Hi,

 I was performing NPT simulation of a water - solid interface and used
Berendsen Semi-Isotropic coupling. The solid is a SAM - 11 atoms long and 
packed together closely to form a surface in the x-y plane.

   
Pcoupl              =  Berendsen
Pcoupltype          =  semiisotropic            ; pressure geometry
tau_p               =  0.5  ; p-coupling time
compressibility     =  1e-15  4.5e-5
ref_p               =  1.0  1.0 ; reference pressure

The simulation runs and the box doesn't distort nor does the X-Y area
change. However, when I calculate the pressure using g_presure, this is
what I get 

Energy         Average       RMSD     Fluct.      Drift  Tot-Drift
-------------------------------------------------------------------------------
Pressure-(bar) 10002.5    270.246    266.208 -0.0806168   -161.234
Pres-XX-(bar)  15670.1    538.695    534.178  -0.120572   -241.143
Pres-YY-(bar) 14335.4    451.628    447.136  -0.110053   -220.106
Pres-ZZ-(bar) 1.03256    238.912    238.824 -0.0112258   -22.4517

Why is the pressure so different from the reference pressure in the
lateral directions.  Is this the pressure in the solid or in water?