Quantum tunneling and classical barrier reduction for a mesoscopic spin

I show that for a large spin with uniaxial anisotropy it is, in principle, impossible to distinguish classical energy-barrier reduction due to a transverse magnetic field from an increase in the tunneling rate of excited levels. Given a suitable definition of the quantum energy barrier, I derive an expression for the field dependence of the barrier height that agrees with the classical result up to a constant of order unity. Numerical results show that for a mesoscopic spin $\left(S\sim 10\right)$ the barrier decreases in a series of steps, which recent experiments may have revealed.