MOLPRO patch 96.2 --- 3/4/96
============================

In order to install the patch, change to the base molpro96 directory and
issue the commands

gunzip < <DIRECTORY>/patch.96.2.gz | /bin/sh
make quicktest

followed by

make install

when you are happy everything works properly.

Changes in this version include:

Fixed bugs in make install
Support for up to 1000 basis functions
Lowered pair truncation threshold in MRCI
Enhanced matrix multiplication routines
Fixed bug in angstrom/bohr conversion in output of XYZ files
Fixed bug in basis function library (Al av5z set)
Updated benchmark directory
Input of XYZ format geometry
Change in format of density matrix storing in MCSCF
Various compilation and other minor problems fixed