THE SCIENCE OF FRONTIER DELIMITATION. 
215 
the neighbourhood of the place where you expect to find it. A forest 
of trees has grown up beside your single tree, and your conical bill 
has turned into a twisted and contorted ridge. As you ascend the hill 
every tree you pass looks something like your object, but none resemble 
it exactly. When you are in this difficulty your compass is the only 
thiug to help you. You get yourself into alignment between the back 
station aud your point by the bearing, and you will soon find your 
tree. Not uncommonly you will find yourself on the wrong hill 
altogether, and will have to move forward or backward. On several 
occasions, during my work in West Africa, we set up the theodolite at 
points which we were confident were those observed, but the compass 
said “ no.” We endeavoured to account for the want of agreement by 
local attraction or an error in reading the bearing, but we always found 
that the compass was right, and that we had mistaken the point. 
Many errors were saved in this way. A further advantage in using 
magnetic bearings is that in plotting the work the direction of the 
sides of the triangles can be checked and errors in computation and in 
plotting discovered. 
Care should be taken to keep the principal triangles well proportioned, 
and to avoid very acute or obtuse angles. The better the triangles, 
the more will any errors of measurement be minimised. 
It is important to intersect all distant conspicuous points from several 
stations, and it is specially important to cut in by intersections any 
distant point, which is known or is on a route which has been 
explored. 
As the work of the triangulation is required for immediate use, the 
triangles must be computed each day as the field-work is done. To 
avoid error it is most desirable that two officers should work out all these 
solutions independently, and compare results, so that errors can be 
checked. The same precaution should be adopted in all astronomical 
work. It will save much re-calculation in the end, for under the 
circumstances of delimitation errors are easily made. 
The triangles should be plotted as soon as computed, and a chart 
should be kept up of the triangulation from which the topographer 
should take his points day by day. 
From time to time observations for latitude and time should be made. 
These will give results less accurate than those obtained from the 
triangulation, but they will serve to locate errors and to check each 
stage of the work. 
The triangulation must close as it began on fixed points. If there 
are no such, the closing point must be fixed in latitude and longitude. 
To test the correctness of the work it is useful, if time admits, to fix 
29 
