We present two new dynamical tests of the biasing hypothesis. The first is based on the amplitude and the shape of the galaxy-galaxy correlation function, xi g(r), where r is the separation of the galaxy pair. The second test uses the mean relative peculiar velocity for galaxy pairs, v12(r). This quantity is a measure of the rate of growth of clustering and it is related to the two-point correlation function for the matter density fluctuations, xi (r). Under the assumption that galaxies trace the mass (xi g = xi), the expected relative velocity can be calculated directly from the observed galaxy clustering. The above assumption can be tested by confronting the expected v12 with direct measurements from velocity-distance surveys. Both our methods are checked against N-body experiments and then compared with the xi g(r) and v12 estimated from the APM galaxy survey and the Mark III catalogue, respectively. Our results suggest that cosmological density parameter is low, omega m approximately 0.3, and that the APM galaxies trace the mass at separations r > or = 5 h-1, where h is the Hubble constant in units of 100 km s-1 Mpc. The present results agree with earlier studies, based on comparing higher-order correlations in the APM with weakly nonlinear perturbation theory. Both approaches constrain the linear bias factor to be within 20% of unity. If the existence of the feature we identified in the APM xi g(r)--the inflection point near xi g = 1--is confirmed by more accurate surveys, we may have discovered gravity's smoking gun: the long awaited "shoulder" in xi, generated by gravitational dynamics and predicted by Gott and Rees 25 years ago.