Early-type Spirals Research Articles

UV to radio centimetric spectral energy distributions of optically-selected late-type galaxies in the Virgo cluster

We present a multifrequency dataset for an optically-selected, volume-limited, complete sample of 118 late-type galaxies (>= S0a) in the Virgo cluster. The database includes UV, visible, near-IR, mid-IR, far-IR, radio continuum photometric data as well as spectroscopic data of Halpha, CO and HI lines, homogeneously reduced, obtained from our own observations or compiled from the literature. Assuming the energy balance between the absorbed stellar light and that radiated in the IR by dust, we calibarte an empirical attenuation law suitable for correcting photometric and spectroscopic data of normal galaxies. The data, corrected for internal extinction, are used to construct the spectral energy distribution (SED) of each individual galaxy, and combined to trace the median SED of galaxies in various classes of morphological type and luminosity. Low-luminosity, dwarf galaxies have on average bluer stellar continua and higher far-IR luminosities per unit galaxy mass than giant, early-type spirals. If compared to nearby starburst galaxies such as M82 and Arp 220, normal spirals have relatively similar observed stellar spectra but 10-100 times lower IR luminosities. The temperature of the cold dust component increases with the far-IR luminosity, from giant spirals to dwarf irregulars. The SED are used to separate the stellar emission from the dust emission in the mid-IR regime. We show that the contribution of the stellar emission at 6.75 micron to the total emission of galaxies is generally important, from ~ 80% in Sa to ~ 20% in Sc.

The Orbits of Cluster Galaxy Populations as Evolutionary Constraints

We determine the mass profile of a synthetic cluster built from the combination of 59 nearby clusters observed in the ESO Nearby Abell Cluster Survey (ENACS). We use ellipticals and S0s as tracers of the cluster potential, and solve the Jeans equation assuming isotropic orbits. Such an assumption is justified by the analysis of the shape of the velocity distribution of ellipticals and S0s. We find that the cluster mass profile is consistent with the Navarro, Frenk and White (NFW) model. We use this cluster mass profile to search for equilibrium solutions for the other cluster galaxy populations: very bright ellipticals (M R ≤ −22 + 5 log h), early-type spirals (Sa-Sb), and late-type spirals and irregulars (Sbc-Ir), together with emission-line galaxies. We find equilibrium solutions for both the early- and the late-spirals, but not for the very bright ellipticals. The dynamics of very bright ellipticals is probably affected by dissipative processes which invalidate the use of the collisionless Jeans equation. The equilibrium solution found for the early-spirals implies them to move on nearly-isotropic orbits. Late-spirals are instead found to be on mildly radial orbits, with the radial anisotropy increasing outwards. We discuss the implications of these results for the evolutionary histories of the different populations of cluster galaxies.

Seyfert 2 Galaxies with Spectropolarimetric Observations

We present a compilation of radio, infrared, optical and hard X-ray (2-10kev) data for a sample of 90 Seyfert 2 galaxies(Sy2s) with spectropolarimetric observations (41 Sy2s with detection of polarized broad lines (PBL) and 49 without PBL). Compared to Sy2s without PBL, Sy2s with PBL tend to be earlier-type spirals, and show warmer mid-infrared color and significant excess of emissions (including the hard X-ray(2-10kev), [O {\sc iii}]$\lambda$5007, infrared (25 $\mu$m) and radio). Our analyses indicate that the majority of Sy2s without PBL are those sources having less powerful AGN activities, most likely caused by low accretion rate. It implies that the detectability of the polarized broad emission lines in Sy2s may depend on their central AGN activities in most cases. Based on the available data, we find no compelling evidence for the presence of two types of Sy2s, one type of them has been proposed to be intrinsically different from Sy2s claimed in Unification Model.

Star Formation in the [ITAL]Infrared Space Observatory[/ITAL] Atlas of Bright Spiral Galaxies

