is not an impossible task and could be 
     
     accomplished for example by using maps such as 
     
     those of Lake in 1863 or Meacham in 1880 — by 
     
     which time, as just discussed, the island had been 
     
     more or less fully settled. However, since neither 
     
     figure is at hand, as a substitute we may make use 
     
     of the length of the present-day road-system — 
     
     5,211 km (or 3,236 miles)246 — which in fact 
     
     should not differ greatly from that shown in 
     
     Meacham in 1880. We can calculate that the 
     
     creation of this road system must have resulted in 
     
     the destruction of 9,530 hectares (or 23,548 
     
     acres) of forested land, which works out at 1.7% 
     
     of the area of the island that would have once 
     
     been under forest247 — quite a small proportion in 
     
     comparison with the amount of forest lost to 
     
     agricultural clearance (56.7%).248 
     
    
   
   
    
     A more direct comparison of the relative effect on 
     
     the forest of the clearance due to road-making 
     
     with that due to farming can be obtained by taking 
     
     what was considered the standard farm size of 
     
     100 acres with its frontage on the road of 10 
     
     chains (660 feet).249 The road in front of such a 
     
     farm would take up 0.91 acres of land (equal to 
     
     0.91% of the total farm area), and when we split 
     
     this into two to take account of the farms on the 
     
     opposite side of the road, this fails to 0.45 acres 
     
     per farm. This means that for each 100 acre farm, 
     
     there will have been the loss of an extra half acre 
     
     of woodland for a road. If we then take the 
     
     11,439 farms recorded in the 1861 census, the 
     
     amount of forest clearance for road-works directly 
     
     associated with farming at that time comes to 
     
     5,200 acres, which works out at 1.4% of the total 
     
     amount of cleared agricultural land recorded in the 
     
     same census (365,995 acres).250 This percentage 
     
    
   
   
    
       
     
    
   
   
    
     2‘6 This figure has been provided to me by William Glen and is 
     
    
   
   
    
     based on government records. 
     
    
   
   
    
     2‘7 MacDougall et al. (1988) state that the total land area of the 
     
     island is 573,525 hectares (or 1,367,782 acres). Subtracting 
     
     19,991 hectares of salt-marsh, coastal beach and dune sand; and 
     
     4,879 of treeless bogs leaves 548,655 ha (or 1,355,726 acres) as 
     
     the area that was once wooded. 9,530 divided by 548,665 gives 
     
     1.74%. 
     
     2‘8 The 1911 census recorded the apogee of agricultural 
     
     clearance on the island, with a total of 769,140 acres (or 311,267 
     
     ha) classed as ‘improved land', which is 56.7% of the area once 
     
     wooded. (The census figure comes from Glen (1995)). 
     
    

    
     2‘9 Both Johnstone (1822) and Lawson (1851) make reference 
     
     to this as the standard size and frontage for a pioneer farm, and 
     
     examination of any township map in 'Meacham's Atlas' (Allen 
     
     1880) shows large numbers of farms of this size. 
     
    
   
   
    
     250 
     
    
   
   
    
     l have taken the number of farms in 1861 from Clark (1959) (p, 
     
     95), and the numbers of acres of ‘arable' from Glen (unpublished, 
     
     see footnote 200). Both numbers derive from the census of 1861. 
     
    
   
   
    
     42 
     
    
   
   
    
     will have fallen with the continuing clearance of 
     
     woodland on each farm for agriculture that we 
     
     noted in the last section. And of course there 
     
     would also have been additional connecting roads 
     
     running through areas of forest that were 
     
     unsuitable for clearance, that would need to be 
     
     added into our calculation. 
     
    
   
   
    
     A second effect of road-building on the forest was 
     
     that the creation of a linear 60 foot wide opening 
     
     through old-growth forest exposed the trees still 
     
     standing on either side to wind-throw.251 As a 
     
     result it became necessary to cut down the taller 
     
     trees beyond the 60 foot width so as to prevent 
     
     fallen trees from blocking the newly created track- 
     
     way. In the early days this probably was not 
     
     done, which is why Walter Johnstone could write 
     
     of the obstructions caused by windfalls, as well as 
     
     of the actual dangers of traveling along the 
     
     island’s roads during periods of high winds.252 
     
     Wherever it was fully carried out, the cleared zone 
     
     would have extended a further 80 or 90 feet on 
     
     either side of the track way.253 Of course with the 
     
     arrival of settlers taking up farms along the road, 
     
     all roadside trees, at least in the hardwood areas 
     
     favoured for settlement, would have been cut 
     
     down anyway during the process of laying out the 
     
     new farms and clearing the land, so in many areas 
     
     the extent of the tree felling along the roadside 
     
     verges was only a question of the timing. 
     
    
   
   
    
     A third way in which road-building had an effect 
     
     on the forest was due to the need to create a 
     
     corduroy surface of logs for roads passing through 
     
     swampy or wet areas. As Johnstone (1822) 
     
     stated in the passage just quoted, ”small softwood 
     
     trees" were used for this purpose, while Selkirk 
     
    
   
   
    
       
     
    
   
   
    
     251 Though I expect that sometimes the cleared width would have 
     
    
   
   
    
     been less than 60 feet (e.g. Mann (1829) describes a section of 
     
     road in present—day Fernwood where “going through the woods 
     
     was like going through a cave"), a letter of April 1824 signed by 
     
     ‘Viator' to the Prince EdWard Island Register (cited by Vass 1986), 
     
     implies that the 60 foot width was followed, since he advocates that 
     
     it be reduced to 15 feet, because, he says, “the felling of the forest 
     
     for the roads 60 feet has been found very injurious in winter'. (He 
     
     does not state in what sense it was is ‘injurious‘.) 
     
    

    
     252 Johnstone‘s passage on windfalls has been extracted 
     
     (Johnstone 1822), but the following passage (p. 137) has not: “as 
     
     I passed along the St. Peter’s road some [trees] actually fell so 
     
     near me that I had to keep a sharp look out to avoid accidents 
     
     when the gale raged in its utmost height". 
     
    

    
     253 There are very few records for the heights of trees in the 
     
     island’s old—growth forests: we have no estimates at all for 
     
     hardwood trees, but for pine, hemlock and spruce (see Appendix 
     
     1) the maximum heights in the records are of the order of 100 feet 
     
     — though this may simply be a rounding to an impressive sounding 
     
     number.