Comment on "Short-range magnetic interactions in the spin-ice compound Ho2Ti2O7"

In their recent communication (Phys. Rev. B 64, 060406(R) (2001)) Cornelius and Gardner have reported the results of magnetization and susceptibility studies on single crystals of the pyrochlore Ho2Ti2O7. The observed field dependence of magnetization is totally unexpected, as it seems to suggest that the magnetic moments in this compound do not obey the 'ice-rule'. We have re-measured the magnetization curves for Ho2Ti2O7 single crystal for the three principal directions of an applied magnetic field and found nearly perfect agreement with the predictions for a nearest-neighbor spin-ice model.