IN G EN I 0 R VI DEN S KA BEL I G E S K R I FT ER Nr. A 46 THE GILDING PROCESS AND THE METALLURGY OF COPPER AND LEAD AMONG THE PRE-COLUMBIAN INDIANS BY PAUL BERGS0E COPENHAGEN TRANSLATED FROM DANISH BY C. F. REYNOLDS DANMARKS NATURVIDENSKABELIGE SAMFUND I KOMMISSION HOS G. E. C. GAD . VIMMELSKAFTET 32 K0BENHAV N 1938 TO MY SON AND FRIEND SVEND BERGS0E GRATEFULLY DEDICATED COPYRIGHT: DANMARKS NATURVIDENSKABELIGE SAMFUND COPENHAGEN AKC!OENSTRYKKERIET POUTlKENS HUS In an earlier treatise from 1937, "The Metallurgy and Technology of Goln and Platinum among the Pre-columbian Indians" (1), I gave an account of the investigations made by me in the case of certain gold objects found in La Tolita, Ecuador. These finds were supplemented by others from the neighbourhood of the town of Atacames, a few kilometres farther south and situated likewise on the coast. The finds proved to be of importance. The extremely large number of the objects - over a thousand - allowed of very exhaustive metallurgical examinations; the fact that these objects were of the minute character peculiar to the spot placed no hindrance upon the work. What especially was of unusual importence for the task was the incredibly large number of half-finished articles found, from which it was possible to reconstruct the whole series of the metallurgical processes E:mployed, from the first melting of the native gold right down to its final melting into ingots and subsequent drawing into wire and conversion into sheets. By sheer luck so much platinum work in its initial stages was .found amongst the material that it was also feasible to follow the metallurgical methods employed with platinum from its very beginning to the fashioning of the small platinum plated objects which occupy a unique position in the whole of ancient metallurgy, that of Europe-Asia not excepted. The discovery of such unique skill in the treatment of metals must, in my opinion, lead to a totally new conception of the Pre-columbian Indians' abilities in this direction, at any rate as regards the dwellers along the Esmaraldas coast. Of earlier publications regarding metal finds from Ecuador may be mentioned first and foremost the works of Saville (3), (4) and Farabee (5), (4), while the discovery of the occurence of plat~num must properly be ascribed to Wolf (2). Orchard (12) has described the small gold beads. Finally. it should not be forgotten that Bollaert (19) already in 1860 in his work, "Antiquarian Ethnologique and Other Researches", mentions La Tolita and the remarkable minute gold objects, but says nothing about platinum. On page 83 he writes that at La Tolita the rain washed out six pounds weight of gold objects annually. I did not receive that portion of the material which did not consist of gold and platinum until the work on the finds of gold and platinum was 6 already in the press and could therefore not be extended to embrace a description of these objects, all the more as I realized that a fresh examination would take longer than I had originally supposed. On making the announce ment that I had encountered lead articles which were doubtless pre-columbian and had discovered how the gilding of copper was performed, I was invited by America to issue a new publication. On January 2nd, 1938, at a meeting of "The Danish Academy for Technical Science" in Copenhagen I made a short preliminary statement regarding this gilding process, which also appeared in "Nature" (20) and also annouced that the work would be published in time for the Anthropological Congress in Copenhagen to be held in the sum mer of 1938. DESCRIPTION OF THE NEW MATERIAL FROM LA TOLlT A Copper Objects: 1 very large axe 6 X 19·5 cm weight 1250 gram. (Plate II Fig. 9) 8 smaller axes and adzes, No. 1. 178 gram, No. 2. 172 gram, No. 3. 281 gram, No. 4. 274 gram, No. 5. 123 gram, No. 6. 107 gram, No. 7. 61 gram, No. 8. 59 gram. (Plate II Fig. 1-8). 1 adze with hollow grip, 208 gram (Plate II. Fig. 10) 6 fragments, possibly axes, from 223 gram, to 20 gram 1 copper ball 175 gram (Fig. 11) 1 with chisel marks, 62 gram (Fig. 12) 1 partly hammered,perhaps commencement of a bell, 15 gram (Fig.13) 1 large bell, 6 cm, 68 gram (Plate III Fig. 14) 18 smaller bells from 12 gram to 0·4 gram (Fig. 15) 1 fragment of hair tweezers, 3 gram 32 sewing needles from 4 gram to 0-4 gram (Fig. 16) 4 chisels or broaches, 26 gram, 8 gram, 5 gram, and 0·5 gram (Fig. 17) 9 picklock shaped objects, 2 gram to 0·9 gram (Plate IV Fig. 18» 1 arrow head, sugarbag-shaped, 0·4 gram (Fig. 19) 9 fragments, filigree work (Fig. 20) Various objects composed of fragments of wire and sheet-work, the nature or use of which it is impossible to conjecture. 7 Gilded Copper Objects 6 small knives (tumis) 0·9 gram to 0·2 gram (Plate IV 21) anow head(?) 