The natural forest as an indicator of soil quality — 
     
     Early emigrants from Europe arriving anywhere in 
     
     North America soon realized that there was a great 
     
     variation in the suitability of the forested lands for 
     
     conversion to farmland, and in an age before soil 
     
     surveys and a knowledge of soil chemistry, they 
     
     soon became aware that the forest trees 
     
     themselves were a valuable indicator of the quality 
     
     of the soils that lay beneath them.158 The same 
     
     was very early realized on Prince Edward Island, 
     
     and it is for this reason alone that many of the 
     
     early manuals and handbooks for immigrants 
     
     included information on the trees.159 As well, 
     
     undoubtedly a lot of advice of this nature must 
     
     also have been given casually to new arrivals by 
     
     established and experienced island residents 
     
     without it ever having been written down. For us 
     
     today, a beneficial side-effect of what did get 
     
     recorded, is that were it not for this use of the 
     
     natural forest as a pragmatic indicator of soil 
     
     quality — for most of the writers were not really 
     
     interested in the forest per se — we would have 
     
     much less information than we do, on the tree 
     
     species make-up of the forests of the island. 
     
    
   
   
    
     In Table 3 I have assembled all of the examples 
     
     that l have come across in the records of the use 
     
     of the natural forest as an indicator of the inherent 
     
     quality of the soil. Table 4 presents information of 
     
     a similar nature, though collected for a different 
     
     purpose: it contains the evidence recorded by the 
     
     Land Commission of 1875 on the opinions of 
     
     tenant farmers and others on the value for 
     
     conversion to farmland of the land then still under 
     
     forest — this was in order to assist the Commission 
     
     in its attempt to put a monetary value on the 
     
     estates of the remaining landlords, including any 
     
     land still forested and unoccupied, such as 
     
     occurred especially in parts of the west of the 
     
     province. 
     
    
   
   
    
     The best generalized classification in Table 3 is 
     
     that of John Stewart (1806) who sub-divided the 
     
     soil into four classes based on differences not just 
     
     in the tree species make—up of the natural forest, 
     
     but also on the size and spacing of the trees, and 
     
     on the soil depth as evidenced by the visibility of 
     
    
   
   
    
     ‘58 For the generality of this, see Williams (1989) (p. 60). 
     
    
   
   
    
     '59 Stewart 1806; Johnstone 1822; MacGregor 1828; Hill 1839; 
     
     Lawson 1851. Also, Gesner (1846) ([pt 18]), in a passage not 
     
     extracted, refers to the same phenomenon, when he said ”the 
     
     trees of the stately forest, and the indigenous plants have arranged 
     
     themselves in groups on the kinds of soils most favourable to their 
     
     growth" — though he did not go on to give any examples. 
     
    
   
   
    
     24 
     
    
   
   
    
     the tree roots on the surface. Johnstone (1822) 
     
     presents a simplified variant of Stewart’s scheme, 
     
     though it is Selkirk (1803) who gives the most 
     
     detailed information on the soil indicator value of a 
     
     wider range of individual tree species — the source 
     
     of his information, as he acknowledges, being the 
     
    
   
   
    
     surveyor-general Thomas Wright, who 
     
     accompanied him to the Belfast area. 
     
     As can be seen from Table 3, the primary 
     
    
   
   
    
     distinction in all of the classifications was between 
     
     the soils under hardwood forest, which were found 
     
     to be of a better quality compared with those 
     
     under the various conifer trees, which were 
     
     thought to be ”inferior” and sometimes ”hardly 
     
     worth cultivating” — this is in fact as far as several 
     
     of the schemes in Table 3 9016‘). However, the 
     
     fact that within the range covered by the term 
     
     'hardwood forest’, there were more subtle 
     
     differences in the soils which could be ’read’ from 
     
     the trees present, was recognized by several 
     
     recorders‘e‘. For example, Selkirk (1803) noted 
     
     that beech or maple on its own was a sign of dry 
     
     land, while the presence of yellow birch indicated 
     
     more moist soils, while Stewart (1806) stated that 
     
     where the hardwood forest contained ’dwarf yew’ 
     
     (i.e. Canada yew Taxus canadens/s), and a 
     
     scattering of conifers (presumably white pine, red 
     
     spruce or hemlock), the soils were considered to 
     
    
   
   
    
     be the best of all. 
     
    
   
   
    
     There are a few contradictions among the various 
     
     classifications, the most notable being the differing 
     
     opinions as to the quality of the soils under 
     
     hemlock‘ez. Another factor was that the soils 
     
     under pure hardwoods, especially the "beech 
     
     lands” referred to by Selkirk (1803), tended to be 
     
     too dry for conversion to hay pasture for livestock, 
     
     whereas ”swampy or sappy" land (presumably 
     
     under wet forest) had a value unrelated to its 
     
     ability to be cultivated, especially in the early 
     
     pioneer years, when the farm livestock could be 
     
     turned out into such woods for grazing, and after 
     
     their clearance, they could be converted into 
     
     meadow land for further grazing purposes 
     
    
   
   
    
     160 
     
    
   
   
    
     Curtis 1775; MacGregor 1828; Bouchette 1832; Hill 1839; 
     
     Lawson 1851. 
     
    
   
   
    
     161 
     
    
   
   
    
     Stewart 1806, Johnstone 1822, Selkirk 1803, Hill 1839. 
     
    
   
   
    
     ‘52 On the one hand Selkirk (1803) and Lawson (1851) reckoned 
     
     that where hemlock was "prevalent" it was a sign of bad land; 
     
     however, Johnstone (1822) and Seymour (1840) seemed to 
     
     consider it more favourably. It may be that it was the amount of 
     
     hemlock that was important, its presence in the hardwood forest 
     
     being a sign of a better soil, but where it predominated, the soils 
     
     were not considered as good,