ARCHAEOMETALLURGICAL STUDY OF THE

VILLAFÁFILA TREASURE

This study has been carried out in the laboratory of the Scientific Cabinet of the Museum of America (Madrid). The metal composition of the objects has been determined spectrometrically by X-ray fluorescence (dispersive energy), with the analyzer of the Institute for the Conservation and Restoration of Cultural Assets, provisionally operating in the aforementioned Museum. The technique is non-destructive.

METALS AND ALLOYS

The pieces that make up the Villafáfila Treasure can be grouped into two groups: a) Gold pieces. These are three crosses of the so-called "kicks", with a very simple lamellar structure. b) Bronze and brass pieces. They are a container and its lid, and a metal rod or handle. All the gold pieces present a ternary Au-Ag-Cu alloy with very similar characteristics. With regard to its grade in gold, we obtain the following figures:

The results of the quantitative analyzes are noted in the table at the end.

A first question that we ask ourselves is whether it is an intentional alloy or, on the contrary, it is native gold. And in the latter case, if it is possible to know something about its geographical origin. The works by Axel Hartmann on the different qualities of gold throughout Prehistory and Protohistory are a must. Unfortunately, there are no other systematic (nor isolated, practically) studies on medieval metalwork.

The presence of small concentrations of tin in the Villafáfila crossings point to alluvial gold and not mine gold. Tin usually appears as a trace element in gold due to two very different causes: in alluvial gold, due to contamination with tin oxide (cassiterite) present in riverine sands; in gold alloyed with copper, as a natural or artificial companion to the latter (II ARTMANN, 1971: 130). In the case at hand, the second cause can be dismissed, since the copper content of the alloys (around 2%) could not be responsible. If so, it would mean that the silversmith linked the gold with copper containing about 3% tin (poor bronze), which is really unusual. On the other hand, copper alloys are made with the double intention of saving noble metal and varying its mechanical qualities to work with it more safely (fine gold is excessively soft). The presence of 2% copper serves neither of these two effects in practice. Therefore we are inclined to think, in principle, that the gold of the Villafáfila crosses is alluvial. Regarding its origin, it fits quite well in the type N of Hartmann (1970), whose components oscillate in the following intervals:

Tin between 0.50 and 0.25 per 100.

Copper between 1.2 and 2.8 per 100.

Silver between 9 and 24 per 100.

Type N is widely distributed in Central Europe and Ireland. This would lead us to suppose that it could be an imported metal.

However, we cannot rule out the hypothesis that it is native peninsular gold to which silver has been linked with copper, the usual method of lowering the gold grade. In what follows we will try to show that such a hypothesis can be true.

Figure 1 shows in Cartesian coordinates the regions occupied by the N and L types of gold. The latter is characteristic of the Iberian Peninsula. The limits of the regions are the extreme values ​​found by Hartmann (1970). The points corresponding to the Villafáfila crosses are located in the N region or its vicinity. It is evident that these points would come to fall in the region L if certain amounts of silver and copper were subtracted from them.

Using the mean values ​​of the Hartmann intervals and of the Villafáfila alloys, we can establish a theoretical equation that links the transformation, determining the theoretical mean value of the 'Ag-Cu alloy that has been added to the native peninsular gold to convert it into type N " extra-peninsular". The points to consider would be:

The line joining the points P, (13.25, 0.29) and P, (18.85, 1.76) has a slope m = 0.262. With this value we can already establish the line that passes through them, whose equation is:

y-0.262x4-3.18=0

from which we can find the average law of Ag-Cu that has been used, solving the system:

y-0.262x+3,I 8=0 x+y=100

The values ​​x=81.76 per 100 of Ag are obtained; y=18.24 per 100 of Cu, which are frequent in silverware of all times. We are not aware of analyzes of Visigothic silver objects, except for a trilaminar fibula fragment containing 83% silver and 12% copper (ROVIRA and SANZ 1985: 236). Low silver pieces are quite common. Many of them are not correctly identified until they are subjected to a suitable cleaning treatment, since the corrosion products of copper (which hardly bonds with silver) produce a "copper" or "bronze" patina that, often confused on a cursory observation.

With the calculated theoretical values ​​and a discreet margin of tolerance extracted from the differences between real and theoretical values, all the Villafáfila points come to fall within type I., de-gold. Consequently, the Villafáfila crosses may have been made with native peninsular gold (perhaps from León or Asturian placers), alloyed with silver and copper.

Two of the pieces (105.1 and 105.3) feature surprisingly similar alloys for the crosses and suspension elements. It is safe to say that both pieces were made from the same metal castings and, therefore, at the same time or within a reasonably short period. It also seems clear that the craftsman selected one type of metal for the cross plate and a different one for the chains and suspension elements. This choice has little to do, in our opinion, with technical aspects. Although gold is softer the purer, in this case the difference in mechanical qualities would not be decisive at all. Perhaps it is simply due to the fact that he had already drawn wire in the workshop ready to be used in each case with slight modifications, and with it he built the suspension elements. For the crosses he had to prepare an "ad hoc" metal acorn.

