MCSCF problem

[Dr. Pablo Wessig]

Dear MOLPRO users,
I have a serious problem using the MCSCF program multi. In order to learn
the correct handling I compared the results of CAS-run's with those
obtained with Gaussian. In most cases I obtained the same results but in
few cases, there are remarcable differences in the energies. Now, I'm
supposing that anything is wrong in my MOLPRO inputs. Here the MOLPRO and
the corresponding GAUSSIAN input:

MOLPRO:

memory,25,m
basis=6-31G*;cartesian
geomtyp=xyz
geometry{
14 
TS1
C          -0.616370   -0.264260    0.535170
C           0.169250    0.906140   -0.822230
C           1.391020   -0.066360   -0.496130
C           0.730040   -0.968980    0.607170
C          -0.388130    1.212850    0.551940
C          -1.516640   -0.652080   -0.420550
H           2.229380    0.525790   -0.139700
H           1.702880   -0.618460   -1.371890
H           0.676180   -2.010800    0.320060
H           1.195790   -0.899420    1.582540
H           0.321050    1.547130    1.299090
H          -1.532130   -1.654450   -0.811450
H          -1.279520    1.821410    0.548800
H          -2.335700   -0.014130   -0.693580}
rhf
multi;occ,26;closed,18;rotate,20.1,18.1,0;rotate,19.1,17.1,0;
rotate,25.1,35.1,0;rotate,26.1,36.1,0;    !gave 231.76455347

GAUSSIAN:

%chk=TS1.chk
# hf 6-31G(d) nosymm  scf=nosp

 orbitlale mit geo TS1-prop2-ca8-new1b

 0 1
 C                    -0.61637  -0.26426   0.53517 
 C                     0.16925   0.90614  -0.82223 
 C                     1.39102  -0.06636  -0.49613 
 C                     0.73004  -0.96898   0.60717 
 C                    -0.38813   1.21285   0.55194 
 C                    -1.51664  -0.65208  -0.42055 
 H                     2.22938   0.52579  -0.1397 
 H                     1.70288  -0.61846  -1.37189 
 H                     0.67618  -2.0108    0.32006 
 H                     1.19579  -0.89942   1.58254 
 H                     0.32105   1.54713   1.29909 
 H                    -1.53213  -1.65445  -0.81145 
 H                    -1.27952   1.82141   0.5488 
 H                    -2.3357   -0.01413  -0.69358 

--link1--
%chk=TS1.chk
%nosave
# cas(8,8) 6-31G(d) nosymm guess=(read, alter) geom=check
scf=(maxcycle=300, nosp)

TS1, gave -231.7734761

0 1

20,18
19,17
25,35
26,36

Does anybody know what could be the difference of these two runs ?
Thanks for any help.
Pablo Wessig

__________________________________________________

  Priv.-Doz. Dr. Pablo Wessig                           
  Institut fuer Chemie                       
  Humboldt-Universitaet zu Berlin            
  Hessische Str. 1-2                         
  D-10115 Berlin                             
  Tel. 030/20937 275                         
  Fax. 030/20936 940                        
  e-mail:  Pablo=Wessig at rz.hu-berlin.de      
  www:     http://141.20.79.96/wessig

  Please do not use the following e-mail address:
  wessig at pro122lin.chemie.hu-berlin.de        
___________________________________________________