We consider a nonlinear differential equation of second order which has been proposed as a model for the periodic motion of a satellite in its elliptic orbit. We use some elementary methods to demonstrate the existence of an odd periodic solution for a considerably larger parameter set than considered earlier.
