Page:Voyage in search of La Perouse, volume 1 (Stockdale).djvu/185

] it off in pieces of twenty-five or thirty feet in length.

Mot of the large trees near the edges of the ea have been hollowed near their roots by means of fire. The cavities are generally directed towards the north-eat, o as to erve as places of helter againt the outh-wet winds, which appear to be the mot predominant and violent in thee parts. It cannot be doubted that thee cavities are the work of men; for had they been produced by any accidental caue, uch as the underwood taking fire, the flames mut have encompaed the whole circumference of the tree. They eem to be places of helter for the natives whilt they eat their meals. We found in ome of them the remains of the hell-fih on which they feed, and frequently the cinders of the fires at which they had dreed their victuals. The avages, however, are not very afe in thee hollow trees; for the trunk being weakened by the excavation, may eaily be thrown down by a violent gut of wind; neither are their eats very commodious, as the ground is very uneven, and we oberved no contrivances to render it more level. Anderon peaks of hearths of clay, made by the natives in thee hollow trees. Whenever I have found any clay in them, it did not appear to me to have been placed there by the avages; but one