Hubble Tuning Fork

Reading the fork at a glance

Point a telescope at enough galaxies and a pattern jumps out before you measure anything: most fall into a handful of recognisable shapes. Some are featureless luminous blobs, brightest in the middle and fading smoothly outward. Others are flat, rotating pinwheels laced with bright arms. And a large fraction of those pinwheels have a straight bar of stars slicing through the centre. Edwin Hubble's genius in the 1920s was to notice that these shapes line up naturally into a branching diagram shaped like a tuning fork.

The single straight handle on the left holds the ellipticals, ordered by how squashed they look — round E0 at the far left, increasingly flattened up to E7. Where the handle ends, the diagram forks. The upper prong runs along the ordinary spirals Sa → Sb → Sc; the lower prong runs along the barred spirals SBa → SBb → SBc. At the junction where all three meet sits the lenticular class S0 — a disk that looks like a spiral stripped of its arms. Anything that refuses to fit gets dumped in the irregular bin, Irr.

The power of the fork is that a galaxy's position encodes several physical properties at once. Move from the handle toward the tips of the prongs and you go from old, gas-poor, slowly-rotating spheroids to young, gas-rich, rapidly-rotating disks bright with new stars. The diagram is, in effect, a low-dimensional map of how stars and gas are arranged in the universe's most common large structures.

A short history of a misleading diagram

Hubble published the scheme in 1926 ("Extra-galactic nebulae", Astrophysical Journal) and gave it its famous fork shape in his 1936 book The Realm of the Nebulae. He had only recently — in 1923–24, using Cepheid variables in M31 — proven that the spiral "nebulae" were entire galaxies far outside the Milky Way, so classifying them was the natural next step. He borrowed the vocabulary of stellar spectral classification, calling the smooth ellipticals "early-type" and the open, knotty spirals "late-type."

That borrowing was a lasting mistake. In stellar astronomy "early" and "late" had once carried an (also wrong) evolutionary connotation, and readers transplanted the idea to galaxies, imagining that an elliptical gradually flattens, sprouts a disk, and unwinds into an Sc. We now know morphology does not flow that way along the fork. The terms "early-type galaxy" (E and S0) and "late-type galaxy" (spirals and irregulars) survive purely as convenient shorthand — they describe shape, not age or destiny.

The handle: ellipticals E0–E7

Ellipticals are smooth, featureless spheroids of mostly old, red stars with little cold gas or dust and almost no ongoing star formation. Hubble graded them by their projected flattening with a single integer:

type = E n,   where  n = round[ 10 × (1 − b/a) ]

  a = projected semi-major axis
  b = projected semi-minor axis

  E0:  b/a ≈ 1.0   (looks circular)
  E7:  b/a ≈ 0.3   (flattest observed elliptical)

An important subtlety: this is the apparent flattening, not the true three-dimensional shape. A genuinely flattened elliptical viewed face-on projects as a near-circle and gets logged as a low-n type. So the En classes scatter a real galaxy across several bins depending on viewing angle, and population studies must correct for this projection statistically.

The surface-brightness profile of an elliptical is described remarkably well by the de Vaucouleurs R^(1/4) law (1948), a special case of the more general Sérsic profile:

I(R) = I_e · exp{ −b_n [ (R/R_e)^(1/n) − 1 ] }

  n = 4    →  classic de Vaucouleurs elliptical
  n = 1    →  exponential disk (spiral)
  R_e      =  effective (half-light) radius
  b_n ≈ 2n − 0.327

The single parameter n — the Sérsic index — turns the qualitative E-versus-spiral distinction into a number: bulge-dominated systems cluster near n ≈ 4, pure disks near n ≈ 1. Giant ellipticals span an enormous mass range, from dwarf ellipticals of ~10⁷ M_☉ to the cD monsters at cluster centres exceeding 10¹³ M_☉, with half-light radii from under a kiloparsec to tens of kiloparsecs.

The prongs: spirals, bars, and the Sa–Sc ordering

Spirals are flattened, rotation-supported disks with a central bulge and trailing spiral arms outlined by young blue stars, HII regions and dust lanes. Moving along a prong from "a" to "c" is governed by three correlated criteria that Hubble laid out:

1. Bulge-to-disk ratio. Sa galaxies have a large, dominant bulge; by Sc the bulge is a small kernel and the disk dominates the light.
2. Arm winding (pitch angle). Sa arms are tightly wound (small pitch angle, ~5°); Sc arms are loose and open (pitch ~25–30°).
3. Arm resolution. Sa arms are smooth; Sc arms break into bright clumps of star-forming knots.

The lower prong, SBa → SBb → SBc, repeats this ordering for galaxies with a prominent stellar bar — a straight, rotating concentration of stars through the nucleus that funnels gas inward along its leading edges. Bars are not rare oddities: in optical light roughly one-third of disk galaxies are strongly barred and another third are weakly barred, and in the near-infrared (which sees through dust to the old stellar population) the barred fraction approaches two-thirds. The Milky Way itself is a barred spiral, classified SBbc.

The disk light follows an exponential profile, I(R) = I_0 · exp(−R/h), where h is the disk scale length — typically 3–5 kpc for an L* spiral. Disk stars orbit at flat rotation speeds of ~150–300 km/s, evidence of the dark-matter halo that the visible disk is embedded in.

