Page:The Melanesians Studies in their Anthropology and Folklore.djvu/324

302 floor has sunk it below the outside level, and that this again has been raised close round the house by accumulation of various rubbish. When the new house is to be built, the hollow inside is filled with the outside accumulation, and the result is a little elevation of the site. If then an ancient site is seen some four or five feet high, it must represent a pretty long occupation of the ground.

In two islands far apart, in Ysabel and in Santa Maria, there are very remarkable structures to be seen. In Bugotu, Ysabel, Bishop Patteson slept in 1866 in a fortified place thus described. 'The site for the village has been chosen on a hill surmounted by steep, almost perpendicular coral rocks; the forest has been cleared for some space all round, so as to prevent any enemy from approaching unperceived; there is a wall of stones of considerable height on that side where the rock is less precipitous, with one narrow entrance, approached only by a smooth slippery trunk of a tree, laid at a somewhat steep inclination over a hollow below.' So also at Tega the people built a toa, 'an impregnable fort on a rocky knoll in the midst of the village.' These forts are made for protection against head-hunters. The stone buildings in a village in Gaua, Santa Maria, are very extraordinary; nothing like them has been seen elsewhere in these islands. There are three small gamal houses on platforms about ten feet high, built up with stones untouched by any tool, and some of them three feet long by two deep. The building is wonderfully square and regular; the style quite Cyclopean, the large stones ingeniously fitted, and the interstices filled with small stones. Besides these platforms there are two or three obelisks about four feet high, and a little dolmen of three stones. There are also two wona platforms, such as are always seen near a gamal (page 101), but much larger, and built of large stones very squarely put together. In one of these is a passage for pigs with a stone lintel. These remarkable works are shown in the frontispiece. That such stone-work exists elsewhere in these islands cannot be positively denied, but none has been heard of, and in the neighbouring islands there