“…From the discussion following Theorem 1.1 it is immediately seen that, for a given p, the condition This means that, in contrast with (2), oscillations of g around 0 at ±∞ are, in fact, allowed, but the length of the intervals I ± where g does not change sign is determined by g and p, and cannot be arbitrarily small. For instance, when g(u) = sin u, there are well known examples of forcing terms p with zero average and c = 0 such that the problem has no solutions (see [1], [7], [9]).…”