2 Preparation

Before applying the procedures to be described below, one must decide which published data should be adopted and what initial corrections to those data should be made. Taylor (1998a, hereafter Paper I) has considered these issues in some detail, so only a brief summary of his discussion will be given here. For each datum to be included, there is pertinent information that can be found only in the source papers. Those papers are therefore consulted; no attempt is made to rely on Cayrel de Strobel et al. (1997) alone. Almost all the initial adjustments that are then applied to the data are of the following kinds:

1.

corrections to a single temperature scale,

2.

non-LTE corrections, and

3.

corrections to a preferred system of solar equivalent widths.

These corrections are not all applied to every datum, but are instead applied as required. (The rationale for adopting this list of corrections is discussed in Sects. 5 and 6 of Taylor 1998a.)

Many published data are given relative to the Sun. However, some data are given relative to other stars instead. Zero-point stars that have often been used include $\epsilon$ Vir, $\alpha$ Boo, $\beta$ Gem, and the Hyades giants. Mean values of [Fe/H] for these stars are given by Taylor (1998b, hereafter Paper II), who includes a detailed discussion of the way those mean values are derived.