Jmol alignment incorrect if PDB file has alternate CA positions

Spun off JAL-1622.
Some PDB files have alternate positions for the CA in some residues, e.g. 1BB9 residues 30, 38, 81.
ATOM 143 CA AHIS A 30 23.595 0.410 2.617 0.75 26.68 C
ATOM 144 CA BHIS A 30 23.606 0.385 2.631 0.25 32.57 C
(.75 and .25 are the respective occupancy rates for the two positions)

This results in an incorrect structure alignment in Jmol (e.g. align 1PHT to 1BB9):
- the atomspec for the residue matches both CA atoms
- as a result the resolved list of atoms to match in the structure is longer than intended
- Jmol just aligns the two lists of atoms in order, resulting in a shift of the alignment after the 'dual CA'
e.g.
 atom 1 [ASP]31:A.CA/2.1 #162 atom 2 [TYR]14:A.CA/1.1 #117 ASP is misaligned to TYR
 atom 1 [HIS]30:A.CA%B/2.1 #144 atom 2 [ASP]13:A.CA/1.1 #108 second CA 'aligns' to ASP
 atom 1 [HIS]30:A.CA%A/2.1 #143 atom 2 [TYR]12:A.CA/1.1 #94 correct alignment
 atom 1 [GLN]29:A.CA/2.1 #133 atom 2 [LEU]11:A.CA/1.1 #85 correct alignment