We investigate star formation along the Hubble sequence using the ISO Atlas of Spiral Galaxies. Using mid-infrared and far-infrared flux densities normalized by K-band flux densities as indicators of recent star formation, we find several trends. First, star formation activity is stronger in late-type (Sc - Scd) spirals than in early-type (Sa - Sab) spirals. This trend is seen both in nuclear and disk activity. These results confirm several previous optical studies of star formation along the Hubble sequence but conflict with the conclusions of most of the previous studies using IRAS data, and we discuss why this might be so. Second, star formation is significantly more extended in later-type spirals than in early-type spirals. We suggest that these trends in star formation are a result of differences in the gas content and its distribution along the Hubble sequence, and it is these differences that promote star formation in late-type spiral galaxies. We also search for trends in nuclear star formation related to the presence of a bar or nuclear activity. The nuclear star formation activity is not significantly different between barred and unbarred galaxies. We do find that star formation activity appears to be inhibited in LINERs and transition objects compared to HII galaxies. The mean star formation rate in the sample is 1.4 Msun/yr based on global far-infrared fluxes. Combining these data with CO data gives a mean gas consumption time of 6.4 x 10^8 yr, which is ~5 times lower than the values found in other studies. Finally, we find excellent support for the Schmidt Law in the correlation between molecular gas masses and recent star formation in this sample of spiral galaxies.

Nuclear Stellar Disks in Spiral Galaxies

We report evidence for nuclear stellar disks in three early-type spirals, namely, NGC 1425, NGC 3898, and NGC 4698, revealed by Wide Field Planetary Camera 2 F606W images out of a sample of 38 spiral galaxies, selected from the Hubble Space Telescope data archive. Adopting the photometric method introduced by Scorza & Bender, we derived their central surface brightness and scale length by assuming them to be infinitesimally thin exponential disks. No nuclear disk was found in barred galaxies or galaxies of Hubble type later than Sb. The external origin of the disk in NGC 4698 is strongly suggested by its orthogonal geometrical decoupling with respect to the host galaxy.

1.65 $\mathsf{\mu m}$ (H-band) surface photometry of galaxies

As a result of a systematic NIR H-band (m) imaging survey of normal galaxies in the local universe that includes objects both in the Virgo cluster and in the “Great Wall” (including A1367, A1656 as well as the “isolated” population in the bridge between the two clusters), we are able to measure in a highly homogeneous way photometric and structural properties for a sample of 1143 galaxies. We base our analysis on a quantitative structural parameter, the concentration index C31 (defined as the ratio between the radii that enclose 75% and 25% of the total luminosity), instead of relying on the galaxies' morphological classification. The C31 parameter provides a model independent, quantitative and continuous characterization of the light distribution within galaxies, and it is thus to be preferred to either the Hubble type or a parameter like the bulge-to-disk or bulge-to-total light ratio. Low C31 objects are typically found among disk galaxies, while high C31 describes bulge-dominated systems. We confirm our previous claim that C31 correlates strongly and non-linearly with the galaxy total luminosity. values are found only at (giant ellipticals mixed with early-type spirals), while at galaxies have (dwarf Irregulars mixed with ellipticals). At high luminosity, low C31 are allowed (bulge-less giant Scs). Thus C31 and the total luminosity are not sufficient to fully characterize the family of galaxies. However we find that galaxies can be completely described by three parameters, namely: a scale parameter (the H-band luminosity), a shape parameter (the concentration index C31) and a colour parameter (e.g. the colour). At low luminosity dEs and dIs, having similar C31, are colour-discriminated, while at very high luminosity different C31 discriminate S0s from Scs, otherwise undistinguishable on the basis of their colour. A single, monotonic relation exists between luminosity and in the H-band, as opposed to the two separate regimes that are generally observed in the B-band. As NIR luminosity traces quite accurately the galaxy mass distribution, this relation re-enforces the indication in favour of a scale-dependent mass collapse mechanism which produces higher surface-brightness and more centrally peaked galaxies with increasing mass. However, the presence of high-luminosity low-C31 galaxies hints at other machanisms and physical properties (such as angular momentum) playing an important role in galaxy formation.

Chemical Evolution of Spiral Galaxies from Redshift 4 to the Present

ISM abundances in nearby spiral galaxies are well known from HII region studies (Zaritsky et al. 1994). While early type spirals, Sa, Sb, have rather uniform abundances and a narrow range of present star formation rates (SFR) the galaxy-to-galaxy variations both in HII region abundances and in present SFR increase towards late spiral types Sc, Sd (see e.g. Kennicutt & Kent 1983). ISM abundances of spiral galaxies or their progenitors up to the highest redshifts can be studied via the absorption properties imprinted in the spectra of background QSOs. While MgII- and CIV- absorption lines are produced in the low column density gas of the extended haloes around galaxies, the Damped Lyα Absorption (DLA) is believed to originate in (proto-)galactic disks. High resolution spectroscopy of a large number of metal lines associated with DLA systems reveal the redshift evolution of ISM abundances from z ≳ 4 to z ~ 0.6.

