Zambczia (1979), VII (ii).TRADITIONAL IRON-WORKING IN CENTRAL AFRICAWITH SOME REFERENCE TO THE RITUALISTIC ANDSCIENTIFIC ASPECTS OF THE INDUSTRY *F. W. J. MCCOSHFaculty of Education, University of RhodesiaTHE ARCHAEOLOGICAL AND metallurgical features of African iron-workinghave been studied in depth,1 and research has been conducted on the econo-mics and demography of the industry.2 Yet there is little more than incidentalconcern with the accompanying ritual and none at all with the scientificimplications. An otherwise authoritative writer has complained of a regret-table result due to ca tendency to follow the views of an earlier school ofarchaeologists and anthropologists who regarded iron-working as an economicritual controlled by the need for secrecy, sexual taboo and exotic mystery'3 Ša statement which I consider to be unfortunate in suggesting that traditionaliron-smelting and forging need not be explained in terms of ritual. Believingthat in a study of traditional industries the ritualistic features cannot bedivorced from the mechanical descriptions of processes, the following is avery general paper on traditional iron-working but with more emphasis onthe ritual and on the possible scientific interpretations of this industry. Thepaper is, therefore, divided into the following sections:A brief history of iron-working in Africa.Two contrasted methods of iron extraction in Zambia and Rhodesia.The ritualistic aspect.The scientific aspect: practical and theoretical considerations.Conclusion.* I am deeply grateful to Dr J. M. MacKenzie of the University of Lancasterand visiting lecturer in the Department of History of the University of Rhodesia forpermission to quote from his recorded interviews in various Tribal Trust Lands, andfor an exchange of correspondence which, for me, clarified several aspects of Hwedzairon smelting. The preliminary draft of this paper was read by Professor D. H. Reader,Head of the Department of Sociology, Professor R. S. Roberts, Head of the Departmentof History, and Dr D. N. Beach, Lecturer in the Department of History, in theUniversity of Rhodesia and I wish to acknowledge their interest, encouragementand critical comments.' W. Cline, Mining and Metallurgy in Negro Africa (Menasha, Wis., Bauta, 1937):a good summary but with little reference to Southern Africa; R. F. Tylecote. Metallurgyin Archaeology (London, Arnold, 1962) ; S. V. Pearce, The Appearance of Iron and ItsUse in Protohistoric Africa (no details, mimeo : based on Univ. of London M.A.thesis, 1960).2 J. M. MacKenzie, 'A precolonial industry: The Njanja and the iron trade',NAD A (1975), XI ii, 200-20.3 M. D. Prendergast, 'Research into the ferrous metallurgy of Rhodesian IronAge societies', Journal of the South African Institute of Mining and Metallurgy (1973-4),15S156 TRADITIONAL IRON-WORKING IN CENTRAL AFRICAA BRIEF HISTORY OF IRON-WORKING IN AFRICAAlthough the metal iron never possessed the charismatic property of gold,it was, nevertheless, a symbol of inexplicable and stupendous preternaturalpower. Did not the first native iron arrive dramatically from the sky whereit was possibly associated with thunder and lightning as well as the un-predictable falling of stars? A Hittite text of the fourteenth century B.C.speaks of 'black iron of heaven from the sky',4 and as late as the close ofthe nineteenth century the Bedouin of Sinai believed that a man who suc-ceeded in forging a sword of meteoric iron was invulnerable in battle.5There appears to be no legendary history of the discovery of man-madeiron although it is possible that Pliny's account of the discovery of glassis analogous. A ship's crew is said to have made a fire on a sandy beach inPhoenicia and supported their cooking pots on lumps of native soda whichwere a part of the cargo. It is possible that nodules of iron ore, usually ironsulphide, were part of a vigorous camp fire, and as a temperature of 700°-800°C is possible in a small hole in the ground without the aid of an artificialblast, it is likely that small pieces of wrought iron were produced in thismanner for the first time. The original use of the metal is believed to haveoccurred in the first half of the third millenium B.C. in Egypt 'and the NearEast where iron beads were made from meteoric iron,6 but man-made ironprobably appeared between 1900 and 1400 B.C. Not until it could beproduced as a mild steel was it of great use; this was accomplished by furtherheating wrought iron to which charcoal was added.The Kushites, in the seventh century B.C., abandoned lower Egypt to anAssyrian army equipped with iron weapons, and established themselves atMeroe in Nubia where there were ample supplies of iron ore and wood whichmaintained an iron industry from the fourth century B.C. to the fourthcentury A.D.7 The diffusion of smelting and forging techniques from theSudan to southern Africa is still a subject of controversy and one whichis fraught with difficulties owing to the corrodable nature of iron, especiallyin acid soils. Direct evidence of iron-working is usually revealed by thepresence of slag, which being a silicate is indestructable, and by curved frag-ments of furnace wall together with the remains of broken baked clay pipes,or tuyeres, which were part of induced-draught furnaces where air enteredat the bottom of the furnace. The diffusion of iron-working may have fol-lowed two routes; westward from the Nile valley to the Chad region in the* J. W. Mellor, A Comprehensive Treatise on Inorganic and Theoretical Chemistry(London, Longmans Green, 12th edit., 1957), 483.