[molpro-user] save/relaod orbital domains?

[Berger Raphael]

Dear Molpro-users,

for some reason I'm not able to understand the text in the
manual regarding the loading and saving of orbital domains
for a lmp2 calcluation. I want to calculate the interaction energy 
contributions of the correaltion energy of a (dicationic)dimer:
First at long distance, save the orbital domains and then redo the 
calculations for shorter distances with the same monomer orbitals.
Somehow I seem to confuse the directive/command realms. Here is my input:

------------------------------------------------
***
memory,400,m
angstrom;
orient,noorient;
symmetry,nosym;
geometry={
Ag1;
Ag2,Ag1,r;
}
set,charge=2
basis={Ag=VDZ-PP;
set,dfmp
default=avdz/mp2fit
s, Ag , 0.15353000000E-01
c, 1.1, 1
p, Ag , 0.20861200000E-01
c, 1.1, 1
d, Ag , 0.35359100000E-01
c, 1.1, 1
f, Ag , 0.57147400000E-01
c, 1.1, 1
g, Ag , 0.16825000000
c, 1.1, 1
h, Ag , 0.63331600000
c, 1.1, 1
i, Ag , 1.1927000000
c, 1.1, 1
}
    r=5.0
    gdirect
    hf

    {df-lmp2,scsgrd=1,nocheck,df_basis=dfmp;
     enepart,6,1;
     local,save=6000.2;
    }

dist=[3.1,3.0,2.9]
do i=1,#dist
    r=dist(i)

    gdirect
    hf

    local,start=6000.2;
    {df-lmp2,scsgrd=1,nocheck,df_basis=dfmp;
     enepart,6,1;
    }
enddo

---------------------------------------------------

any help is exteremly appreciated!

best regards
 	Raphael