THE SCIENCE OF FRONTIER DELIMITATION. 
221 
co-ordinate. If the total distance is found too long or too short by 
subsequent observations, it is easy to recompile the topography. 
(5) A line 'parallel to a road, river or mountain range and at some 
considerable distance from it .— 
This is probably the most difficult and unsatisfactory form of 
boundary to mark out on the ground. To draw, for example, a line 
parallel to a river at a distance from it of 10 miles, 10 kilometres, 
or even one mile is a task enough to make one’s hair stand on end. 
Yet this is a form of boundary sometimes given. To measure the 
distance from the feature a road at right angles to it must be cut. In 
such a case the only course, as far as I can see, is to triangulate care¬ 
fully the whole district, to fill in the guiding feature as accurately as 
possible, and then to follow the line, at the given distance from it, as 
nearly as can be judged, and at any places, where roads appear to 
cross the frontier, to intersect the position with some fixed point and 
measure its distance on the map from the feature which gives the line. 
But this course, though it sounds simple on paper, may be found very 
difficult on the ground, and some other method may have to be adopted. 
If there are only one or two points where beacons ought to be erected, 
it may be possible near each point to measure a short base and to run 
one or two triangles so as to fix the distance from the feature. 
If the feature is to be followed at a very short distance, 500 yards 
or so, measurement is then not so difficult a matter. 
The defects of all these conventional lines are great, and they can 
only be met by compromise, and by agreeing beforehand how the 
difficulties, which are sure to be met, are to be dealt with. The line, 
it will be found, will frequently select the highest peaks, the lowest 
abysses and the most difficult ground, which the delimiting commission 
must follow. It will cut through towns and villages, separating families 
and households, leaving possibly a man on one side of the frontier and 
his family on the other. Now, although such divisions are not always 
regarded as an unmixed calamity by the natives, it is rare that the right 
separation is made, and however much a man may dislike his relatives, 
they become very dear to him when there is anything, even a grievance, 
to be made out of them. Too much, of course, must nob be made out 
of the division of towns by boundary lines. In an uncivilised country 
a town is not the fixture it is in Europe. There, when one builds a 
house, one expects it to remain more or less in the same position as 
long as it lasts. In uncivilised land, it is very different. In Africa 
you never know where a town may turn up next. It is no uncommon 
thing to find a spot formerly occupied by a town deserted, and on 
inquiry to be told that the chief has taken his town to some other 
place. This peculiarity of African towns is a source of frequent errors 
to cartographers; once they have got a place located, they expect it is 
going to remain there always, whereas it probably makes several moves. 
But though the town should be movable, the delimiter’s difficulty is 
not solved. If the boundary cuts it, it may be moved, but to which 
side ? This and other matters must be agreed upon by the com- 
