f the distance of the
small star-like planet from a star is therefore characterised by great
accuracy.

But there is another and, perhaps, a weightier argument in favour of the
determination of the scale of the solar system by this process. The real
strength of the minor planet method rests hardly so much on the
individual accuracy of the observations, as on the fact that from the
nature of the method a considerable number of repetitions can be
concentrated on the result. It will, of course, be understood that when
we speak of the accuracy of an observation, it is not to be presumed
that it can ever be entirely free from error. Errors always exist, and
though they may be small, yet if the quantity to be measured is minute,
an error of intrinsic insignificance may amount to an appreciable
fraction of the whole. The one way by which their effect can be subdued
is by taking the mean of a large number of observations. This is the
real source of the value of the minor planet method. We have not to wait
for the occurrence of rare events like the transit of Venus. Each year
will witness the approach of some one or more minor planets sufficiently
close to the earth to render the method applicable. The varied
circumstances attending each planet, and the great variety of the
observations which may be made upon it, will further conduce to
eliminate error.

As the planet pursues its course through the sky, which is everywhere
studded over with countless myriads of minute stars, it is evident that
this body, itself so like a star, will always have some stars in its
immediate neighbourhood. As the movements of the planet are well known,
we can foretell where it will be on each night that it is to be
observed. It is thus possible to prearrange with observers in
widely-different parts of the earth as to the observations to be made on
each particular night.

An attempt has been made, on the suggestion of Dr. Gill, to carry out
this method on a scale commensurate with its importance. The planets
Iris, Victoria, and Sappho happened, in the years 1888 and 1889, to
approach so close to the earth that arrangements were made for
simultaneous measurements in both the northern and the southern
hemispheres. A scheme was completely drawn up many months before the
observations were to commence. Each observer who participated in the
work was thus advised beforehand of the stars which were to be employed
each night. Viewed from any part of the e