An error in the normalization of the Aplanarity and D-parameter has been found. Insufficient precision on the eigenvalue calculation can cause lambda(3) to be negative, exclusively in two-jet events, yielding an underestimation of the inclusive two-jet cross section.