Caltech Faint Galaxy Redshift Survey. XV. Classifications of Galaxies with 0.2 < [CSC]z[/CSC] < 1.1 in the Hubble Deep Field North and its Flanking Fields

To circumvent the spatial effects of resolution on galaxy classification, the images of 233 objects of known redshift in the Hubble Deep Field (HDF) and its flanking fields that have redshifts in the range 0.20 0.5 are seen to be more chaotic than those of their nearer counterparts. Furthermore, the spiral structure in distant early-type spirals appears to be less well developed than it is in nearby early galaxies.

The Progenitors of Subluminous Type I[CLC]a[/CLC] Supernovae

We find that spectroscopically peculiar subluminous SNe Ia come from an old population. Of the sixteen subluminous SNe Ia known, ten are found in E/S0 galaxies, and the remainder are found in early-type spirals. The probability that this is a chance occurrence is only 0.2%. The finding that subluminous SNe Ia are associated with an older stellar population indicates that for a sufficiently large lookback time (already accessible in current high redshift searches) they will not be found. Due to a scarcity in old populations, hydrogen and helium main sequence stars and He red giant stars that undergo Roche lobe overflow are unlikely to be the progenitors of subluminous SNe Ia. Earlier findings that overluminous SNe Ia (dM15(B) < 0.95) come from a young progenitor population are confirmed. The fact that subluminous SNe Ia and overluminous SNe Ia come from different progenitor populations and also have different properties is a prediction of the CO white dwarf merger progenitor scenario.

The Host Galaxy and Environment of thez = 1.195 Quasar 3C 190

We present deep imaging and spectroscopic observations of the z = 1.195 quasar 3C 190 and its immediate environment. The quasar is surrounded by an asymmetric, low surface brightness envelope in which there are also several galaxies with characteristic dimensions of 3-5 kpc. Some of these form a linear structure some 50 kpc long in projection, offset from the quasar and almost perpendicular to the radio axis. Spectroscopy of this feature indicates the presence of a starburst component, as well as stars a few hundred million years old. Velocities of the emission-line gas associated with H II regions in the linear feature are difficult to reconcile with any reasonable model that is also consistent with the observed morphology. One possible scenario is an edge-on disk in an extremely massive (~1012 M☉) halo; another is two independent chain galaxies, with the apparent alignment being fortuitous. Taking into account all of the types of extended material we find in the vicinity of 3C 190, we appear to be witnessing a relatively brief phase in the development of a spheroidal galaxy, in which merging is proceeding nearly simultaneously in a variety of different regimes. This system may be one of the clearest examples yet found for the mechanism by which many elliptical galaxies and bulges of early-type spirals form in the early universe.

Signatures of Galaxy-Cluster Interactions: Spiral Galaxy Rotation Curve Asymmetry, Shape, and Extent

The environmental dependencies of the characteristics of spiral galaxy rotation curves are studied in this work. We use our large, homogeneously collected sample of 510 cluster spiral galaxy rotation curves to test the claim that the shape of a galaxy's rotation curve strongly depends on its location within the cluster, and thus presumably on the strength of the local intracluster medium and on the frequency and strength of tidal interactions with the cluster and cluster galaxies. Our data do not corroborate such a scenario, consistent with the fact that Tully-Fisher residuals are independent of galaxy location within the cluster; while the average late-type spiral galaxy shows more rise in the outer parts of its rotation curve than does the typical early-type spiral galaxy, there is no apparent trend for either subset with cluster environment. We also investigate as a function of cluster environment rotation curve asymmetry and the radial distribution of H II region tracers within galactic disks. Mild trends with projected cluster-centric distance are observed: (i) the (normalized) radial extent of optical line emission averaged over all spiral galaxy types shows a 4%+/-2% increase per Mpc of galaxy-cluster core separation, and (ii) rotation curve asymmetry falls by a factor of two between the inner and outer cluster for early-type spirals (a negligible decrease is found for late-type spirals). Such trends are consistent with spiral disk perturbations or even the stripping of the diffuse, outermost gaseous regions within the disks as galaxies pass through the dense cluster cores.

