An error in ORTEP has been discovered by Dr. Louis Farrugia, the author
of Ortep-3 for Windows.  The error dates back to the 1976 release of
OR TEP-II (ORNL-5138 manual).  The result of this error is a reversal 
of handedness of chiral molecules in ORTEP drawings when BOTH of the
following occur:

  1. All atoms in the atoms list are related by n-fold symmetry
     with n > 2.

  2. A 506 instruction (inertial axes coordinate system) is used.

This situation gives a product moment matrix with two equal
eigenvalues.  When this occurs, the coding error generates a
left-handed coordinate system that leads to the incorrect handedness of
the ORTEP drawing.  Thus, the drawing does not accurately depict the
true atomic coordinates input to ORTEP.

We point out that this error is most significant when absolute
configuration is being displayed and state again that it only occurs
when both conditions listed above are met.  The fact that it took
23 years for this error to be discovered suggests the rarity of these
conditions in practice.

To extend a famous quote by W. H. Zachariasen - "Don't trust your
theoretical models, don't trust your experimental results, [don't trust
your computer programs,] and don't take yourself or your work too
seriously."

Version 1.0.3 (Jan. 31, 2000) of ORTEP-III corrects the error described
above.  It may now be downloaded from ftp.ornl.gov in pub/ortep or from
our web site http://www.ornl.gov/ortep.  Check with the authors of
software packages that incorporate ORTEP for the status of their
programs.

Carroll K. Johnson and Michael N. Burnett
Oak Ridge National Laboratory
Oak Ridge, Tennessee USA
e-mail: ortep@ornl.gov