Annu. Rev. Astron. Astrophys. 1999. 37: 603-648 Copyright © 1999 by Annual Reviews. All rights reserved

7.5. Summary and Key Issues

Progress on the UVX problem during the last ten years has been excellent. The theoretical, spectral, and imaging evidence has recently converged toward the view that the UVX originates from He-burning, extreme horizontal branch stars, their post-HB progeny, and post-AGB stars in the dominant, metal-rich stellar population of E galaxies. The mixture of these types apparently varies from object to object, perhaps in a systematic way with global mean metallicity or mass, but in most cases the EHB/post-EHB channels are the more important. The simplest explanation for the correlation between the UVX and optical line strengths is that the mass-loss parameter _R increases with Z or that Y / Z 2.5.

Although evolutionary synthesis models successfully predict UV spectral properties in the ranges observed, progress in understanding the UVX, and in refining estimates of ages and abundances derived therefrom, is hampered by our lack of knowledge of two basic processes: mass loss on the giant branch and helium enrichment. Both of these are critical to the efficiency with which an old population can generate UV-bright stars. We urgently require a more complete and predictive physical theory of giant-branch mass loss. This is the highest priority for UVX theory in the near term. The question of the value of the helium enrichment parameter (Y / Z) near and above solar abundance also needs to be addressed. Both areas demand extensive observational programs on nearby systems as well as fundamental improvements in theoretical modeling. The same is true of diffusion in hot atmospheres, which is important to interpreting the UVX line spectrum.

These are the most serious gaps in our astrophysical understanding of the UVX, but there are other troublesome issues as well, three of which are worth mentioning:

1. The behavior of the UVX seems to be firmly linked to that of the lighter elements such as N, Mg, and Na and decoupled from the Fe-peak (Section 5.2). This adds an additional dimension to modeling space, so far unexplored, which is not at present well supported by nucleosynthetic theory (Worthey 1998).

2. The internal spatial gradients in 1500-B color discussed in Section 4.2 do not correlate with gradients in Mg₂ (Ohl et al 1998). Metallicity is evidently not the sole parameter governing the UVX. This may be related to the decoupling of the Fe-peak noted in paragraph 1 above, or it may reflect the influence of other changing parameters within galaxies, such as age or Y abundance. M32, with a large and reversed UVX gradient (see Figure 3), is an important case since there is considerable independent evidence for an intermediate age ( 8 Gyr) population there and possibly an age gradient in which the central regions are younger (O'Connell 1980, Freedman 1992, Rose 1994, Hardy et al 1994, Faber et al 1995, Grillmair et al 1996).

3. There has been very little work on the dependence of the UVX on galaxy morphology despite suggestions of differences between E galaxies and S0 galaxies (e.g. Smith & Cornett 1982). Bright, nearby spiral bulges could readily be studied in the UV with HST, and comparisons with E galaxies could help distinguish some of the underlying drivers of the UVX phenomenon. If, for example, bulges have a wider range of ages than E galaxies (e.g. Wyse et al 1997), then the younger ones should have smaller UV upturns than E's.