O 38 
AERIAL TRIANGULATION 
By 
E.H. THOMPSON 
UNIVERSITY COLLEGE LONDON 
Invited paper to Commission III 
Lausanne, 1968 
1. Introduction 
In this review of the present position of aerial triangulation 
I do not propose to list and describe systematically the methods in 
general use to-day. Not only would it be tedious for the reader, but 
it would not be easy,even if their merits and demerits are carefully 
set out, to bring to the notice of workers in this field where I 
think progress can be most usefully made during the next few years. 
To do this is, I consider, the most useful result that can be 
expected of such a review; and it would, if successful, provide a 
working guide for the activities of Commission III from now until 
the Xllth Congress. If not so successful it should, at least, pro 
vide material for discussion at the Lausanne Congress, 
I have not made exaggerated attempts to be impersonal. The 
views expressed are largely my own, and although they are, I hope, 
based upon an objective view of the current situation they are, none 
the less, controversial and there will be many who will not agree 
with me. 
I have endeavoured to divide the subject into two main 
groups: the acquisition of data and the computation. It is extremely 
important, in my view, to keep these two matters entirely separate; 
more so since the automatic computer has encouraged the proliferation 
of programmes many of which, I fear, see the light of day for no 
better reason than that they have been written. The old cry of 
M I got a horse” is being rapidly replaced by "I got a programme". 
That is not to say that the computing problem has been solved, but 
it should, I think, be regarded in a somewhat different light: 
complete automation has its points but there are other aspects to 
be considered. 
We are, I think, at a crucial stage in the development of 
aerial triangulation. No significant increases in accuracy seem 
to have been made for some time, but it would certainly be rash to 
suppose that the end has been reached. It is now fairly well 
agreed by astronomers that no further improvements in the positions 
of stars are going to be made with the traditional methods: zenith 
tubes and meridian circles. But what is surprising is that there 
is reason to suppose that accuracies can be significantly improved 
by employing photographic triangulation. A strip of overlapping 
photographs of the celestial sphere will give the relative positions 
of the photographed stars (and all star positions are basically 
317