Abstract
Introduction. In two recent papers [4], [30], Erdos, Stewart and the author showed that certain diophantine equations have many solutions. In this way they indicated how far certain results are capable for improvements at most. First we mention some relevant results from the literature on upper bounds for the numbers of solutions of diophantine equations and then we sketch how our method leads to opposite results. 1. Thue and Thue-Mahler equations. Let f(x; y) = a 0 x n + a 1 x n\\Gamma1 y + ::: + an y n 2 Z [x ; y ] be a binary form (i.e. homogeneous polynomial) of degree n 3: Put A := max j=0;:::;n ja j j. Let m<F39.