García-Calderón, Mateos, and Moshinsky Reply: van Dijk and Nogami [1] argue that their result for Pt t ÿ3 follows from a novel analytical expansion of r; t in the stationary solutions uk; r of the Schrö dinger equation; see Eq. (7) of Ref. [2]. They claim that the difference with our approach arises because our expansion in terms of the resonant states u n r and Moshinsky functions Mk n ; t, see Eqs. (12) and (14) of Ref. [3], is unsuitable to describe the long-time regime due to the outgoing character of resonant states at r R. They affirm, instead, that in their expansion a complex pole k n can have a positive or a negative real part, leading, respectively, to an outgoing or an incoming wave at r R. They express r; t as a linear combination involving both types of poles and the Moshinsky functions Mr ÿ R; k n ; t and obtain that at long times r; t t ÿ3=2 . By substitution of this quantity into the definition of Pt, they get the behavior mentioned above.These authors seem to ignore that poles with a positive real part, k n a n ÿ ib n , a n > 0 are related to poles with a negative real part, k ÿn ÿa n ÿ ib n , by time reversal considerations, i.e., k ÿn ÿk n , and, consequently, that both types of poles follow from a purely outgoing boundary condition [4]. Indeed, a careful reading of our Letter [3] shows that our expansion of Pt runs over the full set of poles, k n and k ÿn , as it does also the expansion of the time dependent Green's function, Eq. (9), that when substituted in Eq. (3), noting that Mk n ; t M0; k n ; t [5], leads to the expression,
