On The Geometry Of The Kepler Problem