= M. EKade, The Forge and the Crucible transj. S. Corrin (London, Harper andRow, 1962), 27, citing W. E. J. Bramley, The Bedouins of the Sinai Peninsula(London, Palestine Exploration Fund, 1906).s Pearce, The Appearance of Iron, 9, 28.7 A. J. Arkell, 'The Iron Age in the Sudan', Current Anthropology (1966), VII,F. W. J. McCOSH 157first few centuries before and after the birth of Christ, from whence thetechniques travelled south and west; or southwards from Libya where theBerbers, who had learnt iron-working in the Sudan, passed on their skills toWest African peoples and thence southwards again to those of the savannahsand the tropical forest.8 Vague references to the iron mines in the hinterlandof Sofala were made by Ibn'al-Wardi of Cairo in 13409 and even earlier byEdrisi of Sicily in the twelfth century.10The disappearance of the industry in the first two decades of thiscentury may be ascribed to reasons which are commercial, humanitarian andpolitical. The introduction of factory-made imported steel articles contributedlargely to its decline. Thereafter, the iron-worker dispensed with the arduoustask of constructing a furnace although forging continues to this day fromscrap metal such as the leaves of car springs. The factory-made hoe is notas susceptible to rusting because it is fashioned from a homogeneous caststeel whereas the traditional hoe contained areas of wrought iron, slag andseveral grades of mild steel which, in the presence of water charged withatmospheric carbon dioxide, gave rise to those electrochemical reactionswhose end-product is rust." The humanitarian reason concerns the ban onthe use of goat-skin for the bellows12 which was obtained by flaying theanimal whilst alive. It was believed that the most efficient bellows wereobtained from goats which survived this ordeal until nightfall.13 Althoughthe flaying of live goats was forbidden in Rhodesia from 1897 the use ofimported factory-made hoes may have encouraged the abandonment of thispractice. The political reason arose from the manufacture of bullets fromlocally forged iron rod which were fired from rifles brought from Kimberleyby Africans returning from the diamond mines. Gunpowder was locallymanufactured from the faecal droppings of the Rock Dassie (Procaviacapensis) or its dried urine found in caves, both of which contain saltpetre.These were mixed with charcoal or slag, the latter probably containing resi-dual charcoal. With the extension of colonial rule it was not unexpectedthat the forging of bullets was discouraged by the confiscation of guns,14although iron-working at Hwedza, Rhodesia, had already been banned bytwo neighbouring District Commissioners. Thus, the industry which gaveits name to a widespread culture came to an end and only the occasionalexhibition by surviving members of smelting families enables us to speculatea Pearce, The Appearance of Iron, 12-13, 75.a H. Salt, A Voyage to Abyssinia (London, Rivington, 1814), 57.10 G. Ferrands, 'Sofala', in The Encyclopaedia of Islam (Leyden, Brill, 4 volsand supplements, 1913-38), IV, 470.11 G. H. Stanley, 'Some products of native iron smelting' South African Journalof Science (1931), XXVIII, 131.«2 J. H. Chaplin, 'Notes on traditional smelting in Northern Rhodesia', SouthAfrican Archaelogical Bulletin (1961), XVI, 60.Ł» J. S. Hatton, 'Notes on Makalanga iron smelting", NAD A (1967), IX, iv, 40-1.I* MacKenzie, 'A pre-colonial industry', 218.158 TRADITIONAL IRON-WORKING IN CENTRAL AFRICAon the former Hwedza countryside, described by an English missionary in1893 as 'the Wolverhampton of Mashonaland', a Black Country in whichevery village had several forges.'5TWO CONTRASTED METHODS OF IRON EXTRACTIONIN ZAMBIA AND RHODESIAThe principle of traditional iron extraction is basically the same as thatof modern industrial extraction because it consists of heating strongly amixture of iron ore and carbon. It is the reduction of an ore which is actually,or potentially, an oxide of iron, with charcoal. The reaction inside the furnacecan be simply expressed asIron oxide + Carbon = Iron + Carbon dioxideŠ perhaps too simply stated because an intermediate stage is the formationof carbon monoxide gas which is the effective reducing agent.For the purpose of this paper iron-working techniques can be dividedinto those where shaft furnaces were employed, as in Zambia, and thosein which the smaller 'beehive' furnaces were used as in Rhodesia. I haveomitted the more primitive 'open bowl' or 'pit' type of furnaces as used inpre-Roman Britain and found in various parts of Rhodesia, such as Inyanga'6and Khami, which suggest an earlier technique for which there is no evidenceof ritual. That the shaft furnace was a development of the pit furnace isshown by the Roman introduction into Britain in the second century ofa furnace which was no other than the pit type with its sides built up toa height of just over one metre and blown by induced draught, whereas inthe simple pit furnace bellows were used.17The shaft furnace, typical of Africa north of the Zambezi, has beencarefully documented from an example in south-west Tanzania where theheight was about 3 metres, the diameter increasing from 0.75m at the topto 1.5m at the bottom, the walls having a uniform thickness of 13cm.18Although the custom was often relaxed, smelting was generally afamily skill, a new shaft furnace being constructed when a young memberof the family was considered physically suitable to pursue the trade. Hisskilled relatives constructed a new furnace with clay, inside of which the>5 I. Shimmin, 'Journey to Gambisa's', in F. W. McDonald, The Story of Mashona-land and the Missionary Pioneers (London, Wesleyan Mission House, 1893), 51, 55.is R. Summers, Inyanga, (Cambridge, Univ. Press, 1958), 61, 114.