1·4 gram· (Fig, 22) 32 fish-hooks 3 gram to 0·05 gram (Fig. 23) 17 nose-rings 5 gram to 0·3 gram (Fig. 24) 5 small spoons as described in (1) Fig· 16, 0·5 gram to 0·2 gram (Fig. 25) 10 nails (possibly) 1·1 gram to 0·3 gram. 50-60 various objects, often resembling the gold objects. (Fig. 26) 1 bead as in (1) Fig. 42, 0.3 gram 5 pendants as in (1) Fig. 37, 1·5 gram to 0·3 gram (Fig. 27) 1 object in the shape of an anchor (11 gram). (Fig. 28) Lead Objects 1 round rod 6·8 X 04 cm, 7·8 gram. (Plate V Fig. 29) 1 " 2 X 0·2 cm, 0·3 gram (Fig. 30) 2 nose-rings 1·3 and 0·7 gram (Fig. 31) , ; Non-metallic Objects 2 Mirrors of Obsidian 189 gram and 90 gram (Plate V Fig. 32) 20 objects of polished pyrites with holes, 35 to 2 gram (Plate III Fig. 33) 3 pyrites beads 2 - 0·2 - 0·1 gra~ 4 stone beads of other kinds. 1 whetstone (?) Regarded en masse there is something remarkable about this find; the forms are at once recognizable from the gold objects and are peculiar to the district. But it is immediately noticeable that the copper objects are larger, the reason for economy not being so urgent in the case of copper as it was with gold. Upon the objects mentioned as gilded only traces of gilding remain in places; on some, indeed, it has entirely disappeared. But, as will be explained later, it is quite easy to distinguish these copper cores from copper which has not been gilded. The first object mentioned, namely, the very large axe (Plate Il Fig. 9), is a magnificent piece, made with amazing accuracy - a modem 8 factory-made steel axe could not be more regular in shape. Both this axe and those of smaller dimensions are made of what even today we should call pure copper. The three copper balls (Fig. 11-13), one with the deep marks of a chisel, might with a little imagination be supposed to be bells in their initial stage, but nothing definite can be said on the point. The hells (Plate In Fig. 14-15) are all provided with two holes fOl.' suspension; this formjs. well k~own to be peculiar to Ecuador. Some of them are made of hammered gilt sheets, but the gilding was so thin that nothing of it is left except the black colour such pieces assume. The sewing needles correspond in all respects to the gold specimens; as a large number of them are met with - over 30 - while the gold needles are of rarer occurrence, we may assume that the Indians used copper needles for preference, only using the needles of gold alloy when for some reason they were in need of a specially hard needle. Analysis goes to indicate that the needles were not made of quite such pure copper as the axes; this may be owing to the use of scrap metal. The eight which were analysed gave 98·6 %, 99·9 %, 98·6 %, 97·6 %, 98·9 % 98·7 %, 98·4 %, 97·7 %, - all to within an error of ± 0·25. "Chisels or Broaches" (Fig. 17) are already upon the borderline of the definable; on only one of the pieces is it possible to demonstrate with certainty hammermarks on its blunt end. It appears to have been used as a punch for work on fairly soft material. As regards the remainder of the ungilded objects, their use is practically unassignable. There are so many specimens of some of them that they can hardly be fragments, but it is difficult to express any opinion as to their use. The gilded objects bear an amazing resemblance to similar articles made of gold. There are the same fish-hooks, (Fig. 23), nose-rings, (Fig. 24), eyelets, (Fig. 20), the small spoons, (Fig. 25), the use of which is so enigmatic, the well known pendants, (Fig. 27), filigree work, and beads. Some of the objects bear a striking resemblance to nails. Besides these objects there are, however, others of which no gold counterparts are found. Certain small knives (tumis), (Fig. 21), an arrow-head, (Fig. 19), a good number of fringes in the Greek style, and a few articles, the use of which it is impossible even to hazard a conjedure. A few specimens of gilded sewing needles have come to light, thanks to meticulous control, and for the sake of completeness it should be mentioned that there may have been more articles of this kind, but the thinness of the material has caused them to corrode completely and they have therefore not been found. OnE specimen is quite remarkable. It is made in the shape of an anchor of two pieces of wire 8·8 mm thick (Fig. 28). Superficially, the piece 9 exactly resembles bronze; it is, however, in reality gilded copper, but, unlike the other specimens, the gilding has not been aftertreated with the hammer. It has therefore been spared corrosion. There is not one single piece of bronze amongst the material, but on the· other hand four specimens of lead have been found, two of which are nose-rings. (Plate V). They will be mentioned