We have considered a two-particle Bell experiment to visualize the conflict between rotational invariance of physical laws and a specific local realistic theory. The experiment is reproducible by using a local realistic theory obtained in a two-setting Bell experiment. The generalized Bell inequality [J. Phys. A: Math. Theor. 40, 13101 (2007)], which is derived under the assumption that there exists a rotationally invariant local realistic theory, turns out to disprove such a local realistic model existing with a two-setting experiment. This implies that such a model is riot rotationally invariant and should, therefore, be ruled out in some situations.