[GANONG ] BOUNDARIES OF NEW BRUNSWICK 415 
At the first division of the province in 1785, eight counties were 
established with the lines shown on Map No. 35, and to these seven 
have since been added. Though at first sight the reason for their 
peculiar arrangement may not be obvious, a closer inspection shows 
that they are correlated in a general way with the topography of the 
province. The county lines are evidently so drawn as to make the navig- 
able waters of the province the centres of the counties, for which pur- 
pose the lines of separation must follow approximately the watersheds. 
This was, of course, by far the best arrangement in a new, rough, and 
heavily-forested country like New Brunswick, where at first all travel 
was necessarily by water, and all settlements were beside and centered 
in navigable waters. The very topography of New Brunswick divided 
its population naturally into communities, centering about the prin- 
cipal waterways and more or less isolated from one another, and the 
county divisions are simply a formal expression of this condition. 
That this is the general principle on which the county boundaries were 
established is rendered still more plain as we trace the formation of 
the later counties down to the present, and especially as we note the 
rearrangements of the county lines as the topography of the province 
became better known. The fact that the county lines do not corres- 
pond exactly with the watersheds (see Map No. 36), is no objection 
to this view, for it is obvious that three minor causes have operated 
to prevent such a correspondence. First, owing to the fact that most 
of the surface of New Brunswick is of the character known to physio- 
graphers as peneplained (that is composed of fragments of ancient 
plateaus), the watersheds are rarely pronounced ranges of hills easily 
seen, but are more frequently in a flat county and are very winding 
and difficult to follow or recognize. Hence, boundaries must be 
marked by artificial lines, which, for convenience of running and mark- 
ing, are best made straight. While following, however, the general 
courses of the watersheds, the lines must often deviate, sometimes 
considerably, from the details of their wanderings. Second, many of our 
rivers head far back into the natural basins of others, even to an 
extreme degree, in which cases it is more convenient to include 
their heads with the rivers they approach. This has been the case with 
the South-West Miramichi, the Restigouche and some others. Third, 
the imperfect topographical knowledge of the time, reflected as it was 
in the imperfect maps, led to the establishment of some lines along sup- 
posed watersheds, which later were found to lie elsewhere; in some 
such cases the lines were afterwards altered, but in others, where the 
discrepancy was not serious (or as in the case of the St. Croix, was 
actually advantageous), they were allowed to remain. Thus the Kings- 
