The authors pointed out the disagreement between their calculation of dynamical correlations and that obtained using the exact result as in [2] can be attributed to two of following reasons: (i) The linear ansatz overestimates the higher coefficients and, given the nonoscillatory growth rate of coefficients with respect to n, suppresses the oscillation [1] in a0(t) thus damping it out faster. (ii) The inverse Laplace transform calculation is typically accurate to 10-4 or so unless a higher accuracy is specifically being sought [1]. © 1994 The American Physical Society.