Elliptical Galaxy

What an elliptical galaxy is

An elliptical galaxy is a stellar system that looks, photographically, like a smooth featureless ellipse. There are no spiral arms, no obvious dust lanes, no bright knots of star formation. Run a long-slit spectrograph across it and you see absorption-line spectra dominated by red giants and old main-sequence stars, with very little Hα or other emission to indicate ongoing star formation. Drop a 21-cm radio receiver on it and you get little or no neutral hydrogen — typically less than 10⁹ M_sun of H I, compared with the 10¹⁰ – 10¹¹ M_sun a comparable disk galaxy would have.

The mechanical structure is also distinctive. Stars in a spiral disk move on nearly circular co-planar orbits; stars in an elliptical move on a mess of orbits at all inclinations. The velocity dispersion σ (the spread of line-of-sight stellar velocities at a point) typically dominates over any net rotational velocity v_rot. An elliptical is pressure-supported: the random motion of its stars holds it up against gravity, much as the random motion of molecules holds up a hot gas cloud.

The E0 to E7 sequence

Edwin Hubble in 1926 sorted ellipticals along an apparent-flattening axis. The Hubble class E_n is defined by the projected axis ratio b/a:

n = 10 × (1 − b/a). So E0 has b/a = 1.0 (round), E5 has b/a = 0.5, E7 has b/a = 0.3.

The sequence stops at E7. Beyond that flattening, observed galaxies have ordered rotation and a thin disk — they are classified as lenticular galaxies (S0), which form a separate branch of the morphology diagram. The E0–E7 axis is a projection axis: an intrinsically round galaxy looks E0 from every angle, but a triaxial galaxy can look E2 along one line of sight and E5 along another. Statistical inversion of large samples gives intrinsic axis ratios with typical values around b/a ≈ 0.7–0.9 and c/a ≈ 0.5–0.8, i.e. mildly triaxial.

The de Vaucouleurs profile

The surface brightness of an elliptical falls off smoothly with radius according to the de Vaucouleurs law, log I(R) ∝ −(R / R_e)^(1/4). In magnitudes per square arcsecond:

μ(R) = μ_e + 8.327 × [ (R / R_e)^(1/4) − 1 ],

where R_e is the effective (half-light) radius and μ_e is the surface brightness there. The shape — very steep central rise, slow outer fall-off — captures the bulk of normal ellipticals to within a few tenths of a magnitude. The more general Sersic profile, log I ∝ −(R/R_e)^(1/n), generalises this; ellipticals span n ≈ 2 (compact, "diE-like") to n ≈ 6 or more (massive cD galaxies with extended envelopes). Bigger ellipticals have larger Sersic n: more centrally concentrated and more extended at large radii.

Anatomy at a glance

Property	Typical value	Notes
Apparent flattening	E0 → E7	b/a from 1.0 to 0.3, projection-dependent
Sersic index n	2 – 6+	De Vaucouleurs n = 4 is typical; n grows with mass
Stellar age	8 – 13 Gyr	Old, red, on the red sequence
H I gas mass	≲ 10⁹ M_sun	Often undetected; much lower than spirals
Velocity dispersion σ	100 – 400 km/s	Pressure support; v_rot/σ < 1 typically
Stellar mass range	10⁶ – 10¹² M_sun	Dwarfs to giant cDs
SMBH mass	10⁶ – 10¹⁰ M_sun	Tightly correlated with σ (M-σ relation)
Environment	Cluster centres, group cores	Density-morphology relation: dense → elliptical

How ellipticals form

The standard story is "two-phase" hierarchical assembly. The first phase, mostly in situ, produces a compact red-and-dead galaxy at z ~ 2–3 when intense, rapid star formation in a gas-rich progenitor builds the bulk of the stellar mass and a supermassive black hole grows alongside. AGN feedback quenches further star formation. The galaxy is then dry — gas-poor — and continues to grow by minor and major dry mergers, accreting smaller systems whose stars get smeared into an extended envelope. This produces the observed size growth from z ~ 2 to z = 0 (a factor of 3–5 in R_e) at roughly fixed stellar mass.