«7 Tylecote, Metallurgy in Archaeology, 220. However, the pit furnace, althougha poor heat conserver, could be used for the rapid smelting of successive charges ofore and charcoal without breaking down a part of the furnace as in shaft and beehivetypes, but as the slag was not tapped it was necessary to remove it before smeltingcould be repeated.>s B. P. Brock, 'Iron working amongst the Nyiha of south-western Tanganyika'South African Archaeological Bulletin (1965), XX , 98.F. W. J. McCOSH .159apprentice stood until it was at a height which he could just reach. A polewas placed over the top opening and the relatives, supported on a scaffolding,would raise him very carefully out of the furnace because it was a bad omenif any of the soft clay was disturbed during this ritual. Once outside thefurnace he danced and sang, 'Now I am grown up I am taller than thethatching grass'. He was then entitled to call himself n'anga (doctor), andwas alone responsible for the stacking of firewood, iron ore and charcoalwhen smelting was about to start, and also responsible for starting thefurnace fire. Before the clay of the furnace had dried, a cockerel's headwas chopped off by the senior 'doctor' and the blood sprinkled over thewalls of the furnace to ensure that it would not crack or collapse duringsmelting.19 The clay-pipe tuyeres, or air pipes, of 5-10 centimetres diameterwhich allowed air to enter the bottom of the furnace, were prepared bymen although pottery was normally a woman's trade.20 To ensure anefficaceous draught a furnace would have from 40 to 80 tuyeres.The ore was either a haematite (ironstone) consisting mainly of ferricoxide, or limonite (bog iron), a hydrated ferric oxide found around theedges of marshy depressions. There appear to have been no taboos in themining and transportation of ore, and the absence of women from open-castmining may have been for physical rather than ritual reasons. Ore depositswere sought by the older men after which young men were sent to excavatewhile women assisted in transporting ore to the furnace site in oval basketscarried on their heads or on the backs of oxen.21 Pulverizing of the ore waseffected by roasting, prior to which the ore was upgraded by hand-picking."Charcoal was obtained by burning logs from trees selected for their hardtimber, broken up with rakes and carried by women to the furnace inbaskets.23The careful stacking and firing of the furnace might suggest that theseprecautions were necessary to avoid the effects of counter-magic by a rivalsmelter, yet there is never any mention of such sorcery, perhaps because ofthe smelter's prestige; it was he who knew the 'medicines' necessary forsuccess. These were mainly preventitive, countering the disastrous effectis W. V. Brelsford, 'Rituals and medicines of Chishanga ironworkers', Man(1949), XLIX, 27-9.2° H. B. Barnes, 'Iron smelting among the ba-Ushi', Journal of the Royal Anthropo-logical Institute (1926), LVI, 190.2' G. Kay and D. M. Wright, 'Aspects of the Ushi iron industry', Northern Rhode-sia Journal (1962), V, 30-1.22 R. F. Tylecote, 'Iron smelting in pre-industrial communities', Journal of theIron and Steel Institute (1965), CCIII, 340.23 MacKenzie, 'A pre-colonial industry", 214, mentions the Hwedza trees usedfor charcoal burning, mushava (Monotes glaber), mukarate (Burkea africana) andmyange (=muvanga?) (Pericopsis angolensis). E. Mambaiye, of the Gwelo Teachers'College, includes all these with the addition of mupembere (Combretum molle), 'Ku-cherwa Nekupfurwa Kwesimbi Kare' (unpubl. essay).I6O TRADITIONAL IRON-WORKING IN CENTRAL AFRICAof the presence of menstruating women, or of men who had recently indulgedin sexual intercourse, or even of those married unlawfully according to localcustom. But counter-magic, interfering with the production of hoes, andtherefore food, could not have been tolerated. Even the members of thesmelting crew were forbidden sexual intercourse during the period of smelt-ing; they slept by the furnace while their wives were regarded as temporarywidows, a male 'wife' being employed to cook for the smelters. The centralbelief was that the furnace was a symbol of woman kind who gives birthto iron which can be fashioned into implements of war or agriculture,providing protection from enemies and ensuring an adequate sustenance.The furnace was the 'wife' of the smelters while iron production was inprogress and therefore sexual intercourse was regarded as adultery. Purifica-tion rites were then obligatory, or medicines were placed in the furnacebefore firing. As miscarriages and still-births were attributed to the un-faithfulness of the husband, so sexual intercourse could result in a smeltingfailure where most of the iron was found not in the furnace but in themolten slag which flowed from it, an analogy with menstrual blood.24 Slagwas truly a waste product; it found no use, not even as a fertilizer, and wasreferred to by the Ngoni of Malawi as 'faeces'.25 Iron smelting was ananthropocentric ritual analogizing production as reproduction. As withother traditional projects the success of the undertaking depended partly onskill but also on the approval and acquiescence of the ancestral spirits towhom prayers were addressed before mining commenced, before a furnacewas constructed and before firing a furnace. Sometimes it was the spirits offormer smelters who were petitioned.26 An interesting compromise isrecorded during the 1939-45 wartime revival of iron-working in Zambia,due to a shortage of hoes, when a smelter led in prayer the children froma local mission school.27Rhodesian iron-working was similar in principle to that of Zambia butit was distinguished by