Correction author: David Pearlman
Correction Date: 8/91

Bug: When SANDER was run with NTT=5 (i.e. requesting separate solute/
     solvent temperature scaling), the relaxation constant for scaling
     of the solvent was not being properly set (not being set at all, in
     fact).

Consequences: Runs performed with separate solute/solvent temperature
     scaling in SANDER might be unreliable. For such runs already performed,
     one should examine the values of T_VENT (reported at each energy
     report step) and ensure they seem reasonable.

Cause: The relaxation constant for the solvent temperature scaling was
     not being set, except when NTT=2. But it should have been set when
     NTT = 5.