The major-merger origin is supported by several pieces of evidence:

• Merger simulations. Toomre & Toomre (1972) showed that two disk galaxies on a parabolic encounter relax into a featureless r^(1/4) ellipsoid, with tidal tails resembling those of observed pairs.
• Tidal features. Many ellipticals show faint shells, ripples, or low-surface-brightness streams — relics of recent merger activity (e.g. NGC 1316, NGC 3923).
• Boxy/disky isophotes. The departure from perfect ellipses in the isophotes traces merger history: disky systems (a₄ > 0) tend to be intermediate-mass with mild rotation, while boxy systems (a₄ < 0) tend to be most massive with little rotation.
• SMBH coalescence signatures. The central "cores" of massive ellipticals — flat-density regions inside ~100 pc — are interpreted as scoured by the gravitational wake of merging supermassive black hole pairs.

Some canonical ellipticals

• M87 (NGC 4486). The Virgo Cluster's central giant, classified E0/cD. M_BH = 6.5 × 10⁹ M_sun (EHT 2019); halo mass ≈ 6 × 10¹² M_sun. Hosts a one-sided relativistic jet visible from radio to X-ray.
• NGC 1316 (Fornax A). A merger-disturbed S0/E galaxy in the Fornax Cluster. Multiple stellar shells, tidal arcs, and an embedded radio AGN. Quintessential dry-merger remnant.
• M49 (NGC 4472). The brightest Virgo galaxy, classified E2. Hosts the most massive globular cluster system in the Virgo Cluster.
• NGC 4889. A Coma Cluster cD with one of the most massive black holes known, M_BH ≈ 2.1 × 10¹⁰ M_sun.
• NGC 1399. Central Fornax Cluster cD, E1, surrounded by hot X-ray gas tracing the cluster potential.
• M32. A compact elliptical companion to Andromeda. Very high central density, possibly the stripped core of a more massive galaxy.

The fundamental plane

Ellipticals obey a remarkably tight three-parameter scaling relation. Defining R_e (effective radius in kpc), σ (central velocity dispersion in km/s), and ⟨I_e⟩ (mean surface brightness inside R_e), the locus of ellipticals in (R_e, σ, ⟨I_e⟩) space is essentially a plane:

log R_e ≈ 1.24 log σ − 0.82 log ⟨I_e⟩ + const, with scatter ~ 15%.

This is partly a consequence of the virial theorem (which would give exponents 2 and −1 exactly) modified by systematic variations in mass-to-light ratio, structural homology, and stellar populations. The tightness of the plane makes ellipticals strong distance indicators — independent of the Cepheid or Tully-Fisher routes — and a sensitive probe of cosmic evolution: the tilt and zero-point of the fundamental plane shift slightly with look-back time, encoding the passive ageing of stellar populations.

From dwarfs to cDs

The elliptical family spans more than five orders of magnitude in mass. Dwarf ellipticals (dEs) at 10⁷ – 10⁹ M_sun share the smooth ellipsoidal morphology but are diffuse, with low surface brightness and modest σ. They populate galaxy groups and cluster outskirts. At the high-mass end, cluster-central cD galaxies sit at the bottoms of their hosts' potential wells; they have extended diffuse envelopes (Sersic n > 6), bright X-ray haloes of hot intracluster gas, and they have accreted dozens of smaller galaxies over a Hubble time. The total stellar mass of a cD plus its intracluster light can exceed 10¹² M_sun.

Dwarf ellipticals and giant ellipticals are not on the same formation pathway. Dwarfs probably descend from gas-poor dwarf irregulars stripped by their environments; giant ellipticals are the merger end-products of disk galaxies. The morphology is similar; the histories are not.

Common pitfalls

• Inferring intrinsic shape from Hubble class. The E_n number is a projection; an E2 could be a triaxial system with intrinsic axis ratios c/a ≈ 0.7. Statistical samples are needed to deproject.
• Confusing E with S0. Lenticular (S0) galaxies look similar but contain a thin stellar disk and ordered rotation. Detailed kinematics or velocity-dispersion maps from IFU spectroscopy distinguish them.
• Assuming ellipticals have no gas. They are gas-poor in cold H I, but most host hot X-ray-emitting halos at 10⁶ – 10⁷ K, and small amounts of cold molecular gas have been detected in many systems.
• Calling them "old galaxies." Their stars are old, but the systems themselves are continually assembled — many ellipticals at z = 0 underwent major dry mergers as recently as z = 0.5.
• Treating them as purely pressure-supported. Many lower-mass ellipticals (especially "disky" ones) have detectable rotation, with v/σ approaching 0.3–0.5. The pressure-vs-rotation balance is itself a useful classifier.
• Reading the de Vaucouleurs profile as a physical model. It is an empirical fit that works because of the violent relaxation produced by mergers — there is no first-principles derivation of n = 4.