nor does the form of the earth itself present any peculiarity which
might be used as an initial point. If the refinements of geodesy
should finally lead to the conclusion that the figure of the earth is
an ellipsoid with three axes, yet the question of the direction of
either of the equatorial axes must remain to such a degree uncertain
that the extremity of the axis could not be assumed as the point of
departure for counting longitude. Indeed, as an initial meridian must
above all things be fixed in position, it would not answer to make its
position depend upon any physical constant which is itself in the
slightest degree uncertain; for in these days, when refinements in
physical measurements are constantly leading to more and more accurate
results, each advance in accuracy would necessitate an annoying change
in the initial meridian, or, what would more probably result, the
retention of the first chosen meridian, which would thus lose its
dependence upon the original definition, and become as arbitrary as
if taken by chance in the first instance.

We may then say that, from a purely scientific point of view, any
meridian may be taken as the prime meridian. But from the standpoint
of convenience and economy there is undoubtedly much room for a
choice.

Considering this question of convenience in connection with the
necessary condition of fixity already referred to, the prime meridian
should pass through some well-established national observatory.

In making the choice of a prime meridian which is to serve for a great
period of time, it is important to so fix and define it that the
natural changes of time may not render it in the least degree
uncertain. To this end, the nation within whose borders the chosen