In attempting to fix a date for the first Melanesian settlement in Fiji the widest field lies open to the lover of speculation, for it is unlikely that when a few years have passed, and the last guardians of tradition have made way for young Fiji, any fresh evidence will come to light. The only monuments of a past age are rude earthworks in the form of moats and house foundations, a few stone enclosures known as _nanga_, no older than the period covered by tradition, and a stone cairn or two erected by the worshippers of the _luve-ni-wai_. The Melanesians buried their dead in their own houses if they were chiefs, leaving the house to fall to ruin over them; in the open if they were commoners, or in limestone caves wherever there were to be found, and there is no trace of tombs or hewn stone such as are found in Tonga and other islands colonized by Polynesians. Until the stalagmitic floors of the limestone caves have been examined systematically it is not safe to say that Paleolithic Man never inhabited the islands, but it is at least very unlikely. The earliest trace of human occupation yet discovered is a polished hatchet found in alluvial deposit on the bank of the River Mba about twelve feet below the surface, during excavations carried out in the erection of a sugar mill; but in a river subject to heavy annual floods, during which great quantities of soil are brought down from the hills, the depth is no proof of age. In the island of Waya (Yasawa) a _cache_ of polished hatchets was discovered in 1891. Three of these were gouge-shaped for cutting away the wood on the inside of canoes or drums, and of elaborate finish, but there was nothing to show that they were of ancient date.