the type of smelting furnace used. Whereas theZambian furnace was a tall induced-draught kiln, the Rhodesian 'beehive'furnace was a smaller forced-draught unit about one metre high. Althoughsmelting in Zambia was effected in an induced-draught furnace, the sub-sequent forging employed a smaller forced-draught furnace using drum-bellows which were constructed by stretching an animal skin over a clay orwooden bowl. In Rhodesia both the smelting and the forging furnace useda forced draught supplied by goatskin bag-bellows. Reproductive symbolismin Rhodesian iron-working was emphasized by moulding stylistically theclay furnace as a woman about to give birth, with breasts, tribal cicatrices" Kay and Wright, 'Aspects of the Ushi iron industry', 36.25 S. N. Stannus, 'Nyasaiand: Angoni smelting furnace', Man (1914), XIV, 131-2.2« Barnes, 'Iron smelting among the ba-Ushi', 192.*? Chaplin, 'Notes on traditional smelting'. 54.F. W. J. McCOSH 161or nyora, umbilicus and genital organs. The goatskin bag-bellows and claypipe tuyere represented testicles and penis and thus, again, iron smeltinganalogizing production as reproduction.Although it appears that the production of iron implements wasgenerally for local use there were certain export centres. It can be readilyassumed that the larger production of iron in the shaft furnaces of Zambiaimplied an export market but more definite evidence concerns the Hwedzaindustry of Rhodesia where a labour-intensive organization successfully ex-ploited the rich haematite reefs of Hwedza mountain, although this sourcewas in enemy territory belonging to the Mbire tribe. Iron ore was alsoexported, not only from Hwedza but from Shamva and Mangula (mhangura= iron ore), but it was among the Njanja tribe, near Hwedza mountainthat there evolved in the eighteenth and nineteenth centuries an industry des-cribed as the 'Wolverhampton of Mashonaland'. The missionary who coinedthis phrase arrived at Gambisa's kraal with a broken clasp on the end ofthe disselboom, the pole to which the oxen were harnessed. That the claspwas repaired showed that the Njanja were adept at welding and drillingholes in iron.The construction of a 'beehive' furnace was well described after anexhibition of iron-working in the grounds of the Queen Victoria Museum,Salisbury, in 1944, superintended by Chief Ranga of the Njanja tribe.Ranga first dug a circular trench as a foundation; from this hepiled up the clay into a conical shape, resembling a female torsobearing tribal markings. This was built up as a solid mass duringthe day and left to set and dry overnight. The next day a stick wascarefully put down the centre from the top for slowly hollowingout the inside. Ranga then burrowed underneath the cone andshaped the hearth, which was hollowed out well below the base.For the next few days a small fire was kept going in the fresh-madefurnace, to dry out the clay. Ranga was careful not to mould theinside too quickly in case it collapsed.28It has been alleged, with some justification, that at some such exhibitionsthe result of the smelting was invariably a lump of poorly reduced iron or'bloom', or a total failure. The photographs accompanying the article fromwhich I have quoted show clearly the successful production of adzes, spear-heads and hoes. Yet at the Witwatersrand Agricultural Society's show aiJohannesburg in 1930 when an Njanja smelting crew from Hwedza wasperforming, it was recorded that the bloom was unrecognizable as iron,having charcoal and slag enmeshed in its cellular structure. Many reheatingsand forgings were required before implements could be fashioned from it."28 E- G> 'Iron smelting and smithing in Africa', The Outpost (1944), XXI, vii,*» Stanley, 'Some products of native iron smelting', 132.162 TRADITIONAL IRON-WORKING IN CENTRAL AFRICAThe Hwedza industry produced hoes, knives, axes, adzes, spear-heads, razors, .'chiefs' badges, hooks, needles, arm and leg bands, bullets, and a great Ispeciality was the manufacture of the musical instrument, the mbira.Production was followed by marketing over a radius of 100 kilometres, thesalesmen returning with cattle in exchange for hoes. Both cattle and hoes (were used as rovora or lobola (bride-wealth).Evidence of any transition from the Zambian induced-draught furnaceto the Rhodesian forced-draught furnace, or vice versa, is slender. Chilundusmelters at Kalobo in Zambia worked ore deposits until 1948 when theore pits were flooded by a rising water table. From a brief description anda photograph it appears that the smelters had used a large beehive-typefurnace with drum-bellows, the furnace bearing feminine symbolism, breasts,etc.30 The Ushi, also of Zambia, were prepared to change from induced- "draught furnaces to forced-draught types but abandoned the measure whenthe flaying of goats was banned, although the importation of cast steel hoes Iappears to have been the deciding factor in abandoning the industry. Butit is clear that the induced draught furnace not only avoided the ban on 'goat flaying but required less labour during smelting.31 The Kaonde of |Zambia in the early 1920s were shown the superiority of bag-bellowsover drum-bellows but they refused to adopt the new device for fearof arousing the anger of their ancestral spirits.32 The shaft furnace wasthe typical smelting furnace north of the Zambezi, examples being foundnot only in Zambia but in Malawi, Zaire and Tanzania, and further northfrom the Horn of Africa to the Niger delta.33 It is believed that the forced-draught 'beehive' furnace was introduced to Rhodesia by a fairly recentimmigration of Njanja from Mocambique in the eighteenth century.34The accounts of strange