An Investigation into the Prominence of Spiral Galaxy Bulges

From a diameter-limited sample of 86 `face-on' spiral galaxies, the bulge-to-disk size and luminosity ratios, and other quantitative measurements for the prominence of the bulge are derived. The bulge and disk parameters have been estimated using a seeing convolved Sersic r^(1/n) bulge and a seeing convolved exponential disk. In general, early-type spiral galaxy bulges have Sersic values of n>1, and late-type spiral galaxy bulges have values of n<1. In the B-band, only 8 galaxies have a bulge shape parameter n consistent with the exponential value of 1, and only 5 galaxies do in the K-band. Application of the r^(1/n) bulge models results in a larger mean r_e/h ratio for the early-type spiral galaxies than the late-type spiral galaxies. Although, this result is shown not to be statistically significant. The mean B/D luminosity ratio is, however, significantly larger (> 3-sigma) for the early-type spirals than the late-type spirals. This apparent contradiction with the r_e/h values can be explained with an iceberg-like scenario, in which the bulges in late-type spiral galaxies are relatively submerged in their disk. This can be achieved by varying the relative bulge/disk stellar density while maintaining the same effective bulge-to-disk size ratio. The absolute bulge magnitude - log(n) diagram is used as a diagnostic tool for comparative studies with dwarf elliptical and ordinary elliptical galaxies. At least in the B-band, these objects occupy distinctly different regions of this parameter space. While the dwarf ellipticals appear to be the faint extension to the brighter elliptical galaxies, the bulges of spiral galaxies are not.

Star formation in extragalactic HII regions. Determination of parameters of the initial mass function

A method for using the colors of star-forming complexes to derive the slope and upper mass limit of the initial mass function (IMF) and the age of the complex is proposed in the framework of synthetic evolutionary models of star-cluster populations. The star-formation parameters of 105 complexes in 20 spiral and irregular galaxies are determined. The IMF slopes in different star-forming complexes differ appreciably, and their dependence on the luminosities and masses of the complexes is derived. The duration of the star-formation period increases with the luminosity of the complex, and complexes with longer star-formation periods are richer in metals. The slope of the integrated IMF in a Galaxy depends on the mass spectrum of its complexes, and the upper mass limit of the IMF is lower in early-type spirals.

On the Nature and Redshift Evolution of DLA Galaxies

We extend our spiral galaxy models, which successfully describe nearby template spectra as well as the redshift evolution of CFRS and HDF spirals, to include – in a chemically consistent way – the redshift evolution of a series of individual elements. Comparison with observed DLA abundances shows that DLAs might well be the progenitors of present-day spiral types Sa through Sd. Our models bridge the gap between high redshift DLA and nearby spiral HII region abundances. The slow redshift evolution of DLA abundances is a natural consequence of the long SF timescales for discs, the scatter at any redshift reflects the range of SF timescales from early to late spiral types. We claim that, while at high redshift all spiral progenitor types seem to give rise to DLA absorption, towards low redshifts, the early-type spirals seem to drop out of DLA samples due to low gas and/or high metal and dust content. Model implications for the spectrophotometric properties of the DLA galaxy population are discussed in the context of campaigns for the optical identifications of DLA galaxies both at low and high redshift.

On the bulge-to-disk size ratio for spiral galaxies

From a detailed structural analysis of the sample of spiral galaxies from de Jong and van der Kruit (1994), we present new results on the bulge-to-disk size ratio as a function of morphological type. A Sersic r 1/n bulge and an exponential disk model, both convolved with an appropriate Gaussian PSF, were simultaneously applied to the galaxy light profiles to obtain the best-fitting bulge effective radius, r e , and disk scale-length, h. The Hubble sequence of spiral galaxies is scale-free if the ratio of these two quantities is uncorrelated with morphological type. This was in fact claimed to be the case by de Jong (1996) and also Courteau et al. (1996) using R-band data; although, applying exponential bulge models to their K-band data resulted in their mean r e / h ratio for the early-type spirals being smaller than the mean r e / h ratio for the late-type spirals at the 98% significance level. However, early-type spiral galaxy bulges are known to be better described with models having Sersic shape parameters n>1(Andredakis, Peletier and Balcells, 1995; Graham and Prieto, 1999). It was therefore of interest to re-evaluate the above claim allowing for the full range of bulge profile shapes that exist and can be described by the Sersic r 1/n model. In so doing, the mean r e / h ratio for the early-type spirals (Sa,Sab,Sb) was found to be significantly (3σ) larger than the mean r e / h ratio obtained when using the exponential bulge models. Intriguingly, the mean r e / h ratio for the early-type galaxies was discovered to be larger than the mean r e / h ratio for the late-type (Scd-Sm) spirals at only the 1.5–2σ significance level in the optical bands and at the lσ significance level in the K-band. These results suggest that the Hubble sequence for spiral galaxies is indeed scale-free. A fuller discussion is presented in Graham (2001).

