Comment on Local-Field Distribution in Random Dielectric Media