On the Validity of Perturbative Evolution of Structure Functions from Low Q²

The validity of using QCD perturbation theory to generate dynamically the parton distribution functions of hadrons, starting from a valencelike input at low $Q^2$, is discussed. In particular, we consider the prescription of Barone {\em et al} who evolve from $Q^2 = 0$, and that of Gl\"uck {\em et al} who start evolution from $Q^2 \simeq (2\Lambda_{QCD})^2$.