The observed spectra were modelled using the ZEEMAN2 (Landstreet
1988; Wade et al. 2001) spectrum synthesis code. Chemical abundances
and projected rotational velocities v sin i were determined
by direct fitting of the synthesized spectra. For each of the five stars,
several spectral regions were examined. Uniform vertical and horizontal
abundance distributions were assumed. While a uniform horizontal
distribution is almost certainly not the case for some elements
in some stars, with only one observation per star, it is impossible to
accurately model surface distributions.