). Galleries andtunnels are skillfully cut althoughrarely supported with timbering.Many were still in good conditionat the time of the authors' visit.

Smelting took place in largeopen hearth furnaces, thought tohave been without bellows,composed of a deep fire chamber infront and a sloping, bowl-likereduction space in back which was


heated from beneath by flues. Theliquid metal trickled out of the bowland fell down through the firechamber to be collected in a hollowat the bottom. LaTouche and Browndo not speculate on the reason forusing such a furnace, althoughBrown comments that the processinvolved was clearly of the "roastand reaction" type—that is, a smelting procedure where leadsulfide is roasted to an oxide whichin turn reacts with and decomposesmore sulfide. Brown, an authorityon Chinese metallurgy, notes thathe has seen no furnaces like thesein Yunnan.

Cupellation, separating thesilver from the lead, is believed tohave been accomplished in twostages, the first in round beehive-shaped furnaces and the second insquare furnaces enclosed in stonebuildings, presumably for security.The lead oxide or litharge left overafter the extraction of the silver wasprobably resmelted to metal: noneof the sources mention finding largequantities of litharge. The slagresulting from the initial smelting,however, had a very high leadcontent. Between 1909 and 1919 a British firm recovered more than180,000 tons of old slag atBawdwin; this had an average leadcontent of 60 percent (Bender 1983:176) but a silver content of only1.7-1.8 ounces/ton (Brown 1917).

Mining at Bawdwin is thought

to have begun by AD 1412, the dateof a Chinese inscription found there.All miners and smelters wereChinese, although some lead smeltingby local Kachins, using the same leadslags as the later British company,took place there in the late 19thcentury. Crawfurd (LaTouche andBrown, p. 236) heard in 1827 thatthe mine produced 960,000 ticaisweight of silver annually and that itemployed 1,000 miners. Oldham(1855: 345) puts the number ofminers at 10,000.

Comment:

No one who had worked at themine, which closed in the 1860s foruncertain reasons, seems to have beenstill in the area at the time ofLaTouche's and Brown's visit. Theirreconstruction of processes is thereforebased on educated guesswork,supplemented by the small amount ofexisting historical information. Both ofthe authors, on the other hand, wereexceptionally knowledgeable aboutAsian ethnometallurgy. Their guessesare likely to be fairly close to thetruth. As noted above, they do notspeculate about the reasons why theunusual smelting hearths were used,but Brown's comment about a "roastand reaction" process may providethe clue. While such processes werenotoriously inefficient in extractinglead, they were economical of fueland were capable, if properly run, of

extracting virtually all of the silveroriginally present in the ore. Thiswould seem to have fitted the needsof the Chinese smelters perfectly. Fuelwas scarce, and the distancesinvolved would have made it veryexpensive to transport the lead to anymarket able to absorb such massivequantities of that metal.

Interestingly, although Scott(1900: 304) felt that the minesprobably had been exhausted by1860, they were reopened in 1914and are still in operation. In themid-1930s Bawdwin was producing70-80,000 tons of lead annually,along with an equal quantity of zincand perhaps 175,000 ounces of silver(Bender 1983: 178). As noted above,run-of-mine ore during this periodstill had an average silver content of18 ounces per ton. This would havebeen a "payable" but not spectacularore in the eyes of most preindustrialsilver miners. On the other hand, ifthe ores exploited during the 15th-19th centuries resembled the 88ounces/ ton sample picked up in1900, Bawdwin would have beenone of the largest producers ofprecious metal in Asia.

LEAD AND SILVER SMELTING1865

Kyaukthat, Yawng Hwe State, S.Shan States, E. Burma. Personal


observation by Fedden, quoted infull by Scott and Hardiman (1900:403-2).

Nothing is known of miningoperations, as mine locations were keptsecret. At the time of Fedden's visit in1865, the ore was brought to Kyaukthatwhere it was sold to smelters. The oreis a lead mineral, presumably galena,with an unspecified silver content.