peoples supplied by explorers, traders andmissionaries, upon which nineteenth century ethnology was founded, havebeen criticized for being too descriptive and insufficently interpretative. Thesame criticism might be levelled at what, so far, I have written, and there-fore I shall conclude this very necessary description and proceed to a moreinterpretive phase in discussing the ritualistic and scientific aspects of thissubject.THE RITUALISTIC ASPECTJust as in the pre-Enlightenment, pre-seventeenth century Europe, thealchemists' experiments depended much on the mental condition of thepractitioner, so the ritualistic aspect of the smelter's work, the blacksmith's30 J. Housden and M. Armor, 'Indigenous iron smelters at Kalobo', NorthernRhodesia Journal (1959-61), IV, 135-8.31 Chaplin, 'Notes on traditional smelting', 58.sz F. H. Melland, In Witch-Bound Africa (London, Seeley Service, 1823), 137.33 Pearce, The Appearance of Iron, 16.34 MacKenzie, 'A pre-colonial industry', 203, 211.F. W. J. McCOSH 163work and the smelting process appeared one and the same subject, but it issimpler to deal separately with these three topics.Generally, the smelter was much respected and often feared becausesometimes he was not only a smelter but also a chief or senior headmanby virtue of a large family following owing to his ability to maintain numer-ous wives, acquired by hoes, the fruits of his skill.35 There were exceptionssuch as the Masai of Kenya where God offered Man a herdsman's staff,a bow and a smith's hammer. The poor smith was left with the hammeras the other gifts were claimed by the ancestors of pastoral and huntingtribes. Or, their neighbours, the Suk, who pity the smiths among them becauseGod gave them only brains but no sheep. Yet the Chagga in the same areabelieve that marriage to a smith's daughter by an outsider brought him toan early grave, such was the dangerous potency of smith's blood.36 Smithsand smelters were synonymous in these cases and therefore they have beenmentioned in this section devoted to smelters.Not all smelters were diviners or n'anga; often the Rhodesian smelterhad his own special n'anga who was called upon to guarantee the successof the smelting operation. The n'anga would speak on the work to beaccomplished that day and comment again at the end of the day. Sometimes'medicines' would be administered orally to the smelters and a goat sacrificedto the Great Earth Mother, Zimai-revhu, or Zimai-remhangura, the GreatIron Mother, who must not be offended. Where this deity stood in thehierarchy of spirits is not clear. The goat, consumed by the smelters, ensureda successful fire in the furnace.37 The Ila of Zambia obtained the sameresult by what Frazer would have termed homeopathic magic. Their furnace'medicine' consisted of a piece of hippopotamus hide and some guinea-fowlfeathers because a successful fire simulates the cries of hippo and guinea-fowl.38Reading through the literature on iron-working one is surprised by therich mythology of smithing compared with that of smelting. The confusionin some reports between the smith and the smelter may partially accountfor this discrepancy but I may suggest that the social status of each accountsfor the difference. The smelter was a skilled worker as well as a chief andlittle or no additional prestige was needed. But there were smiths in everyvillage and they were only associated with the metal itself, whereas thesmelter was concerned with the actual birth of the metal inside its mother,the furnace. The smith perhaps needed a mythological boost.as G. Dieterlen, 'A contribution to the study of blacksmiths in West Africa' in P.Alexandre (ed.), French Perspectives in African Studies (London, Oxford Univ.Press, 1973), 40; MacKenzie, 'A pre-calonial industry', 207.ss Cline, Mining and Metallurgy, 114, 115.37 Mambaiye, 'Kucherwa Nekupfurwa Kwesimbi Kare'.ss E. W. Smith and A. M. Dale, The Ila-Speaking Peoples of Northern Rhodesia(London, Macmillan, 2 vols, 1920), I, 210.164 TRADITIONAL IRON-WORKING IN CENTRAL AFRICAWhereas the smelter, except for exhibition purposes, has all but dis-appeared from the African scene, the blacksmith has survived because ofhis ability to effect repairs and to forge implements from scrap iron. Teachersin African schools will recollect woodcarving classes where each pupilbrought his own locally made adze> forged from a leaf of a discarded carspring. It was among the Dogon of Mali that the prestige of the smith foundits zenith which can be told briefly as follows. The Creator of the world,Amma by name, 'found' a 'womb' to receive his sperm from which issuedtwo sets of 'fraternal' or two-egg twins, Igo and his sister Yasa, and Nommowith his brother who is never named is referred to as the Blacksmith.Ogo committed incest with his 'mother' for which he was transformed intoa fox, representing all that is evil, including drought, sterility, disorder,impurity and death, whilst Nommo stood for fertility, order, purity and life.Amma took the heart-blood of Nommo and threw it to Earth where it landedupside down as an anvil. The anvil miraculously rose again and embeddeditself in the ground the right side up. Today, whenever a Dogon 'plants'his anvil in a shelter he is acting out this part of the legend. Nommo hadalready been castrated and his penis and testicles were thrown to earth toact as blast-pipe and bellows. But the Blacksmith had yet to learn the artof obtaining iron from ores; therefore Amma threw down the blood ofNommo's spleen which changed to a lump of iron, recognizable as a meteo-rite. It was now Yasa's turn to be useful when an arm was amputated toprovide a hammer and the hand became a pair