Comment on: ``Semiclassical theory of spin-orbit interactions using spin coherent states''