Indians were repeatedly seen upon both Bentinck's and Sweers' Islands; but they always avoided us, and sometimes disappeared in a manner which seemed extraordinary. It is probable that they hid themselves in caves dug in the ground; for we discovered in one instance a large hole, containing two apartments (so to call them), in each of which a man might lie down. Fire places under the shade of the trees, with dried grass spread around, were often met with; and these I apprehend to be their fine-weather, and the caves their foul-weather residences. The fern or some similar root, appears to form a part of their subsistence; for there were some places in the sand and in the dry swamps, where the ground had been so dug up with pointed sticks that it resembled the work of a herd of swine. Whether these people reside constantly upon the islands, or come over at certain seasons from the main, was uncertain; canoes, they seemed to have none, but to make their voyages upon rafts similar to those seen at Horse-shoe Island, and of which some were found on the shore in other places. I had been taught by the Dutch accounts to expect that the inhabitants of Carpentaria were ferocious, and armed with bows and arrows as well as spears. I found them to be timid; and so desirous to avoid intercourse with strangers, that it was by surprise alone that our sole interview, that at Horse-shoe Island, was brought about; and certainly there was then nothing ferocious in their conduct. Of bows and arrows not the least indication was perceived, either at these islands or at Coen River; and the spears were too heavy and clumsily made, to be dangerous as offensive weapons: in the defensive, they might have some importance.