consumed, its bark half burnt, and covered with a sooty blackness; if hard wood, and viewed at a distance, one would think it did not know that summer was come. But more narrowly examined, the trunks are found rotting, the bark partly peeled off, the leafless branches falling down, and the whole verging fast to decay. But if a few years more have elapsed since the fatal flame passed through it, a scene more revolting remains yet to be described. Some of the trees are overturned by the roots, with great mounds of earth attached; others broken at different heights, where most weakened by the fire. These all lying upon the ground, in the most confused manner possible, several courses deep, form a confused mass impassable for either man or beast, but with the greatest difficulty. This is a faint representation of what is lying on the ground, but when the eye is turned to what is still standing, we behold trunks of trees, of various heights, having lost their tops as if they had been cut with chain shot, and here and there a huge pine or hemlock, a great unsightly object, with a trunk thicker than a corn sack, barkless and weather-beaten, raising its top, perhaps a hundred feet high, into the air, and reaching out a few half wasted branches. Thirty years later, John Lawson (1851) noted the ”melancholy look in the spring and autumn [of] the half burnt and blackened stumps”, adding that ”the whole Island was once in the same cheerless state”; while even near the end of the nineteenth century a travel writer, Anna Ward (1887), observing on her train journey through the island the ”burnt districts where, leafless and limbless, blackly charred trunks mingle with thrifty forest trees”, lamented ”the ravages of the fire-fiends who have done their best to obliterate the original forests”. 377 Some ecological points about fire — The lack of any useful documentation, does not prevent us from speculating that the spatial distribution of the starting points of the fires would have been determined to some degree by the distribution of active clearances in new settlements, where fires were being set to burn surplus wood. However, because, as we have seen, forest fire could readily spread many miles beyond the bounds of any settlement area, it is likely that the spatial distribution of all of the burnt areas resulting from such fires will have shown less relation to the 377 As far as I know, there have not survived any visual representations or photographs of the ‘burnt woods' of the island. The nearest that I have come across is in the view of the farm property of John McRae, New Glasgow Road, Wheatley River, Lot 24, in ‘Meacham’s Atlas‘ (Allen 1880) (p. 86), where in the background there appears to be what looks like a forest landscape degraded by fire. 61 settlement locations (as was also evident for the French period fires as shown in Figure 5). For, once a fire had spread to the woods from a clearance burn, other factors such as the forest- types in the surrounding area and the location of barriers to fire, such as rivers and shorelines, as well as chance factors such as weather conditions and wind direction at the time, will have played a major role in determining the area burned in a fire, and thus of the spatial distribution of all of the burned areas taken together. This pattern would of course have been constantly changing from year to year as a result of new fires occurring. Due to the lack of documentation, it is also now impossible to distinguish between the relative loss of the island’s pre-settlement forest to fire and to land clearance operations, either in terms of the area of land affected or of the volume of timber destroyed. In the early years, however, just as it had in the French period, it is likely that runaway forest fire is likely to have affected a far greater acreage of forest than that destroyed by the clearances of the pioneer settlers. We might also expect that coniferous forest, as well as any mixed forest with a high conifer component (or even small patches of conifers within the hardwood forest) would have been more susceptible to fire than pure hardwood forest.378 Even so, we must not imagine that the upland hardwood forests of the island were immune to fire, for such forests often had a minority component of scattered conifer trees (especially white pine, hemlock and red spruce), that was likely to have given them some susceptibility. In fact in his description of the ”burnt woods”, just quoted, Johnstone (1822) refers to both burned ‘hard wood’ trees, and also to "here and there a huge pine or hemlock”, which indicates that the burnt woods that he had in mind included upland hardwood forest. Also, his comment elsewhere that "when the hardwood is burnt upon a farm" the people are without fuelwood, clearly implies a fire in hardwood forest. As well, the autumn ground fires that Selkirk (1803) said scorched only the trunks of the trees without getting up into the crown, appear also to have occurred in deciduous woods, since he tells us that it was only in the following year when the new leaves did not appear on the trees, that it was realised that the trees had been killed.379 Finally, 378 As noted, Johnstone (1822) commented that "woods much mixed with softwood” favoured the spread of fire. 379 There are also references to fire in hardwood forest in the evidence submitted to the Land Commission (1875): James Farrar Stewart, the landlord of Lot 7, referred to ”heavy hard wood