|DAVID A. SCOTT

| |

|

| La Aguada style bronze cast plaque, 84mm x 51mm, dated to about 650 AD - 850 AD from Argentina or Bolivia. Cast in a copper-tin-antimony alloy. Probably cst by the lost-wax process.

| |Abstract | |: | This paper presents some metallurgical |and technical studies of ancient South American metal objects from Colombia, Peru and |Argentina.

|They are presented as fascinating representatives of disparate cultural traditions and different metalworking practices from The pre-Hispanic era. The first material which will be discussed is a collection of copper and copperrich tumbaga alloy fragments from Colombia, representing several different cultural areas, such as Nariño, Sinu, and Muisca. Later in the paper some very different, but technically important objects, in gold, platinum, and silver will be examined from the Recuay and Huari cultural areas of Peru, and finally a bronze plaque, possibly from the La Aguada culture, Argentina or Bolivia.

Resumen: Este artículo presenta algunos estudios metalúrgicos y técnicos de objetos procedentes de Colombia, Perú y Argentina. Estos artefactos son representantes fascinantes de tradiciones culturales dispares y de diferentes prácticas metalúrgicas de la época prehispánica. El primer grupo de objetos que se discuten incluye fragmentos de artefactos de cobre y de tumbaga de Colombia, procedentes de varias regiones como Nariño, Sinú y Muisca. Luego se analizan algunos objetos tecnológicamente importantes aunque muy diferentes, de oro, platino y plata de las áreas culturales peruanas Recuay y Huari. Finalmente se presenta el estudio de una placa de bronce de la cultura La Aguada de Argentina o Bolivia.

Ancient Colombia is well-known as contributing some of the most exquisite lost-wax castings to the corpus of art from the Ancient World (Plazas and Falchetti 1978; Bray 1978). Some of these objects are made in tumbaga alloys, alloys of copper and gold in varying proportions whose composition was directed by the choices of the metalsmiths regarding colour and symbolism, casting properties and the possible alteration of the surface colour by depletion gilding or fusion gilding (Scott 1982,1986a) to name only two possible approaches in the Colombian sphere to methods of surface treatment. Many of these castings require the deliberate alloying of copper and golg to create a whole range of colours and alloy types, from those with 90% - 80% copper content, which mar characterize much Nariño Piartal period metalwork, for example (Plazas 1979; Scott 1982), to those of the Sinu and Quimbaya region where many cast tumbaga alloys mar have only 10% - 30% copper.

The question naturally arises as to where the Indians procured the copper used to make these numerous tumbaga alloys, and whether the copper used was obtained in the native state, or smelted from oxide or sulphide ores of copper. The problem is particularly acute for ancient Colombia, for unlike Peru or Ecuador where there is plenty of evidence for the use of arsenical copper and tin bronze, these two alloys are largely absent from the pre-Hispanic Colombian cultures, and no smelting slags of copper or primitive bowl furnaces have yet been fauna in the ancient Colombian region. The problem is exacerbated by the fact that it can be very difficult, if not impossible to distinguish between copper which has been smelted from malachite, or copper melted from pieces of the native metal. Work on questions of this kind for Old World alloys began in earnest with the Sumerian copper research project in England, reported in Man, the Journal of the Royal Anthropological Institute, which began to illuminate the problems of differentiation with a series of emission spectrographic analyses of native copper and smelted products (Voce 1948). Whilst the work was useful in a chemical sense, it began to raise doubts about the possibility of relating particular objects to specific copper sources, or even being able to distinguish between native copper utilization and the smelting of copper from oxide or carbonate ores. This situation has changed during the 1990´s: in some cases it is now possible, with a combination of metallographic and compositional studies to differentiate between native copper and smelted copper, particularly if the native copper has been shaped into an object without being molten in a crucible. The recent work of Rapp et al (1990), Wayman et al (1985), Hancock (1991) and Maddin et al. (1980), has shown that some progress is possible on this difficult question that the earlier investigators tackled with an optimism that was ill-conceived. In part, these advances are due to the application of sophisticated modem methods of analyses such as thermal neutron activation analysis, and, in this study, inductively coupled plasma-mass spectrometry (ICP-MS). Pavlish et al. (1994), for example, reported on a detailed analytical study, attempting to distinguish between natural and processed copper-based metals within North America. They fauna that a clear separation could be made between copper materials used by the native North American Indians, before European contact and those imported after contact. The results showed that the North American copper could be several orders of magnitude purer in terms of key trace elements such as gold. Following from this premise, the antimony to gold ratios were fauna to be very diagnostic between the two groups of copper. There are greater difficulties than that with the Colombian copper objects examined here however:  we are not trying to distinguish between European smelted copper and native copper, but between indigenous smelted copper and native copper, which is much more problematic.

| |

Figure 1 : Colombia sample C2. X-ray fluorescence spectra for this Nariño bell reveals clear indications for the presence of arsenic. The alloy is otherwise relatively pure copper with a trace of silver.

Previous studies (Wayman 1985) have shown that some of the most important elements to distinguish native copper from smelted copper are: arsenic, nickel, selenium, antimony, gold and sometimes silver. The concentration of elements such as cobalt, zinc, tin, mercury and iron have not been found to be generally diagnostic. In the case of native copper, the levels of arsenic, nickel, selenium and antimony mar be in the low parts per million region or below detection limits, whereas smelted copper mar contain several hundred parts per million of arsenic, a few hundred ppm of nickel, 10-100 ppm of selenium and 10-400 ppm of antimony.

Fifteen samples from different archaeological regions of Colombia were studied during the investigation. These fragments had been selected by Clemencia Plazas, Director of the Museo del Oro, Bogota, as representative of the copper or copper-rich end of the tumbaga spectrum. The samples, with relevant information, are tabulated in Table l. A preliminary analysis of each piece was carried out non-destructively with x-ray fluorescence analysis in order to ascertain if the object was, in fact, primarily fabricated from copper, the abbreviated results of which are also given in Table l. The analysis with x-ray fluorescence spectrometry was carried out using a Kevex 0750A Spectrometer in air with a barium-strontium secondary target and accelerating voltage of 55KY, 3.3mA, with an acquisition time of 240 seconds.