Comment on "Hole digging in ensembles of tunneling molecular magnets"

Tupitsyn et al. [Phys. Rev. B v69, 132406 (2004)] have recently reported results for the relaxation of crystalline systems of single--molecule magnets, such as Fe_8. They claim that, quite generally, (1) the magnetization and hole widths of field--distributions evolve with time t as sqrt{t}, and (2) the holes' line shapes are Lorentzian. We give a counter--example to these conclusions, and show that the main assumption on which they rest is invalid.