The Role of Starbursts in the Evolution of Galaxies

We have determined, for the first time, the nitrogen-to-oxygen abundance ratio (N/O) for a large sample of starburst nucleus galaxies (SBNGs) and compared it to the values observed in normal spiral galaxies. The N/O ratios in SBNGs are ∼ 0.2 dex higher than in normal HII regions observed in the discs of late-type spirals, but are comparable to the values found in the bulges of normal early-type spirals. The variation of the N/O ratio as a function of metallicity follows a primary + secondary relation, but the increase in nitrogen does not appear as a continuous process. Assuming that nitrogen is produced by intermediate-mass stars, we show that our observations are consistent with a model in which the bulk of the nitrogen was formed during past sequences of bursts of star formation which have probably started 2 or 3 Gyr in the past. What we observe, therefore, could be the main production of nitrogen in the bulges of these galaxies.

The Orthogonal Bulge–Disc Decoupling in NGC 4698

The R-band isophotal map of the Sa galaxy NGC 4698 shows that the inner region of the bulge is elongated perpendicularly to the major axis of the disc. At the same time a central stellar velocity gradient is found along the minor axis of the disc. The same properties have also been recognized in the Sa galaxy NGC 4672. This remarkable geometric and kinematic decoupling is a direct indication that a second event occurred in the history of these galaxies suggesting that acquisition phenomena could play a primary role in the formation of early-type spirals.

The masses of black holes in the nuclei of spirals

We use the innermost kinematics of spirals to investigate whether these galaxies could host the massive black hole remnants that once powered the quasi-stellar object (QSO) phenomenon. Hundreds of rotation curves of early- and late-type spirals are used to place upper limits on the central black hole (BH) masses. We find that (i) in late-type spirals, the central massive dark objects (MDOs) are about 10-100 times smaller than the MDOs detected in ellipticals, and (ii) in early-type spirals, the central bodies are likely to be in the same mass range as the elliptical MDOs. As a consequence, the contribution to the QSO/active galactic nuclei (AGN) phenomenon by the BH remnants eventually hosted in spirals is negligible: BH(Sb-Im)<6 104 M Mpc 3. We find several hints that the MDO mass versus bulge mass relationship is significantly steeper in spirals than in ellipticals, although the very issue of the existence of such a relation for late Hubble type objects remains open. The upper limits on the masses of the BHs resident in late-type spirals are stringent: MBH 106-107 M , indicating that only low-luminosity activity could possibly have occurred in these objects.

The Evolution of Early‐Type Galaxies in Distant Clusters. III. \documentclass{aastex} \usepackage{amsbsy} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{bm} \usepackage{mathrsfs} \usepackage{pifont} \usepackage{stmaryrd} \usepackage{textcomp} \usepackage{portland,xspace} \usepackage{amsmath,amsxtra} \usepackage[OT2,OT1]{fontenc} \newcommand\cyr{ \renewcommand\rmdefault{wncyr} \renewcommand\sfdefault{wncyss} \renewcommand\encodingdefault{OT2} \normalfont \selectfont} \DeclareTextFontCommand{\textcyr}{\cyr} \pagestyle{empty} \DeclareMathSizes{10}{9}{7}{6} \begin{document} \landscape $M/ L_{V}$ \end{document} Ratios in the \documentclass{aastex} \usepackage{amsbsy} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{bm} \usepackage{mathrsfs} \usepackage{pifont} \usepackage{stmaryrd} \usepackage{textcomp} \usepackage{portland,xspace} \usepackage{amsmath,amsxtra} \usepackage[OT2,OT1]{fontenc} \newcommand\cyr{ \renewcommand\rmdefault{wncyr} \renewcommand\sfdefault{wncyss} \renewcommand\encodingdefault{OT2} \normalfont \selectfont} \DeclareTextFontCommand{\textcyr}{\cyr} \pagestyle{empty} \DeclareMathSizes{10}{9}{7}{6} \begin{document}

