1996A&A...311..425B

In this paper we address the question whether age and metallicity effects can be disentangled with the aid of the broad-band colours and spectral indices from absorption feature strengths, so that the age of elliptical galaxies can be inferred. The observational data under examination are the indices H_β and [MgFe] , and the velocity dispersion {SIGMA} for the sample of galaxies of Gonzales (1993), supplemented by the ultra-violet data, i.e. the colour (1550-V), of Burstein et al. (1988). The analysis is performed with the aid of chemo-spectro-photometric models of elliptical galaxies with infall of primordial gas (aimed at simulating the collapse phase of galaxy formation) and the occurrence of galactic winds. The galaxy models are from Tantalo et al. (1995). The study consists of four parts. In the first one, the aims are outlined and the key data are presented. In the second part, we summarize the main properties of the infall models that are relevant to our purposes. In the third part we present the detailed calculations of the spectral indices for single stellar populations and model galaxies. To this aim, we use the analytical relations of Worthey et al. (1994) who give index strengths as a function of stellar parameters. In the last part, we examine the age-metallicity problem. In contrast with previous interpretations of the H_β and [MgFe] data as a sort of age sequence (Gonzales 1993), we find that the situation is more complicate when the space of the four variables H_β, [MgFe] , (1550-V), and {SIGMA} is examined. Galaxies in the H_β and [MgFe] plane do not follow a pure sequence either of age or metallicity. The observed (1550-V) colours are not compatible with young ages. Basically, all the galaxies in the sample are old objects (say as old as 13/15Gyr) but have suffered from different histories of star formation. Specifically, it seems that some galaxies have exhausted the star forming activity at very early epochs with no significant later episodes. Others have continued to form stars for long periods of time. This is perhaps sustained by the analysis of the gradients in the H_β and [MgFe] indices across the galaxies. There are galaxies with no age difference among the various regions. There are other galaxies in which large gradients in the mean age of the star forming activity between the central and the peripheral regions seem to exist. The nucleus turns out to be younger and more metal-rich than the outer regions. Finally, there are galaxies in which the nucleus is older but less metal-rich than the external regions. All this perhaps hints not only different histories of star formation but also different mechanisms of galaxy formation difficult to pin down at the present time. From the analysis of the H_β, [MgFe] , (1550-V), and {SIGMA} space, and of the age and metallicity gradients in single galaxies, the suggestion is advanced that the overall duration of the star forming activity is inversely proportional to the velocity dispersion {SIGMA} (and perhaps galactic mass).