After it is crushed, the ore isput into a small "cupola or blastfurnace" made of clay two and one-half to three feet in height and 16inches in diameter. The air blast issupplied by women standing onraised platforms, presumably usingthe usual upright paired pistonbellows; each furnace has two suchwomen.

The reduced lead is ladledout (of a receptacle at the bottom?)and transferred in ingot form tosmall "reverberatory" cupellationfurnaces. These consist of a firechamber separated by clay barsfrom the molten metal below. Asthe lead oxidizes, "it is removed bygently revolving over the surface aniron rod around which the lead inthe form of litharge (ie, lead oxide)accumulates in a number of coatingsor layers, one upon the other." Therelatively pure silver that remains intaken out as a button or plate fromthe bottom of the furnace. The"rollers" of litharge are thenresmelted and the lead metal sold.

Comment:

As in other cases where Scottdoes not specify an ethnic group, thesmelters here are probably Shans. Itwill be noted that several features ofthis particular method of cupellation,including the clay bars and thelitharge "rollers," should beidentifiable archaeologically.

LEAD AND SILVER MINING ANDSMELTING

1890s

Maw Son or Bawzaing, MyelatDistrict, S. Shan States, Burma.Personal observation by J. G. Scott(Scott and Hardiman 1900: 301).

The mines are on a hillsideone mile northeast of Maw Sonvillage. They are shaft mines, mostor all of which "descend to about300 feet before the miners follow upany veins." The lead ore (galena?) isargentiferous but averages only fiverupees weight of silver per 365-pound basket. Mining tools are "asmall hand pick, a mallet and a cold steel chisel. Two men take it inturns picking at the rock, whileothers carry the ore to the surface."There it is sold to the smelters, whocarry it back to their furnaces inMaw Son.

Each furnace has a dailycapacity of about five baskets of ore(about 1825 pounds) but is not

further described. Unlike theircounterparts at Bawdwin, thesmelters of Bawzaing were notinterested only in the silver in theore. Scott says "the main profitswere derived from the sale of lead."

Comment:

If, as Scott seems to implythe miners involved are Shans, thismakes Bawzaing the deepest mineknown to have been dug by nativeSoutheast Asians before moderntimes. The furnaces might bevertical shaft furnaces or might beone of several types of open hearth;that they are of medium size isshown by their daily capacity ofnearly one ton of ore. The silvercontent of the Bawzaing ores isindeed not high as such ores go;Bender (1983:182) says that oresmined there in modern timesaverage only 10 ounces to the tonof 50-55% concentrate. That suchores could be mined economicallyimplies a good market for lead, a low-valued metal, and hence cheaptransportation. Bawzaing is not farfrom Taunggyi and the mainoverland trade route to Chieng Mai.

LEAD-TIN ALLOY MINING ANDSMELTING

1890s

Benang Sta (Setar), near Biserat,Pattani River, Pattani Province, S.Thailand. Personal observation byH. Louis (1894: 235-6).


Comment: WOMEN, MAGIC AND METALMAKING

The ore occurs in theneighborhood of ordinary tin andgalena mines but is, Louis says,"quite unique, as far I know, in theworld. It has been produced by thegradual degradation of cassiterite-bearing granite and of limestonecontaining pockets of galena"; theresult is what the author describesas a natural "pewter mine" withores composed of oxidized leadminerals (anglesite with cerussite,pyromorphite and mimetite)intimately mixed with tinstone. Thedeposit is mined by sluicing awaythe overburden, breaking out theore with crowbars, crushing it withstone-headed tilt hammers workedby human foot power, and washinguntil a sufficient concentration is.reached.

Smelting is done in smallblast furnaces about seven feet high.The product, "besides a good dealof lead fume, is a natural alloy oflead and tin." A coin said to bemade of this metal was analyzed.It showed 69.4% lead, 27.9% tin,0.7% iron and 2.0% "impurities."

The miners and smelters areChinese. They sell the alloy "to therajahs of the surrounding states,who use it to make the small coinscalled puis."

