“…We choose to work, as in the paper [8], in the fixed flavour scheme, n f = 3, with Λ (4) QCD = 231 M eV , and without taking into account, in the Q 2 evolution of α s , quark thresholds. The range ]0, 1] has been divided in M steps by M + 1 points: x 0 , x 1 , ..., x M distributed in such a way that the function ln(x) + 2x varies by the same amount at any step; this function is slightly different from the pure logarithmic distribution commonly used in literature [8,9], but allow, in our case, a more uniform distribution of the numerical errors. The end points are fixed to be x 0 = 1 × 10 −8 and x M = 1; however, for a better reading, in the Figures 1−4 the x−axis ranges from 10 −4 to 1.…”