********>Bugfix.40:
Author: Dave Case
Date: 03/17/2005

Programs: sander.QMMM

Description: When the cap water option is requested, there is no energy 
             computed to go along with the force being applied.  This can
             lead to an apparent lack of energy conservation.

Fix:  apply the following patch to amber8/src/sander/forcecp.f

------------------------------------------------------------------------------
*** forcecp.f	2004/02/17 17:19:07	7.13
--- forcecp.f	2005/03/17 16:45:52
***************
*** 53,59 ****
     logical do_list_update
     _REAL_ r_stack(*)
     
!    _REAL_ enmr(3),devdis(4),devang(4),devtor(4),entr
     
     dimension itrp(3,nt_lim)
     _REAL_ x(*),f(*),ene(30),ener(*),vir(*)
--- 53,59 ----
     logical do_list_update
     _REAL_ r_stack(*)
     
!    _REAL_ enmr(3),devdis(4),devang(4),devtor(4),entr,ecap
     
     dimension itrp(3,nt_lim)
     _REAL_ x(*),f(*),ene(30),ener(*),vir(*)
***************
*** 315,321 ****
     
     !     ----- CALCULATE THE CAP FORCE IF NECESSARY -----
     
!    if(ifcap > 0) call capwat(natom,x,f)
        vir(1) = vir(1)+0.5d0*molvir(1,1)
        vir(2) = vir(2)+0.5d0*molvir(2,2)
        vir(3) = vir(3)+0.5d0*molvir(3,3)
--- 315,324 ----
     
     !     ----- CALCULATE THE CAP FORCE IF NECESSARY -----
     
!    if(ifcap > 0) then
!       call capwat(natom,x,f,ecap)
!       ene(20) = ene(20) + ecap
!    end if
        vir(1) = vir(1)+0.5d0*molvir(1,1)
        vir(2) = vir(2)+0.5d0*molvir(2,2)
        vir(3) = vir(3)+0.5d0*molvir(3,3)
  
------------------------------------------------------------------------------

Workarounds: Don't worry, be happy....Amber has been like this forever.