Arvind Sivasubramanian has detected some typographical errors in the description
of
the Langevin algorithm, Section 2.2: The factor 24 in the amplitude
of the stochastic
force should be 6, and the factor of mass should not appear. Moreover,
in the
velocity update the factor h**2 should be h.