Exciting new evidence shows that the human settlement of New Guinea has an antiquity equal to that of Australia, together with even more startling implications regarding the beginning of horticulture. The evidence comes from uplifted limestone sea-shore terraces on the Huon peninsula, north-east of Lae. The Huon terraces are one of the best sets of fossil Pleistocene coastlines anywhere in the world. The terraces rise like a giant flight of steps out of the sea; each ‘tread’ is an ancient coral reef now raised up high above modern sea-level. The Huon peninsula is bounded by offshore volcanoes and lies at the collision point between three of the earth’s plates. Massive earth movements, earthquakes and volcanic eruptions to which New Guinea is subject have raised the coral reef formed at the shoreline 120,000 years ago to 400 m above sea-level. The rate of uplift is 4 mm per year, or about 4 mevery thousand years. On one of the lower terraces, 80 m above sea-level, more than twenty-four weathered axes have been found. These ‘waisted axes’ are large, heavy, stone tools with a flaked cutting edge. A notch flaked out of each side edge gives them a ‘waist’, or an hour-glass shape. The notches are not simply to provide a grip on hand-held stone tools, since some of them are too wide apart for even the largest hand to stretch across. The notches were probably made to aid the attachment of a handle by hafting. These waisted axes were found in a terrace formed 45,000 to 55,000 years ago. With each volcanic episode layers of ash fell on the landscape, encasing the evidence of human occupation. On this terrace under 2 m of tephra were found two waisted axes (Plate 49), a core and some flakes (Groube et al., 1986). These stone artefacts are covered by a layer of tephra dated to an absolute minimum age of 35,000 years, and lie on another tephra layer dated to in excess of 37,000 years. (The dates are thermoluminescence dates obtained by the Australian National University and are regarded by Chappell as reliable (personal communication).) The absolute minimum age for these artefacts is regarded by Chappell and Groube as 37,000 years, and they are probably a few millennia older. This is the first evidence that human occupation of New Guinea is as old as that of Australia. The environment of the camp site on the Huon terrace was probably similar to that of today, with an average annual rainfall of over 2,500 mm (100 inches). The natural vegetation is coastal rain forest, although much of this has been cleared and the region now contains anthropogenic grasslands. About 40,000 years ago the area was probably clad in rain forest, possibly with some sago and mangroves present (Chappell, personal communication). Fire is the main tool used to clear such rain forest nowadays by the slash-and-burn method. It may be that the massive stone axes found by Joe Mangi, a Papua-New Guinean teaching fellow at the University of Port Moresby, and Groube were used for primitive horticulture. In an interview with Groube, the journalist Reinhardt reported: Groube is convinced the Huon axes were used for primitive agriculture. His argument is advanced cautiously into the uncertain along the following lines: ‘The opening up of the dense coastal rain forests must have been an immense problem for the earliest settlers in New Guinea which the massiveness of the Huon tools may reflect.’ ‘Whythese early colonists were attempting to trim back the forest edges, however, is unclear, whether to facilitate hunting or as an early trend towards utilizing the forest plant resources is unknown.’ But in contemporary Papua-New Guinea, Groube points to isolated clans in the remote west of the country who cultivate bananas in the middle of rain forests, simply by axing back enough vegetation cover for sunlight to reach a banana clump. ‘The possiblity undreamt of a few years ago, that the manipulation of rain forest resources, a form of casual “gardening”, might have a Pleistocene antiquity must be seriously considered, altering our views of the emergence of systematic gardening’ he says. ‘The evidence of botanists that many important garden plants are of New Guineainea origin, e.g. sugar cane, pan-danus, breadfruit, probably one of the bananas, perhaps coconuts, possibly the swamp taro and many tree crops, strengthen these hints.’ (Reinhardt, 1985, p. 91) Agriculture was being practised in the highlands of New Guinea 9,000 years ago. The evidence comes from the Wahgi Valley in the central highlands near Mount Hagen (Golson, 1977). When some tea planters drained a swamp in the late 1960s they discovered ancient digging-sticks, wooden paddle-shaped spades and stone axes. These artefacts were associated with many water-control ditches, probably dug to aid the growing of taro (Colocasia esculenta), cultivated for its edible, starchy tuberous root. The earliest ditch, 1 m deep by 2 m wide and some 450 m long, was radiocarbon dated to about 9,000 years ago. Taro, like the pig, is not native to New Guinea, so it must have been introduced. Other archaeological evidence in New Guinea shows that by 5,000 to 6,000 years ago plant cultivation, based on both native and non-native species, forest clearance, relatively permanent village settlements and complex water-management systems had already been developed. Elsewhere in New Guinea evidence of early human activity has come to light. The Huon artefacts are similar in some respects to 26,000-year-old tools from Kosipe– the earliest site previously known in New Guinea (White et al., 1970). The discovery of Pleistocene occupation at Kosipe was a surprise, for it lies in the south-east corner of New Guinea some 1,400 km from the western Ice Age coastline and in the highlands at 2,000 m above the present sea-level. Kosipe is an open camp site on a flat-topped, steep-sided ridge on the southern slopes of Mt Albert Edward, a 3,990 m peak on which there was a glacier and substantial ice area during the late Pleistocene. It thus seems clear that humans were paying at least seasonal visits to the highlands of south-east New Guinea 26,000 years ago, when the snow-line was only some 1,000 m above the camp site and the temperature would have been about 6 ◦C lower than at present.