5 Results

The V, (B-V) and V, (V-I) diagrams for the stars in Table 2 are shown in Fig. 3. Stars # 10 and 27, which have very red V-I colors laying well outside the diagram borders are not plotted. The situation for these two stars looks confuse (and need additional observations to be settled), because # 10 has a B-V color that places it exactly on the ZAMS, and # 27 (which is too faint for a reliable measure on the Bframes) appears too bright in V to be a pre-ZAMS cluster member.

We started with a fit of the observed V, (B-V) diagram to the Zero Age Main Sequence (Schmidt-Kaler 1982) obtaining an apparent distance modulus m-M = 14.28 and a mean reddening E_B-V = 1.02. Then, the age was determined through trial fits to the observed main sequence with theoretical isochrones of the Padova group (Bertelli et al. 1994) characterized by a standard [He/H] ratio and solar metallicity and scaled to the distance modulus and reddening found before. The best fit, shown in Fig. 3, is obtained for the isochrone corresponding to the age of 5 10⁶ years, slightly older than the value reported by FD. The resulting age is sensitive to the position in the color-magnitude diagram of star #26 (the brightest one) and to a lesser extent of #23. Both stars lie very close to the cluster center, their spectro-photometric parallax and reddening agree with the cluster distance and extinction, and furthermore the wide field spectroscopy of Fig. 2 and Table 3 shows a marked drop of O-B stars outside the region covered by Fig. 1. Thus, it is quite safe to assume both stars as validated members of the cluster. The radial velocity of #26 is $\sim 2.5 \sigma$ off the mean of the other three cluster members in Table 5, which could suggest a binary nature. The contribution to the #26 brightness by the possible companion cannot be directly determined by the data at hand; however it is worth noticing that (a) the companion is not severly affecting the star colour, and (b) its spectral features does not show up in our Echelle spectra. Thus the position of star #26 in Fig. 3 should not be affected by a possible companion by more than a few tenths of a magnitude. Dimming star #26 by 0.25 mag would change by no more than 1 million year the age estimate, for which we can therefore assume a safe 5( $\pm 2)~10^{6}$ years. It may be of interest to note that Feinstein et al. (1986) have suggested that open clusters having stars with Of characteristics like star #26 should not be older than 5 10⁶years.

The distance to NGC 6604 is d = 1.7  kpc, for a R_V = 3.1 standard reddening law. Such a distance is 25% smaller than found by FD, which is mainly based on photographic photometry, but quite in agreement to the value derived by MV and places the cluster at a galactocentric distance 6.9 kpc, on the outer boundary of the Carina-Sagittarius arm. Adopting the ZAMS in the (B-V), (V-I) as tabulated by Munari & Carraro (1996), a ratio

$$\frac{E_{B-V}}{E_{V-I}} = 0.77$$ (10)

is found for the reddening affecting NGC 6604, in agreement with the value for the R_V = 3.1 standard reddening law. The fitting to isochrones in the V, V-I plane is shown in Fig. 3, with V - M_V = 14.28 mag and E_V-I = 1.32 as scaling factors, in agreement with the values from the V, B-V diagram. In view of the larger uncertainities affecting the transformations from the theoretical plane (log L, log T) to the observational one when red photometric bands are involved, the resulting isochrone fitting to the observational data seems satisfactory.

The spectroscopic data of Table 3 for the four stars with detailed spectral classification (# 2, 21, 23, 24) give a mean reddening of $E_{B-V}= 1.02\pm0.01$, the same determined from photometry (the star #26 has not been considered because of its emission line nature). The mean reddening from interstellar lines from Table 4 is $E_{B-V} = 1.00 \pm 0.04$. It seems noteworthy that three independent methods converge within 0.01 mag to the same $E_{B-V}= 1.02\pm0.01$ value for the reddening affecting NGC 6604.

The mean spectroscopic apparent modulus $V - M_{V}=14.13~\pm~0.17$ mag for the cluster members in Table 3 is in good agreement with the V - M_V=14.28 derived from ZAMS fitting.

Finally, the data listed in Table 5 give a cluster heliocentric radial velocity $RV_{\odot} = +20.5~\pm~2.1$ km s^-1, in agreement with the $RV_{\odot} =+19.0~\pm~3.5$ km s^-1 of Liu et al. (1991). The Hron's (1987) rotation curve gives a heliocentric radial velocity of +8.2 $\pm$ 2.5 km s^-1 at the galactic location of NGC 6604. Bearing in mind that the effect of the galactic rotation, as seen from the Sun, nearly vanishes toward the Galaxy center direction (close to which NGC 6604 lies) the resulting difference between model and observational velocities (12 km s^-1) is within the dispersion of the galactocentric radial velocities for extreme Pop I objects (12.5 km s^-1, Binney & Merrified 1998). Therefore the cluster distance and position, its radial velocity and the Hron's model for the Galaxy disk rotation appear in good mutual agreement.