later. Amongst the non metallic objects we notice two large convex mirrors of obsidian (Fig. 32), each 8 cm in diameter, the one made from a roughly hewn block, the other with a flat hewn underside. The polish has become rather dull. Neither specimen distinguishes itself in any way from these described by Nordenskiold (21). Besides these obsidi1l.n mirrors 20 in all have been found of polished pyrites provided with holes for suspension - if indeed they are really mirrors, for they are comparatively small, 3·5 to 2 cm (Plate III Fig. 33)· Also a number of stone beads, some of pyrites and some of other kinds of stone. All the material can be inspected in the Ethnological Section of Denmark's National Museum in Copenhagen, together with the gold and platinum objects described in the first treatise. OXIDA TION AND CORROSION All the copper articles are in an amazingly good state of preserva tion; the small pit-like concavities which indicate heavy corrosion are nowhere to be found. Most of the material is covered with cuprite (Cu0) and 2 malachite, but generally in quite a thin layer. Here and there we find thicker deposits in the layer of malachite, but there is nowhere the faintest sign of the formation of the well known blue compound azurite. Some of the fragments were studded with quite small, shimmering crystals which Professor B0ggild, of Copenhagen University, was kind enough to determine as magnetite (Fe 0.). 3 The preliminary chemical analyses of the first metal boring revealed a copper content of 99·4-99·8, a quite unusual degree of purity, and thus explained the reason for the slight amount of corrosion. It is well known that corrosion depends very greatly upon the purity of the metal. The nature of the impurities present also plays a part. If these impurities form 10 solid solutions and become an integral part of the copper crystallate, as in the case of nickel, for instance, they givE' no cause for corrosion; upon the .:lther hand, inclusions which become imbedded among the crystals of the copper, produce corrosion. A higher degree of corrosion was observable among the smaller copper objects. But they were not composed of such pure copper and were very possibly made by melting down scrap. As regards corrosion, the gilded objects revealed a quite amazing difference from these of copper. The gilding had peeled off in flakes, as a rule to such an extent that only a small quantity remained and, in the case of a few pieces, hardly any remained. A good deal of this flaked-off gilding was found among the gold objects described in my first article (1). Of such flakes, a few were so unbroken that, in spite of being pressed flat and out of shape, it was possible to see that they had been part of nose-rings· On the inner surface the remains of fine gold powder still adhered, this being derived from the transition layer between the gold and copper. But the vast majority of these gold flakes, however, were so small and so thin that it was impossible to come to any decision about them other than that they were the remains of gilding; no copper was found adhering to them. The copper nuclei themselves were heavily corroded, they were often pitted and their colour was quite different to that of the un gilded copper. Malachite was only discernible at isolated spots and the fragments were nearly black. In the great work describing the Cocle find (22) the heavy corrosion which the gilded objects had suffered is also mentioned. In this case the circumstances appear to be exactly similar. I analysed this dark layer which proved to be a mixture of cuprite, malachite and gold powder. The test in an easy one: the object is laid for one hour in dilute hydrochloric acid. This dissolves the malachite and cuprite. If the object is now taken out, it is found to be covered with a loosely adhering black powder which easily comes off onto the fingers and the marks of which it is almost impossible to remove. The black powder is gold, which is easily determined by brushing it off the object and submitting it to the chemical test for gold. It is now possible to explain the whole process of corrosion, which differs so greatly from the slight degree of corrosion taking place on objects of pure copper. The first thing that happens is that the gilding partly or entirely loosens itself from its base· It the gilding is continuous without pores it will provide a real protection for the underlying base metal. But if gilding has been performed so that pores are left, the attack will be delivered here and the result will be a loosening of the gold sheath, the products of oxidation taking up a greater space than the original metal. When the gold sheath has split off one might expect that thc heavy corrosion would cease,