“…Indeed, one of the assumptions made in [11,Theorem 1.4], namely assumption (H3) there, demands that the principal symbol p = p 1 +ip 2 of the operator in question should be such that ∂ α p 1 = O(1), |α| ≥ 2. This assumption is not satisfied in the magnetic Schrödinger case, for p given by (1.3) and indeed, a crucial role in our proof of Theorem 1.2 is played by the recent work [4] of the first-named author, where resolvent estimates of subelliptic type are established for the operator P in (1.9), in an unbounded parabolic neighborhood of the imaginary axis, away from a bounded region near the origin.…”