of pincers.39 No similar legendappears to exist in central Africa, but then the Dogon are a remarkablepeople with an exceptional cosmogony. Anxiety for the success of black-smithing, with its consequences in hoe manufacture and therefore in foodproduction, was a feature among the Ondulu of south-east Angola. Whenthe large blacksmith's hammer, or sledge hammer, was forged it was tiedto the back of a young girl with a cloth, as a baby is tied to its mother. Itwas carried thus to the waiting villagers who welcomed it with appropriatesongs. The hammer received the name of the Chief because it was the pro-vider of food, for without it no hoes could be forged and therefore no cropscould be cultivated.40 Yet on the whole, the blacksmith exercised asecular function in the community; preternatural forces did not go beyondthe smelting process.If there are two general characteristics of smelting ritual throughoutAfrica they are the deep abhorrence of the presence of menstruating women,and secondly, the injunction to the smelting crew to abstain from sexual inter-course during the period of smelting. However, it appears that when increasedproduction was required for export, as in the case of the Njanja of Hwedza,39 Dieterlen, 'A contribution to the study of blacksmiths', 41.4.° V\^f- Read> 'Iron smelting and native blacksmithing in Ondulu country south-east Angola, Journal of the African Society (1902-3), II, 44-9.F. W. J. McCOSH 165there was a relaxation of taboos. Menstrual blood may be only a plasmacontaining various living corpuscles transporting food, oxygen and wastematter, but ritually it is in that fearsome category between the living andthe non-living. Yet it is associated with potential reproduction and thereforeit can be regarded as sacred as well as unclean and contagious, giving riseto a, prohibition which is a taboo in the fullest meaning of the term. At thefurnace a menstruating woman is passing blood, analogous to a runningslag which contains most of the iron from the ore so that a very poor bloomis left in the furnace, resulting in a serious waste of ore, charcoal, firewoodand time.41 The taboo on menstruating women also applies to the cattle kraalwhere it is feared that their presence will encourage the production of still-born or prematurely born calves. But the cattle kraal is family propertyover which there is a greater measure of control over visitors than in thecase of the smelting furnace which, although at some distance from thevillage, can be visited by all and sundry, by known and unknown women,hence the use of 'medicines'. In the pursuit of effective medicines, use wasmade of the strongly toxic alkaloid content of certain Amaryllidaceae bulbsor durura belonging to several plants referred to as mushandwe which flowerin vleis in the rainy season. I would suggest that the mushandwe commonlyused was Boophane disticha or Sore-eye because the large fan-like leaves areclearly seen in winter when smelting took place. A more positive approachto 'medicine' was adopted by the Ila of Zambia who claimed that theirmedicines transformed the ore into iron and that without them the reactionwould not proceed.42 The function of the 'medicines' will be further dis-cussed under scientific aspects.43The magico-ritualistic properties of the human foetus and placentawere recognized by the Tonga of Zambia who cast a piece of the afterbirthinto the furnace to improve smelting; and also by the Chewa of Malawiwho placed 'medicine' inside a stripped mealie cob which was then thrownat a pregnant woman, resulting in a miscarriage following the shock. A'doctor' retrieved the foetus from a refuse heap, mixed it with 'medicine'and then burnt it in a hole in the ground over which the furnace was con-structed. Some of the hoes thus produced were given to the woman who, soit is said, never knew the reason for the gift.4* These two examples mayrefer to an analogy between the embryonic primal substance of the worldand the ore used in the earliest metallurgy. Babylonian texts in the library*i Kay and Wright, 'Aspects of the Ushi iron industry', 36; the prohibition ofmenstruating women applies throughout Africa, not only to Zambia.« Smith and Dale, The Ila-Speaking Peoples, I, 203.as The emphasis on female taboos appears to have varied in place and time.Ultimately, the labour-intensive industry at Hwedza included the smelters' wives whocooked for their husbands at the furnace site and 'some clever women tried thebellows and became good at blowing them', private communication from Dr J. M.MacKenzie.** Cline, Mining and Metallurgy, 119.166 TRADITIONAL IRON-WORKING IN CENTRAL AFRICAof Assurbanipal, king of Assyria in the seventh century B.C. refer to themagical properties of a human foetus added to the smelting furnace45 butthe paucity of such examples in African metallurgy behoves us to adopt asceptical view towards any attempt to correlate African custom with thatof Babylonia, China or even medieval Europe where the analogy betweenthe human foetus and metallic ores was derived from the symbolism of themacrocosm and the microcosm by which man is formed after the image ofthe cosmos, a hypothesis which does not appear to exist in African society.THE SCIENTIFIC ASPECT Š PRACTICALAND THEORETICAL CONSIDERATIONSIt had been assumed that the product of African iron smelting was abloom of wrought iron mixed with siliceous slag, unused ore and charcoal,the impurities being removed by hammering them out of the plastic bloom.Wrought iron was the usual product of British furnaces before their conver-sion to blast furnaces at the beginning of the sixteenth century when theirincreased temperature enabled cast iron