Internal kinematics, length scales, and surface brightnesses have been determined for a large sample of 53 galaxies in the cluster Cl 1358+62 at z = 0.33 from Keck spectroscopy and Hubble Space Telescope WFPC2 imaging over a 1.5 h-1 × 1.5 h-1 Mpc2 field of view. These data have been used to constrain the evolution of early-type galaxies in the cluster environment. We have constructed the fundamental plane using 30 E and S0 galaxies and draw the following conclusions. The fundamental plane at z = 0.33 has the form re ∝ σ1.31±0.13I, similar to that found locally. The 1 σ intrinsic scatter about this plane is ±14% in M/LV, comparable to that observed in Coma. We conclude that, even at intermediate redshifts, E and S0 galaxies are structurally mature and homogeneous, like those observed in nearby clusters. The M/LV ratios of these early-type galaxies are offset from the Coma fundamental plane by Δ log(M/LV) = -0.13 ± 0.03 (q0 = 0.1), indicative of mild luminosity evolution in the stellar populations in the Cl 1358+62 E and S0 galaxies. This level of evolution suggests a luminosity-weighted formation epoch for the stars of z > 1. The precise redshift depends on the initial mass function and on the cosmology. The scatter about the fundamental plane is consistent with that in the color-magnitude relation, indicating that the E/S0's have a scatter in luminosity-weighted ages of 15%. We have also analyzed the M/LV ratios of galaxies of type S0/a and later. These early-type spirals follow a different plane from the E and S0 galaxies: re ∝ σ0.66±0.29I, with a scatter that is twice as large as the scatter for the E/S0's. The difference in the tilt between the plane of the spirals and the plane of the E/S0's is shown to be due to a systematic correlation of velocity dispersion with residual from the plane of the early-type galaxies. These residuals also correlate with the residuals from the color-magnitude relation. Thus for the spirals in this cluster, as well as for the three E+A galaxies in the sample, we see a systematic variation in the luminosity-weighted mean properties of the stellar populations with central velocity dispersion. If this is a relative age trend, then luminosity-weighted age is positively correlated with σ; i.e., more massive spiral galaxies have older stars on average. The residuals from the color-magnitude relation were used to correct the surface brightnesses of the early-type spirals. After this correction for age effects, these spirals fall on the fundamental plane of E/S0's. We conclude that the early-type spirals may well evolve onto the scaling relations of the old cluster members. After correcting the spirals for the systematic trend with color, their scatter about the fundamental plane does not decrease and remains twice as large as that for the E/S0's. This large scatter should be seen in a subsample of present-day cluster galaxies, if these spirals evolve into contemporary S0's, unless some unknown process reduces their scatter. The colors and M/LV ratios imply that many cluster galaxies were forming stars at least up until z ~ 0.5, but that this activity was specific to the spiral population. The E+A's will likely evolve into low-mass present-day S0 and Sa galaxies but be slightly bluer than the present-day (B-V) color-magnitude relation by ~0.08 mag, while fading to ~1 mag below L*.

Galactic bulges from Hubble Space Telescope NICMOS observations: ages and dust

We present optical and near-infrared colour maps of the central regions of bulges of S0 and spiral galaxies obtained with WFPC2 and NICMOS on the Hubble Space Telescope (HST). By combined use of HST and ground-based data, the colour information spans a region from a few tens of pc to a few kpc. In almost all galaxies, the colour profiles in the central 100-200 pc become more rapidly redder. We attribute the high central colour indices to a central concentration of dust. We infer an average extinction at the centre of A(V)=0.6-1.0 mag. Several objects show central dust rings or discs at subkpc scales similar to those found by others in giant ellipticals. For galactic bulges of types S0 to Sb, the tightness of the B-I versus I-H relation suggests that the age spread among bulges of early-type galaxies is small, at most 2 Gyr. Colours at 1R(eff), where we expect extinction to be negligible, are similar to those of elliptical galaxies in the Coma cluster, suggesting that these bulges formed at the same time as the bright galaxies in Coma. Furthermore, the galaxy ages are found to be independent of their environment. As it is likely that Coma was formed at redshift z > 3, our bulges, which are in groups and in the field, must also have been formed at this epoch. Bulges of early-type spirals cannot be formed by secular evolution of bars at recent epochs, because such bulges would be much younger. There are three galaxies of type Sbc and later; their bulges are younger and could perhaps arise from secular evolution of transient bars. Our results are in good agreement with semi-analytic predictions, which also predict that bulges, in clusters and in the field, are as old as giant ellipticals in clusters.