Morphology by the numbers

Type	Shape	Bulge/disk	Cold gas	Stellar pop.	Sérsic n	Example
E0–E7	Spheroid	All spheroid	Very low	Old, red	≈ 4	M87, M49
S0 (lenticular)	Disk, no arms	High	Low	Old, red	3–5	NGC 5866, NGC 3115
Sa / SBa	Tight arms	Large bulge	Moderate	Mixed	~2–3	M104 (Sombrero), M81
Sb / SBb	Moderate arms	Medium bulge	Higher	Mixed	~1.5–2	M31 (Andromeda), Milky Way
Sc / SBc	Loose, clumpy arms	Small bulge	High	Young, blue	~1	M101 (Pinwheel), M33
Irr	No symmetry	None	Very high	Young, blue	—	LMC, SMC, NGC 4449

The trend down the rows tells a coherent story: gas fraction and the proportion of young blue stars both rise from ellipticals to irregulars, while the prominence of an old central bulge falls. This is why ellipticals glow uniformly red while late spirals sparkle with blue, recently-formed stars.

Beyond the fork: the de Vaucouleurs extension

Hubble's fork is two-dimensional and a little coarse. In 1959 Gérard de Vaucouleurs generalised it into a richer system that classifies a galaxy along three independent axes:

• Stage — the position along the sequence, extended past Sc to Sd, Sdm, Sm and Im (Magellanic irregulars). Encoded as a numerical T-type from T = −6 (compact elliptical) through T = 0 (S0/a, with S0 itself near T = −2) to T = +10 (Im).
• Family — the strength of the bar: SA (unbarred), SB (barred), and the intermediate SAB.
• Variety — the presence of an inner ring (r), an s-shaped flow into the arms (s), or both (rs).

Under this scheme the Milky Way reads roughly as SAB(rs)bc or SBbc — a moderately barred spiral of intermediate stage. The T-type makes morphology a continuous number that correlates cleanly with physical quantities: gas fraction, colour, and specific star-formation rate all increase smoothly with T, which is why modern surveys often quote T rather than a single Hubble letter.

What actually sets a galaxy's morphology

If the fork is not a timeline, what physics decides where a galaxy lands? The modern answer is a mixture of initial conditions and environment:

• Mergers. A "major merger" of two comparable gas-rich spirals violently scrambles their ordered disks into a pressure-supported spheroid — making an elliptical. Simulations by Toomre & Toomre (1972) and successors show this is the principal channel for building giant ellipticals. So the real transformation runs spiral → elliptical, the opposite of a naive left-to-right reading.
• Gas supply and angular momentum. A galaxy that keeps accreting cold, high-angular-momentum gas can sustain a star-forming disk and stay a spiral. Cut off the supply and star formation fades.
• Environment. The morphology–density relation (Dressler 1980) is one of the strongest in extragalactic astronomy: the fraction of ellipticals and S0s rises steeply with local galaxy density, while spirals dominate the low-density field. In rich clusters, ram-pressure stripping and high-speed "harassment" can quench an infalling spiral, converting it into a gas-poor S0.
• Internal secular evolution. Bars and spiral density waves redistribute angular momentum over billions of years, slowly growing pseudo-bulges and rearranging gas — a quieter, internally driven path that does not require a collision.

Where you've already seen the fork

• M87 (E0/E1). The giant elliptical at the heart of the Virgo Cluster, ~16.4 Mpc away, host of a 6.5×10⁹ M_☉ black hole and the first imaged by the Event Horizon Telescope. A textbook handle-end elliptical.
• The Sombrero, M104 (Sa). A nearly edge-on early spiral with an enormous bulge and a striking dust lane, ~9.5 Mpc away — the picture most people imagine when they hear "Sa."
• Andromeda, M31 (Sb). Our nearest large neighbour at 0.78 Mpc, a classic intermediate spiral roughly 1.5× the Milky Way's stellar mass, on a collision course with us in ~4–5 Gyr.
• The Pinwheel, M101 (Sc). A grand-design late spiral ~6.9 Mpc away, with a tiny bulge and sprawling, clumpy arms lit by HII regions — the visual archetype of the prong tip.
• The Large Magellanic Cloud (Irr / SBm). A satellite of the Milky Way ~50 kpc away, lopsided and bar-bearing — once filed as a pure irregular, now recognised as a disrupted Magellanic barred spiral.

Common misconceptions and edge cases

• "Early means young, late means old." Exactly backwards. "Early-type" ellipticals are dominated by old stars; "late-type" spirals are forming new ones. The words describe sequence position, a historical accident, not stellar age.
• "E7 is just a very thin elliptical." There is a hard physical wall: no relaxed elliptical is observed flatter than E7. Past that flattening a self-gravitating stellar system needs rotational support and becomes a disk — an S0 or spiral. The fork's geometry encodes this.
• "Lenticulars are a kind of elliptical." No — S0s have a flat, rotation-supported disk like spirals; they simply lack arms because they lack the cold gas to form them. Structurally they are disk galaxies that happen to be red.
• "The classification is objective." Visual typing is subjective and depends on image depth, resolution and observer. Two experts routinely disagree by one stage. Modern quantitative measures — Sérsic n, concentration, asymmetry, Gini–M20 — were developed precisely to make morphology reproducible.
• "It works at all redshifts." The clean fork describes the present-day universe. At z > 1, when the universe was young, regular spirals and ellipticals were rare; galaxies were clumpy, irregular and merger-dominated. The Hubble sequence largely assembled over the last ~8 billion years.
• "Bars are permanent labels." Bars form, dissolve and can re-form over a galaxy's life as the disk's stability changes, so the SA/SB/SAB distinction is a current state, not a fixed identity.