Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 360, 85-91 (2000)

Previous Section Next Section Title Page Table of Contents

1. Introduction

The distributions of colours in E-type galaxies have been studied in the past from two different point of views, obviously not independent. Some authors aimed at detecting unapparent dust patterns in such objects, and to obtain statistics of their occurrence and/or estimates of the amount of dust: see for instance Sparks et al. (1985), Véron-Cetty & Véron (1988), Kim (1989). Other were more concerned with the average run of colours with radius, as Vigroux et al. (1988), Franx et al. (1989) (Fal89), Peletier et al. (1990) (Pal90), Goudfrooij et al. (1994a) (Gal94). These two topics could of course be approached from the same observations, as in Gal94 and subsequent papers by Goudfrooij et al. (1994b, 1994c), and our contributions in previous papers of this series (Michard 1998a, 1998b, 1998c, 1999).

Trends for bluer colours at larger radii in E-type galaxies have long been known from aperture and surface photometry, and were generally attributed to gradients of the average metallicity, as evidenced by systematic changes of metallic line indices (see for instance the review by Kormendy & Djorgovsky, 1989). The Fal89 and Pal90 surveys introduced the U-R colour index, and demonstrated the generality of large outwards negative gradients in this colour, more than twice the corresponding gradients in B-R. Pal90 showed these gradients to be "consistent with a decrease in the [Fe/H] of approximately 0.20 per decade in radius" to quote their paper.

In Gal94 the B-V and B-I gradients were measured for a complete magnitude limited sample of 56 E-galaxies. It seems certain from far infrared observations, that E-galaxies contain more dust than apparent from optically seen patterns. Goudfrooij & de Jong (1995) estimated the amount of diffuse dust from IRAS data, and then used a model of Witt et al. (1992) to infer the B-I colour gradients induced by this dust. These were finally compared with measured gradients in Gal94 and found to be similar or smaller.

Another approach was taken by Wise & Silva (1996), using calculated colour gradients from their own theory of transfer in various models of dusty ellipticals on the one hand, and a consideration of the ratios of available U-R and B-R gradients on the other. They conclude that "measurements are not precise enough in many objects to provide tight constraints on the models".

Perhaps we might comment at this stage that population gradients exist in E-galaxies as proven by line spectra, while dust is certainly also present. It is a merit of the theoretical work quoted above to prove that dust could indeed be much more abundant than estimated from apparent patterns (absent or quite elusive in many cases!) and probably play some role in broad band colours distributions. It is however unrealistic to neglect population changes in the discussion of colour observations, the more so when a metallicity sensitive index such as U-B is considered.

The purpose of the present work is to reconsider the discussion of colour gradients in ellipticals, using simultaneously the 4 optical colours U-B, B-V, B-R, V-I, rather than only one pair. The sample is therefore made up, as a first step, of objects in common to the Gal94 sample and the Fal89+Pal90 samples. This was supplemented with hitherto unpublished observations obtained at the Observatoire de Haute Provence. Some data from Bender & Möllenhof (1987) (BM87) were also used. The number of objects with complete 4 colours data is 26; a few other with only 3 colours available were also considered in the discussion.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: July 27, 2000