THE SCIENCE OF FRONTIER DELIMITATION. 
209 
have a boundary to delimit,, and that the necessary preliminary to 
delimitation is a treaty, an agreement, a protocol, or some diplomatic 
instrument fixing the way the boundary is to be drawn. From the 
point of view of delimitation a great deal depends upon this instrument, 
and especially on the wording* of the clauses which define the frontier. 
It is most important that they should be clear and distinct, and that 
no ambiguous expression or words, which can be interpreted in two 
ways, should be used. The diplomatists should know exactly what 
they mean, and should state it clearly and concisely. A fine word or 
a well turned phrase, like the unintelligible part of a legal document, 
has got to be paid for sooner or later. The simplest words will be 
found cheapest in the end. 
It is also very important that as little as possible should be left for 
commissioners to decide on the spot. In almost all cases time is a 
matter of the first consideration in boundary work. If the task is not 
completed before the rains commence, or the season ends, a second 
commission must be sent out the following year at great expense. It 
is very difficult for commissioners with opposing interests to come to 
an agreement when matters are left entirely in their hands. Each 
officer naturally seeks to do the best he can for his country's interest, 
and it rarely happens that the claims of one side or the other are so 
clear and indisputable as to be acknowledged at once by the opposing 
side. Even if one commissioner gives way, he probably makes a cir¬ 
cumstance of having done so, and expects compensation in some other 
direction, which means the re-opening of a long discussion. In my 
recent experience we barely concluded work before the commencement 
of the rains; if we had had to spend any considerable time in discussing 
claims to particular districts, the commission would almost inevitably 
have been wrecked. It is easier, so far as I understand the matter, 
for diplomatists, who have a broad view of the whole question, to make 
concessions in drafting an agreement than for commissioners, who are 
mainly interested in seeing that they don't get the worst of the bargain, 
to give any point away in delimitating it on the ground. 
It can easily be understood that for every reason it is desirable that 
a boundary should follow natural features. It will generally bo found 
that natural features are used to divide tribes and races, and when 
these have in some instances pushed beyond the natural barriers, it is 
an easy matter to make them withdraw. A river forms, so far as my 
experience goes, an excellent natural frontier, by far the best from the 
delimiter's point of view, for it requires no delimitation. A river 
delimits itself. I am aware that there are certain disadvantages con¬ 
nected with the adoption of a river as a frontier, such as the question 
of water rights, navigation, the right of diverting the course of the 
water, etc. In using the term river, however, I do not intend to refer 
to small streams, but to the main drainage lines of a country. 
A mountain range makes a good natural boundary, particularly if it 
is well defined. But a mountain boundary must be delimited, and 
whether the line is drawn along tbe watershed, or follows straight lines 
joining the principal peaks, the ground must be examined closely. 
A conventional line, i.e. } a line drawn without any reference to 
