Migrating sql/ds to Oracle

I'm migrating a C-application using SQL/DS (IBM/VM) to Oracle under Unix. Could anybody explain that :
I have an sql-table T1 with a column C1 declared as

        C1 char not null ,
and other columns.
Table T1 has been filled by specific Oracle tools, out of the application.  C1 contains uppercased letters.

I want to get the value of C1 for one row of that table, in C-written application:
...
EXEC SQL BEGIN DECLARE SECTION
  char loc_for_C1;
   ....
EXEC SQL END DECLARE SECTION
   ...

EXEC SQL SELECT C1

         INTO   :loc_for_C1
         FROM   T1
         WHERE  where-clause 

Sqlcode is then correct, but loc_for_C1 contains '\0';

Is this construction forbidden under Oracle?

 Am I obliged to declare loc_for_C1 as char loc_for_C1[2] ? I don't want.  Any explication or help ?
 Thanks

-- 
Philippe Olivero                       e-mail pho_at_cc.in2p3.fr
Centre de Calcul de l'In2p3          
29, Bld du 11 Novembre 1918
69622 Villeurbanne-Cedex , France