to be produced. Wrought iron issometimes described as pure iron although it usually contains carbon upto 0.15 per cent which gives it the properties of mild steel. However, bypolishing and acid-etching an iron surface from a traditional furnace, andthen obtaining a microphotograph of the surface, it is found to be a heterogen-ous product containing not only wrought iron but the constituents of steel,i.e. pearlite and ferrite. Usually the microphotograph reveals several gradesof steel together with iron and slag, proving that the iron had not beencast. Traces of cast steel are seen, suggesting that very high but localizedtemperatures were possible.46 This was confirmed during an investigationof old shaft furnaces in Britain when a specially constructed furnace fittedwith temperature measuring thermo-couples at various heights recorded ageneral temperature of l,200°C with a maximum temperature of l,600°Cat the base of the furnace, rendering possible the formation of a smallquantity of cast iron to be absorbed by the bloom. This work is mentionedbecause although not entirely relevant to African practice Š the furnacewas constructed of firebrick blocks and was a shaft furnace supplied withforced draught Š it suggests that some cast iron could have been formedin a traditional furnace.47 Yet what credence can be awarded to eye-witnessaccounts of molten iron pouring into moulds excavated in the ground? Sometestimonies48 are so extremely doubtful that mvura yemahwe (water from« M. L. von Franz, 'The idea of the macro- and microcosmos in the light ofJungian psychology', Ambix (1965), XIII, i, 22."s Stanley, 'Some products of native iron smelting', 131-2.47 R. F. Tylecote et. al., 'The mechanism of the bloomery process in shaft furnaces'Journal of the Iron and Steel Institute (1971), GCIX, 342-63, esp. 344.48 TTn;v. of Rhode=:R. DPD. of History. Oral Interviews Collection J. M MacKenzieJM 1/2, Masweru of Chibi, 26 Feb. 1974.F. W. J. McCOSH .167the stones) may refer to molten slag. But there are statements in which theremoval of the slag is followed by the release of the molten metal into amould, although the informants were elders recounting the recollections ofjuvenile spectators.49 It is doubtful whether this question will ever beresolved; most of the evidence suggests that a bloom of impure wroughtiron was produced, and latter-day demonstrations of smelting confirm thisconclusion.49As is generally known, a flux is a substance which combines with thesiliceous impurity, or gangue, in an ore to form a silicate which, due to itscomparatively low melting point, is free-flowing at the temperature of thefurnace so that it can be separated from the bloom as a slag. No flux wasemployed in pre-Roman Britain with the result that about half the iron waslost in the slag because it combined with the high proportion of silica in theore. Lime was first added as a flux in the first century A.D. and continuedto be used up to the sixteenth century when it appears to have been graduallyreplaced by limestone, first mentioned as a flux in the mid-seventeenthcentury.50 Usually a flux was unnecessary in African smelting where the lowsilica content of the upgraded ore combined with a little of the high ironcontent to form a ferrous oxide content which combined with the silicaimpurity in the ore.5' Burnt bone, rich in calcium carbonate, has been foundnear a smelting site at Inyanga,5* but these are isolated examples whencompared with the general use of bones by the Venda of the Transvaal.Human bones were especially sought after, and there was much consternationamong the smelters at the 1936 Empire Exhibition in Johannesburg whenit was rumoured that their bones were wanted by the Iron and Steel Cor-poration (Iscor).53The removing, or tapping, of molten slag was generally neglected asthere was so little slag which was usually found on the charcoal bed of thefurnace when the bloom was removed. However, there is evidence of slag-tapping from shaft furnaces; the Lungu of Zambia fired their furnaces for26 hours after which molten slag was released through a hole in the furnaceat ground level, but it was not until a further 18 hours had elapsed thatthe bloom was removed and refined in a small open furnace, probably a pitfurnace.5* Again, in the Ushi iron industry, also in Zambia, the slag flowedfrom some of the tuyeres but precautions were taken to ensure that theŁ49 Undoubtedly, smelting expertise varied from place to place and the resultwould depend on a combination of factors such as furnace construction, furnacecharging, the grade of ore, personal smelting skill and the quantity of materials used.50 H. R. Schubert, History of the British Iron and Steel Industry (London,Routledge and Kegan Paul, 1957), 44, 161.51 Hatton 'Notes on Makalanga iron smelting'.=z Summers, Inyanga, 120-1.53 U.S. Kusel, Primitive Iron Smelting in the Transvaal (Pretoria, The NationalCultural History and Open-Air Museum, Studie No. 3, 1974), 5.s* Chaplin, 'Notes on traditional smelting', 54.168 TRADITIONAL IRON-WORKING IN CENTRAL AFRICAtuyeres were not blocked55 as happened at the 1972 Njanja demonstrationat Enkeldoorn, Rhodesia, when the blow-pipe tuyere was blocked by slagand in the attempt to unblock the pipe, by pushing with iron rods, the tuyerewas broken and the smelting operation abandoned.56 It appears, therefore,that the function of a flux was understood but in many cases its use wasfound to be unnecessary.To seek a traditional explanation of the smelting reaction in terms ofWestern science would prove a futile exercise, whether in terms of oxygenor the earlier concept of phlogiston which combustible materials weresupposed