suggests that under the canopy of the mature forest (probably in the hardwoods) the shrub and herb layers were not well—developed and appear to have presented no great obstruction to travellers. An exception is likely to have occurred in the areas covered by ground hemlock — as was recorded by Roma (see above). Were there open treeless areas in the forest? — While sailing along the north coast of the island Jacques Cartier (1534) noted the presence of ’prairies’ (i.e. meadows or grasslands) — these must have been the marram-dominated sand-dunes especially prominent along the northern coast — or possibly in places salt-marshes. However, elsewhere his account seems to suggest the presence of something different from either sand— dunes or salt-marsh: based on what he saw at the four places at which he landed (which appear to have been along the western shore between North Cape and West Point), he says that ”the soil where there are no trees is also very rich and is covered with poys [peas], grouaise/iers blans et rouge [white and red gooseberries — or currants], frasses [strawberries], franboysses [raspberries] and b/é sauvaige comme sei/le, [wild oats like rye] which one would say had been sown and tilled”. These species imply a clearing on normal forest soils as opposed to areas of salt-marsh or sand-dunes. Given that he spent only a short time on the sites (at most hours, or even less, during a single day), what he saw must have been directly on the coast and could not have been far from the shore. His treeless areas might have been coastal cliff-top grassland — though later records from the British period suggest that forest thicket in this area went right to the cliff edge“. It may be that on this western coast, exposed as it was to the prevailing westerly wind, there were areas where even white spruce thicket could not survive. horseback. Only the old trees which are fallen in one place and another could offer any hindrance." Ganong (p. 377) considered this description of the forests of Acadia ‘incorrect', stating: “it is only in a few limited areas, especially in occasional pine or hardwood groves, that they are open; elsewhere they are dense, obstructed, and practically impassable for horses". However, other historical records suggest that Denys was correct and it is Ganong who is in error. (e.g. Bishop Plessis on his pastoral visit to Prince Edward Island in 1812 (this will be included in the source-book for the British period) says that the forests are claires [open]; and J. W. F. Johnston in 1851 (cited by Wynn 1981, p. 158) described a beech forest in New Brunswick that was “entirely free from underwood and in many places so open that it would be easy to ride through it on horseback’. 53 Anon. (1762) Remarks relative to the Sketch of the Island of St John's in North America [Hardwicke Papers, British Library: Add. MSS 35914, 95-99ff; and PARO: 4615.] u Gotteville (1720) records a lot of meadows [quantité de prairies] where there are a few lakes here and there”. If he had said simply quantité de prairies we might assume that he was referring to salt-marshes. It is the combination of meadows with lakes that is peculiar. It is a well known fact that Prince Edward Island has virtually no lakes, and certainly none that Gotteville is likely to have seen in the area of Port La-Joie and its rivers before recording the above. It is thus not clear what he is describing, unless it is the shallow lagoon-type bays that are associated with salt- marshes and sand-dunes along parts of the northern shore of the island. Finally, in 1750 Roma refers to trees in association with what he calls ’prairies nature/les’ (literally 'natural grasslands’, which would normally be taken to refer to the salt marshes so important in Acadian agriculture). What he says is that since there was only a thin scattering of trees around such prairies, their area could be doubled with little effort. However, he cannot be referring to salt- marshes as trees would not have been able to grow on them at all. It may be that he is referring to the grasslands on dune slacks which would have had a thin scattering of trees especially white spruce.‘54 Bog forest — Bonnaventures' (1753) description of the coastal woodland between West Point and the Percival River as ”a quagmire of poor spruce trees or small firs same shaky ground” [un c/oique mauvesce prusiere ou peti sapeins . mesme péy tranb/an]. Although his poor hand-writing and spelling leaves us ourselves on somewhat shaky ground in the translation, we can recognise this as a description of a type of forest that is otherwise not mentioned in the records of the French period, i.e. black spruce bog forest — which in fact still occurs on much of the area that Bonnaventure was describing.55 The only other reference to such wet areas in the forest is Pensens' 1728 mention of the mo/ieres (a word that can mean swamps, bogs or marshes)56 that presented difficulties in 5“ See Erskine 1960, pp. 27-28. 55 See Sobey and Glen 1999. 56 The actual word was partly hidden by the binding when the page was photographed for the PAC microfilm copy so that only the first three letters were visible, i.e. ‘mo/-‘ Margry's transcription from the original document gives the word as mofiere. a word not listed in any French dictionnary. However, given the frequent lack of the cross on the t in eighteenth century French hand—writing, it is more likely to have been moliere, an archaic French word defined by Littré (1969) as “terre grasse ou marécageuse” [i.e. rich land or boggy/swampy/marshy land].