[en] The uniqueness of the direct reconstruction of pp elastic scattering amplitudes from the observables measured in the Nucleon-Nucleon experimental program at SATURNE 2 is discussed. The amplitudes were deduced analytically solving bilinear relations expressing the observables in terms of invariant amplitudes. It is shown that the choice of observables in the experimental program allows, in principle, a complete and unique analytical reconstruction at all energies involved if experimental errors are ignored. This is shown to apply to two different sets of 11 observables and to a set of 10 observables. 18 refs