The treatment of this subject in recent publications is discussed. The procedure described by M. Perks in the present issue of THE CANADIAN SURVEYOR is found to be the most accurate, but it is more complicated than it need be. The procedure adopted at the National Research Council of Canada is then described. This makes use of a sequence of conformal transformations in two dimensions. It has the same accuracy as Perks’s procedure, but it is simpler in both formulation and use.