Algebraic relaxation of an order parameter in randomly coupled limit-cycle oscillators

In their recent paper [Phys. Rev. E 58, 1789 (1998)], Stiller and Radons (SR) study, following our earlier work [Phys. Rev. Lett. 68, 1073 (1992)], the behavior of globally and randomly coupled phase oscillators with distributed intrinsic frequencies. They claim that their simulation results do not confirm the power-law behavior of an order parameter found numerically by the author, attributing its cause to the poor precision of the author's integration scheme. Here demonstrated is that the power law survives even for a scheme better than SR's, provided that finite-size effects are properly taken into account, as was done in our previous work.