We study the convergence of N-1 Sigmaf(1)(T(a1n)x)f(2)(T(a2n)x)f(3)(T(a3n)x
), for a measure-preserving system (X, B, mu, T) and f(1), f(2), f(3) is an
element of L-infinity(mu). This generalizes the theorem of Conze and Lesig
ne on such expressions and simplifies the proof. We also obtain a descripti
on of the limit.