On the other hand, the third cross (105.2) was made with different alloys than the previous ones. Although the grade in gold is similar in the cross and the suspension element, the copper content differs in both and indicates different alloys.

The second group of pieces is made with copper-based alloys. Vessel 105.4 is made of Cu-Sn-Pb ternary bronze, frequent in Visigoth metalwork (ROVIRA and SANZ. 1985: 283; ROVIRA 1987: 382). The difference in composition in two analytical shots taken at the edge and bottom of the piece is due to the lack of homogeneity of the bronze with a high lead content.

The 105.5 cap shows a similar alloy to the previous one. This, together with certain formal coincidences, suggests that it is effectively the lid of container 105.4. The decreasing figures of lead in the analyzes are related to the order in which the smelter was pouring the molten metal from the crucible to the moulds, taking into account the aforementioned segregation of lead. The fact that arsenic was recorded in one analysis and not recorded in the other two may be due to the configuration of the measurement equipment, in which it is difficult to read the arsenic spectral line when lead is present in large quantities.

Finally we have the bar or handle 105.6, made of brass. The subject of late Roman and early medieval brasses was widely discussed on another occasion (ROVIRA and SANZ 1985: 239 et seq.; SANZ and ROVIRA, 1986: 88-89).

CONSTRUCTION TECHNIQUES

We have been able to carry out a metallographic study of one of the crosses, performing a microscopic metallography on the end of one of the arms. The metallic structure corresponds to a cold-rolled and annealed metal. To build the cross, the silversmith had to start from a metal acorn, rolling it with a hammer or with the help of rollers. After each severe drawing the gold hardens and becomes brittle. To restore its original mechanical qualities, it is necessary to heat it up to the recrystallization temperature, thanks to which recovery takes place and it can be worked again without risk of breakage. The presence of cracks in the metal, some of which can be seen with the naked eye, indicates that the metal's resistance limit was once exceeded during the rolling process. Or was it an oversight?

Once the sheet of the desired thickness was achieved, it was cut with scissors according to a previous design drawn on the metal surface. Traces of shearing when cutting the metal, as well as irregularities of the cut, can be observed in the interior angles.

The suspension bags are fastened by cold welding by hammering the metal. Gold is easily soldered in this way.

In summary, the technological aspects of the Villafáfila crosses are very simple. Its execution is far from being the work of a careful craftsman. It seems to be a local workshop with a provincial air. The rest of the pieces of the Tesorillo de Villafáfila are cast. Container 105.4 has certain suspension elements on the rim that remind us of those of incense burners (ROVIRA 1986: 271 et seq.).

RESULTS TABLE

Spectrographic analysis results expressed in '1/4t by weight. The acronym ad stands for element not detected under analytical conditions. It may be present in amounts below the detection limit of the spectrometer (generally less than 0.01 percent except for silver and antimony).

The acronym tr means element present as traces (generally in amounts less than 0.01% except for silver and antimony).

BIBLIOGRAPHY

HARTMANN, Alex:

1970 Prdhistoriche Goldfunde aus Europa. Spektralanalvtische Untersuchugen und iteren Auswertung, Studien zu den Antingen der Metallurgie, Band 3, Gebr. Mann, Verlag, Berlin. 1971 "Analyses of some cigar objects from the North of Portugal", in Revista Guimaraes, Vol. LUX', pp. 129-133.

ROVIRA, Salvador:

1986 "Technological study of the Palestinian censer", in Erytheia. 7, 2.pp. 269-277. 1987 "Metallurgical study of five Visigoth belt clasps preserved in the Lázaro Galdiano Museum (Madrid), in Spanish Medival Archeology. II Congress, Volume II, Ministry of Culture, Community of Madrid, pp. 379-384.

ROVIRA, S. and SANZ, M.

1985 "Metallurgical analysis of the materials from the El Carpio de Tajo Necropolis (Toledo)", in G. RIPOLL, La Necrópolis Visigoda de El Carpio de Tajo (Medi)), EAE, 142, Ministerio de Cultura, Madrid, pp. 227-254. SANZ, M. and ROVIRA, S.:

1986 "L'Analyse spectrographique et métallographique des ohjets Visigothiques", in Dossiers Ilistoire et Archéologie, No. 108, Les 1Visigoths, pp. 84-89.

Archaeometallurgical study of the treasure of Villafáfila

Authors: 

Ignacio Montero Ruiz ,  Susana Consuegra Rodríguez ,  Salvador Rovira Lloréns .

Numantia: Archeology in Castilla y León ,  ISSN  0213-7909,  No. 3, 1990 , pp. 209-216

Transcription and montage:

All text, photographs, transcription and montage, the rights belong to their authors, any type of use is prohibited without authorization.

All text and photography has been authorized for storage, treatment, work, transcription and assembly to José Luis Domínguez Martínez, its dissemination on villafafila.net, and any other means that is authorized.