On non-periodic and non-dense billiard trajectories Part 1 and Part 2

Abstract

This paper shows that the method of Galperin which had been widely accepted in the literature since 1983 for constructing a non-dense and non-periodic trajectory in a triangle can never work. This is done by showing that all possible examples that can be produced by Galperin's method of which there are infinitely many all produce a periodic trajectory.