to contain. Perhaps the only parallel would be the use of'medicines' in the African furnace which could be compared with the un-necessary constituents of many alchemical recipes. Nor can we oversimplifythe reaction as one between ironstone and charcoal, yielding iron and slag,providing there is no menstruating woman present. Menstruation, a negativefactor in this reaction, could not be classed with such positive variables astemperature, pressure and mass which influence a chemical change. Suchethnocentric considerations exclude the possibility of a traditional inter-pretation in terms of a model. The concept of a model, perhaps too obviousto be noticed, provides a more plausible interpretation in the form of thereproductive woman, a model familiar to the 'primitive'. Here we havenot only a mental image but one which is affirmed in the feminine symbolismof the furnace. In the absence of oral or written evidence for a theory, themodel explained smelting in terms of reproduction, and identified the bloomwith the new born child which, when later forged, was carried, baby fashion,on the back of a young girl Š such was the custom in south-east Angola.We can conjecture that the ironstone corresponded to the foetus Š was notthe actual foetus added to the ironstone by the Tonga of Zambia and theChewa of Malawi? In this model of the theory the charcoal at the baseof the furnace might represent semen, as this is the most effective zone inreducing the ironstone to iron. Here, we are in grievous peril of confusingthe model with the theory when its function is merely explanatory, andsome aspects of the model may be completely irrelevant as in the exampleof the kinetic theory of gases where the colour of the billiard balls, represent-ing gaseous molecules, is of no consequence. It is possible that the slaganalogized faeces, as among the Ngoni of Malawi, but it might also simulatea miscarriage, due to sexual intercourse by the smelters who were thus com-mitting adultery towards the furnace. The model appears as a personalizedexplanation of the smelting theory, a type of explanation more acceptableto traditional society, whereas in Western society the impersonal elements,because they are understood, are preferable vehicles of explanation where=5 Kay and Wright, 'Aspects of the Ushi iron industry5, 34. a=s Prendergast, 'Research into the ferrous metallurgy of Rhodesian Iron Ass 'societies', 258. !F. W. J. McCOSH 169population mobility and instability render difficult a personalized theory,as well as a tendency to a more secular way of life.Such speculation cannot be left without referring to Robin Horton'scontroversial and stimulating study of African traditional thought andWestern science57 in which he examines the similarities and the differencesbetween tradtional ritual and scientific methodology. The differences are awarning that we cannot assume that traditional models are similar to thoseof Western science, but the similarities are sufficiently impressive to compela comparative study of ritual and the philosophy of science. Some would,perhaps, regard the differences as of greater moment than the similarities,such as the continued faith in a ritual which has failed to achieve its objectas compared with the scepticism accorded the repeated failure of a scientificexperiment. No attempts appear to have been made experimentally to assessthe effects of sexual taboos on the smelting reaction. Experimentation of alow order, without modern aids, could have shown that menstruation wasnot a factor in this reaction, but perhaps this would have destroyed part ofthe woof and warp of their society where, to cite a well-known line,'. . . inthis web of belief every strand depends on every other strand, and a Zandecannot get out of its meshes because it is the only world he knows . . .'5SCONCLUSIONMuch technical detail has been included in the first half of this paperfor which no apology is required on the grounds that without such in-formation a consideration and understanding of the significance of smelt-ing ritual and the relevance of the philosophy of science is rendered difficult.Although iron-working in Africa covered a period of 1,500 years there hasbeen no explanation of the smelting process in terms of African thought,nor any investigation of the possibility of such an explanation. Emphasishas been on the outward manifestations of African ritual especially withregard to sexual conduct and menstruation, without acknowledging theirfundamental relevance to a process expressed as a personalized theory.Recorded interviews with aged informants who had witnessed iron smeltingin their boyhood revealed no trace of causal action, nor did the smelterswho were re-engaged in Zambia during the Second World War contribute anexplanation. Perhaps they thought it completely unnecessary; the explanationwas there for all to see in the moulded shapes of their furnaces."T?" ^°^iOniK'^can traditional thought and Western science', Africa (1967),AVlx, 50-71, 155-87.=s E. E. Evans-Pritchard, Witchcraft, Oracles and Magic among the Azande(Oxford, Clarendon Press, 1937), 194-5.17O TRADITIONAL IRON-WORKING IN CENTRAL AFRICAFinally, I cannot but contrast the social attitude to traditional iron-working with the public reaction, if any, to steel production today; for societyin general, one was a sacrament, the other a means of livelihood. Thetraditionalist in me salutes that most discerning of philosophers, MirceaEliade, writing of metallurgy and alchemy: 'Modern man is incapable ofexperiencing the sacred in his dealings with matter; at most he can achievean aesthetic experience'.59Eliade, The Forge and the Crucible, 143.