4 Conclusions
It is concluded that the two Geistingen axes have been produced by using different raw materials. The use of three complementary analysis techniques has allowed an optimal determination of the average composition of the bronze used for these two axes. Axe AC20 consists of copper alloyed with 9 at% tin, while the bronze of BH76 is identified as copper with 4 at% antimony and 6 at% nickel.
The microstructures display the presence of matte particles and silver and lead-antimony particles. Based on these features, the melt temperature and the different composition of AC20 and BH76, production by the (non-extensive) re-melting of scrap (part of which was possibly made from imperfect smelted sulphidous ores) for both axes is suggested. The microstructure of both axes proved to be essential for the deduction of characteristics of the thermal cycle during the production process. The composition and morphology of interdendritic inclusions indicate that the material has been molten in the temperature range 1100-1150 °C. The secondary dendrite arm spacing, combined with the composition of the alloy, is used to estimate the cooling rate during solidification. The average cooling rate of both axes is around 30°C/s, which implies that both axes have been quenched with water after casting into a (bronze) bi-valve mould.
Combining the compositional, metallurgical and archaeological considerations it is clear that Geistingen axes were not intended to be used as a tool or a weapon. Five aspects render a functional use of the Geistingen axes improbable:
- their thin walls (Butler 1973, 339-340; Kibbert 1984, 166),
- their low weight (Butler & Steegstra 2002/03, 304; Kibbert 1984, 166),
- the presence of the embrittling element antimony in the alloy (this study),
- they rarely show traces of use (Fontijn 2003, 325) and working (this study), in spite of their fine external finish and sharp cutting edge (Butler & Steegstra 2002/03, 304),
- some show casting flaws that would have impeded hafting altogether (Butler & Steegstra 2002/03, 309).
This leaves ingots ('axe money') or an unspecified votive or ritual function as options. It is possible that the conversion of scrap stock into local types that by composition, knowledge of origin or by visual clues could be identified as non-local, was facilitated or cosmologically rendered acceptable through the deposition of parts of the scrap as hoards, as cogently argued by Fontijn (2008). Consequently, even larger hoards as the eponymous Geistingen hoard may merely represent parts of much larger original quantities. This tallies with the observation that for the Geistingen axes were probably produced using a reusable mould or pattern. In short, sizeable quantities of Geistingen axes were presumably produced, but never with the intention to ever fell a tree with them.