The Watson multiple scattering theory is used to generate the first and second order optical potentials for nucleon-nucleus elastic scattering at high energies. A method is developed for treating exactly the non-locality of the second order potential and the scattering from the exact calculation is compared to that from a local approximation for the second order potential. The results are significantly different both for the angular distributions and the polarisation. Simple methods are used to include in the theory the finite range of the two body interaction and off---two---body--- energy-shell effects. The inclusion of these phenomena are seen to improve the agreement with the data. Neither the exact nor the approximate calculation gives very good fits to the data and some reasons for this are considered. Some comparison is made with the work of other authors, both in terms of the formal theory and the resulting calculations. Differences due to choosing different two body phase shifts are seen to be comparable to the errors due to approximating the second order potential.