Mixed lead-tin deposits are notordinarily considered to be smeltable.In this case, however, the originallead sulfide (galena) had been alteredby weathering to a mixture of leadsulfate and carbonate, which can bereduced to metal in the same way asa tin oxide (cassiterite) ore. Theremust be at least a few other depositsof this kind in the cassiterite-andgalena-bearing areas of SoutheastAsia. The blast furnace and tilthammer represent imported Chinesetechnology, but there is no reasonwhy such ores could not have beenhandled with less sophisticatedequipment in earlier periods. InKedah, pewter pitis coins go back tothe early 17th century; they werewidely used in Kelantan, Pattani,and Trengganu in the late 18thcentury and afterward (Show andKassim Haji Ali 1971). Consideringthat those kingdoms had much tinand little lead, and thus noincentive to make artificial tin-leadalloys, it seems possible that eventhe earliest pewter used in pitis wasa natural alloy made from an oresimilar to that mined at BenangSetar.

As we said in the introduction,we do not intend to discuss all of theimplications of these eyewitnessaccounts for archaeology and history,However, a few points are worthemphasizing.

It is common in other parts ofthe world to find that mining andsmelting operations are surroundedwith elaborate taboos and ritualpractices.

In Africa, taboos againstwomen at mining, smelting andsmithing places were (and still are)very widespread. The mereappearance of a woman at a smeltingfurnace often meant that work had tostop and that lengthy specialceremonies had to be performedbefore smelting could resume. InEurope and the Middle East, explicittaboos connecting women and metalworking seem to have been rare,but the idea of a women acting as a smelter or smith was almost asinconceivable as in Africa.

Nothing like this is reportedin Southeast Asia by any of oursources. Among the Lawa, womennot only engaged in all phases ofmetal making but, according toColquhoun, monopolized some ofthem. He says all smelting wasdone by women. A girl made ironblooms for her marriage dowry, andcontinued to be in charge ofsmelting the ore mined by herselfand her husband throughout hermarried life.

The other sources do not say


specifically that discrimination bysex in metallurgy is either presentor absent. However, the fact thatnone of them mention it can betaken to mean that it was weak orabsent in the places they wroteabout. Suchitta (1983) cites otherThai cases of women engaged inmetal working. It seems valid toconclude that in the ancientSoutheast Asian mainland womenwere not usually barred fromsmithies and smelting places. Theymust sometimes have run thoseplaces themselves.

None of the sources says muchabout metal working magic either,although that topic is very importantin writings on traditional metallurgyin Africa, ancient Europe, Indonesiaand even China. Smelters and smithsin the mainland portions of SoutheastAsia undoubtedly performed magicalceremonies to ensure that their workwould come out well. But the factthat the sources pay little attention tothese ceremonies may imply that theywere less conspicuous and perhapsless essential than in other regions.

ETHNICITY AND METALPRODUCTION

Because the metal makingoperations that survived longenough to be seen by outsideobservers were usually in isolated

areas, it is not surprising that themetal makers involved oftenbelonged to minority ethnic groups.What is surprising, however, is tofind these tribal metal makersselling their product to peoples withmore developed economic andpolitical systems.

That the Shans controlledsome of the most important lead-silver deposits in Southeast Asiaexplains why they could produceand sell these metals to the Burmeseand Thais. But what about the Lawaand Kui? Usable deposits of iron oreare present almost everywhere inSoutheast Asia. So why should thenorthern Thais, Laos and Cambodiansbuy iron from backward tribalpeoples instead of making their own?

We cannot offer a fullysatisfactory explanation. As notedpreviously, a number of other tribalgroups in Asia were formerlysuccessful exporters of iron—theKhasi in northeastern India, the NgajuDayaks of Borneo, and the Toraja andToBada of Sulawesi. All of these, likethe Lawa and Kui, formerly smeltedsubstantial amounts of iron for sale toconsumers in economically developedareas.

