Matthew Svensson HBSc thesis abstract

A collection of fourteen samples of metal made from iron artifacts recovered from Thule Inuit and Dorset culture sites in the Canadian Arctic Archipelago was received from the Canadian Conservation Institute, Department of Canadian Heritage in order to examine the origin of the materials.  In Greenland, three sources of iron in artifacts have been determined:  terrestrial native iron, manmade iron of European origin and meteoritic iron.  Observation of polished surfaces made from the samples by reflected light microscopy revealed the presence of the Widmanstätten structure deformed in all cases by cold working, clearly indicating the meteoritic origin of all of the artifacts.  The Ni-distribution and concentration in the kamacite phase was in the classification of the meteorite thus providing strong evidence pertaining to its origin.  It is reasonable to hypothesize that the Cape York meteorite in southern Greenland is the source of iron for these artifacts. A comparison was done of the Ni-content and the bandwidth of the kamacite lamellae in the artifacts with that of the Cape York meteorite in order to precisely test this hypothesis.  1.0mm-2.0mm sized samples were taken for use in SEM-EDS analysis in order to preserve as much of the artifacts as possible.  A step-scan was performed roughly perpendicular to the kamacite taenite interface to acquire a representative sample of the artifact’s whole composition with the use of the Filomena meteorite as a standard. The deformation in the kamacite lamellae caused by the cold working of the artifacts and the small sample size prevent direct observation of the full bandwidth of the kamacite lamellae, so the bandwidth of the kamacite lamellae was calculated through the use of Goldstein’s (1965) plot of average half-width of kamacite vs the average kamacite Ni-content. The plots in this study are nearly flat, reflecting the approach to equilibrium in diffusion of Ni from kamacite lamellae.  The kamacite bandwidth as measured in SEM images and by determination from Goldstein’s (1965) figure 9 yield a range of widths of 0.042 to 0.0196mm .  These bandwidths are significantly smaller than Buchwald (1975) reported for Cape York at 1.20±0.2mm.  The bandwidths for masses other than Savik I have not been reported, but it is reasonable to expect that they would be significantly smaller and similar to the results of this study.