Lying within sight of the South Australian mainland, the 4000km2 land mass of Kangaroo Island presents a continuing archaeological mystery. In the mythology of southern coastal Aborigines, it was Karta island of the dead (Tindale and Maegraith 1931; Berndt 1940) a spirit place beyond the reach of living people. However, reports of the discovery of stone artefacts on the island by Howchin in 1903 and Tindale and Maegraith in 1931 demonstrated that Aboriginal people once had occupied the island, beyond the recollection of oral history. The suite of stone implements recorded by Tindale was distinctive. It consisted primarily of large, unifaciallyflaked cobble choppers, unidirectional block cores ('horsehoofs' and 'karta'), steep-edged, thick flake tools, pitted hammerstones, occasional massive 'waisted axes' and very few flakes. Most artefacts were made from the island's abundant quartzites and metasandstones, with occasional use of quartz. This assemblage was unlike any stage of the industrial sequence identified at Devon Downs rocksheiter (Hale and Tindale 1930). Tindale named the Kangaroo Island industry the 'Kartan culture'. Scattered finds of similar implements had been made in the Flinders Ranges north of Adelaide, and along the Fleurieu Peninsula to the south. This suggested to Tindale that the Kartan might date to the last ice age, when Kangaroo Island was connected to the mainland through lowered sea levels. Tindale also recognised similar implements in collections of artefacts from Tasmania. The location of Kartan artefacts on relict, stranded shorelines around Murray Lagoon reinforced the idea of Pleistocene antiquity, as did the similarity of the cobble artefacts to examples from Sumatra and Malaya, which were thought to date to the Upper Pleistocene. Putting all of these clues together, Tindale (1937:59) suggested that the Kartan artefacts represented the toolkit of the earliest immigrants to Australia from southeast Asia _ the people who had reached Tasmania and Kangaroo Island while these places were accessible by glacial land bridge. Consequently, Tindale placed the Kartan as the earliest phase of the industrial succession which he described for Australian prehistory (Tindale 1957, 1968, 1981). Fieldwork to locate Kartan sites continued from the 1940's through to the 1960's by Cooper (1943, 1960, 1966, 1968a, 1968b), Bauer (1959, 1970) and others. More than 100 sites were recorded on the island and adjacent mainland peninsulas. However,