When the stars are plotted on a colour-colour diagram, in which one colour indicates effective temperature and the other luminosity, the lower envelope of the stars constitutes the location of the non-evolved stars (ZAMS). Figure 5 (click here) shows various ZAMS defined for the early, intermediate and late photometric regions together with the stars which are members of the clusters in Table 9 (click here). For the early and late regions empirical ZAMS were taken from Crawford (1978, 1975, 1979). We refer to Balona & Shobbrook (1984) for the comparison of several ZAMS for the early region. For the intermediate region, Strömgren (1966) placed the ZAMS at r=0. We modified by hand this relation in order to take into account the continuity with the contiguous regions. Table 11 (click here) shows the ZAMS adopted for the intermediate region.
Table 11: ZAMS defined for the intermediate region
Figure 5: Observational ZAMS for a) early, b) intermediate and c) late regions.
Dots are the members of open clusters in Table 9
Effective temperatures and surface gravities of the stars were
derived from intrinsic colour indices through
Napiwotzki et al.'s (1993) code based on the grids of Moon & Dworetsky (1985)
for solar abundance covering the range 6500-30000 K in 
and
2.5-4.5 in log g.
The grids were built from the stellar atmosphere models of Kurucz (1979).
In line with Jordi et al. (1994), the correction to
log g proposed by Napiwotzki et al. for early type stars was not applied.
Comparison of the values obtained for early type stars with
Moon (1985) and Castelli (1991) codes based on the same grids
were discussed in Jordi et al. (1994). They found that the
systematic differences with determinations of 
from Geneva
photometry were removed using Napiwotzki et al.'s (1993) code, but
that the systematic differences in log g remained.
New grids based on revised stellar atmosphere models (Kurucz 1991)
are currently being built (Smalley 1995). Preliminary results
show the same discrepancy in the surface gravity for early type stars.
The observational ZAMS translated onto the plane (
, log g)
using the methods described above
are shown in Fig. 6 (click here) with the stars belonging to the
sample of open clusters.
Theoretical ZAMS of stellar evolutionary models reported by
Schaller et al. (1992) and Schaerer et al. (1993) at different metallicities are also plotted.
Figure 6: Observational and theoretical ZAMS plotted with the open
cluster star members of Table 9. Thin lines represent the
theoretical ZAMS proposed by Schaller et al. (1992) and
Schaerer et al. (1993) at different metallicities.
The thick line represents observational ZAMS (see text)
A comparison of different ZAMS was presented by Jordi et al. (1994) for the O-B region and a good agreement was found between the theoretical and Crawford's (1978) observational ZAMS up to 20000 K. The differences in log g for intermediate and late A-type stars (regions 2 and 3) were of 0.1 dex, well within the uncertainty of the log g determinations. As for the early region, good agreement was also reported.
At lower temperatures, the observational ZAMS is a good lower envelope of
open cluster stars, but it shows a more pronounced slope
than the theoretical one, even taking into account different metallicities.
The observational ZAMS is placed on the limits
of the validity range of the grids. The determination of 
and log g may be uncertain,
but whether this is the reason for the discrepancy,
or whether it is due to a real difference between observational and theoretical
ZAMS is a matter that requires further research.
The conclusions drawn from this comparison are similar if the theoretical models by Claret & Giménez (1992) are considered instead of those of Schaller et al. For a comparison among recent stellar evolutionary models see Asiain et al. (1997).