Part of the explanation maylie in superior ores. The Kui, forinstance, had access to exceptionallygood iron deposits in the PhnomDeck mountains. Yet there is nosuggestion in the sources that the

majority of successful tribal smeltersutilized unusual ores. Instead, thesources tend to imply that the tribaliron and iron implements weretechnically superior to those madeby the majority ethnic groups. Wemust give serious consideration tothe possibility that certain of themore "primitive" Southeast Asianpeoples were in fact moretechnologically advanced than theircivilized neighbors in terms of metalmaking and fabrication.

Perhaps the most interestingcase is that of the Kui. They arethought to have inhabited theirpresent homeland on the fringe ofthe Phnom Deck for many hundredsof years. The Phnom Deck is thelargest (and highest-quality) depositof iron ore in Cambodia. Thesmelting methods used by the Kuiare complex and ususual, involvingthe use of one item of equipment—the spring pole-operated drumbellows—that was definitelyborrowed from India, presumably atan early date.

What this adds up to is a fairly strong argument for the ideathat in former times the Kui wereimportant suppliers of iron toAngkor. We do not know whetherthere is epigraphic or other directevidence for this. But we do think itquite possible that weaponssupplied by the simple Kui were a


key factor in the military success ofthe greatest of ancient SoutheastAsian empires.

COPPER

Copper, important as a metalin its own right and also as thechief ingredient of bronze and brass,is naturally of great interest toarchaeologists. It is thereforeunfortunate that no outsider everseems to have seen and writtenabout traditional copper mining orsmelting in the non-Chinese partsof mainland Southeast Asia.

The well-recorded historicalYunnanese copper industry (manyreferences are given in Brown 1923)is not precisely relevant to ourunderstanding of early coppersmelting in the area south ofYunnan, for during the 17th-19thcenturies the copper mines andsmelting plants of that provincewere among the largest in theworld.

The previously mentionedrecords of tribal copper smelting atMancayan in the northern Philippinestherefore have special importance.They represent the only case of themining and smelting of copper by a Southeast Asian people to have been

observed and described in print.Interestingly, the Kankanay smeltersat Mancayan employed a notablycomplex method for extracting metalfrom the local arsenical copper sulfideores.

We might note that Jagor'sand Eveland's descriptions ofKankanay smelting are good. Theyare worth studying by anyoneinterested in the Southeast AsianBronze Age and in the more generalproblems of preindustrial technology.Rather similar ores are found inseveral parts of Thailand, Burma andVietnam. Some archaeologists haveassumed that such ores would havebeen avoided in antiquity because ofthe difficulty of smelting them. Andyet the Kankanay, whose materialculture is otherwise quite simple,appear to have handled those orescompetently. It is entirely possiblethat methods as sophisticated as thoseof the Kankanay were known at anearly date elsewhere in SoutheastAsia.

OTHER TOPCS

A number of importantsubjects have not been touched uponin this concluding section: thecapacity of furnaces; the productivityof labor; the geographical distributionof techniques; the locations of mines

and smelters with reference tomarkets and sources of raw materials;the unusual number of natural alloysproduced by traditional SoutheastAsian smelters; the existence ofspecialization in producing rawmetals as opposed to finished metalgoods; and so forth. Most of thosesubjects would require many pages todiscuss properly, and such a discussion might be premature inany case.

The reason why a discussionmight be premature is that newinformation is now coming in fromarchaeologists at an increasing rate.A new generation of specialists,some with extensive training in thetechniques of the metallurgicallaboratory, have become interestedin the study of early SoutheastAsian metals. Excavations focussedon archaeometallurgical problemshave begun to be carried out—forinstance, Piggott's and Natapintu'sas yet unpublished work at PhuLon in northeastern Thailand. Wetherefore expect that within the nextdecade our views on early SoutheastAsian metal production will havechanged drastically. Much of whatwe now think we know will havechanged drastically. Much of whatwe now think we know will proveto be incorrect. New interpretations,some of them still unimaginable,will replace those of the presentday.


Documents

Eyewitness Accounts of the Early Mining and Smelting of