Globular Clusters: Cities of Ancient Suns

If you have ever swept through Cassiopeia or the Pleiades, you have seen an open cluster — a loose, ragged knot of a few hundred young stars, drifting through the galactic disk. Open clusters are stellar daycares: they hatch, hold together for a hundred million years or so, and then shear apart as they orbit the galaxy.

A globular cluster is the opposite of that, in almost every respect.

Where an open cluster is sparse, ragged, and young, a globular is dense, spherical, and ancient. The center of M13 has stars packed roughly a thousand times closer together than the neighborhood around the Sun — close enough that, on a planet near that core, the night sky would be saturated with stars brighter than Sirius. There would be no real darkness.

Open the interactive star map and turn on the Milky Way setting; the open clusters trace the band of the Milky Way. The globulars don't. They are scattered across the whole sky, with a heavy concentration toward Sagittarius — toward the galactic center.

That is the first clue to where they live. Globulars orbit the halo of the Milky Way: a roughly spherical volume centered on the galactic core, extending well above and below the disk where the Sun and the open clusters sit. Their orbits are long, eccentric, and steeply inclined. A typical globular plunges through the disk at intervals of a few hundred million years, then climbs back into the empty halo on its long ellipse.

The reason for the spherical distribution is the reason globulars are so old: they formed in the chaotic early Milky Way, when the proto-galaxy was still a roughly spherical cloud, before gravity and angular momentum had flattened most of the gas into a disk. The disk is the universe's accountant — neat, organized, full of recent business. The halo is the attic, full of antiques.

Every globular has the same broad architecture: a dense core where the stars are crushed together, surrounded by an extended halo of stars that thins gradually outward. But how concentrated the core is varies dramatically from cluster to cluster.

The American astronomer Helen Sawyer Hogg adopted a system, originally introduced by Harlow Shapley, that ranks globulars on a Roman-numeral scale of central concentration:

• Class I — extreme central condensation, almost stellar-looking core (M75, NGC 6388)
• Class V — strong but resolvable central crush (M13, M3, M5)
• Class VIII — loose, easily resolved across the whole face (M71, M22)
• Class XII — barely a cluster at all, more like a rich open swarm (Pal 5, NGC 5466)

This matters at the eyepiece. An 80 mm refractor will shred a Class VIII cluster like M22 into a glittering swarm but reduce a Class I object like M75 to a tight fuzzball. Aperture buys you the core. The rest of the cluster is mostly resolvable in any decent scope.

A few of the most extreme globulars have undergone core collapse — gravitational interactions in the densest cluster centers slingshot lighter stars outward and let the heaviest fall inward, until the core becomes a tiny, ultra-dense knot. M15, M30, and M70 are textbook examples. In photographs the very center looks like a bright unresolved point, almost like a star superimposed on the cluster.

A globular cluster is a closed laboratory in which 12 billion years of stellar evolution have run to completion. Almost every massive star is gone — supernova, neutron star, white dwarf. What is left is a population of stars all near the same age, all from the same low-metallicity gas, but very different masses and life stages.

That uniformity is what makes globulars indispensable to astronomers. Plot every star in a globular on a Hertzsprung–Russell diagram and the picture is gorgeous: a cleanly defined main sequence, a sharp turnoff where the most massive surviving stars are just leaving it, a beautiful subgiant branch, a wide red giant branch, a horizontal branch, and a wispy asymptotic giant branch above. The position of the main-sequence turnoff is essentially a clock — it tells you the age of the cluster to within a billion years.

Among that orderly stellar population are some genuine oddities:

• RR Lyrae variables. These are old, low-mass stars on the horizontal branch, pulsating with periods of about half a day and amplitudes near a magnitude. Globulars are full of them — M3 alone has more than 200. Their absolute brightness is almost the same from one to the next, which makes them excellent standard candles: measure the period, derive the brightness, compare to the apparent brightness, and you have the distance.
• Blue stragglers. Stars that look too blue and too young to belong on the cluster's main-sequence turnoff. They are most likely the result of stellar collisions or mass transfer in close binaries — two old stars that merged and started over. In the crowded core of a globular, near-misses and mergers happen.
• X-ray binaries and millisecond pulsars. Dynamical encounters between stars and stellar remnants are so frequent in cluster cores that neutron stars get dragged into close binaries with companion stars, accrete material, and spin up to millisecond rotation rates. 47 Tuc and Terzan 5 are famous millisecond-pulsar factories — Terzan 5 alone hosts more than 40 known.

Globular cluster stars are some of the oldest objects we can put under a spectrograph, and almost the first thing the spectrograph reports is that they are chemically primitive. Their atmospheres are missing most of the heavy elements that ordinary stars take for granted.

Astronomers measure that with metallicity — the fraction of a star's mass made of elements heavier than helium, usually expressed as the iron-to-hydrogen ratio [Fe/H] on a logarithmic scale where the Sun is zero. A typical Milky Way globular sits between [Fe/H] = −1 and [Fe/H] = −2.3: one-tenth to one-two-hundredth the Sun's iron content. The stars formed before generations of supernovae had seasoned the galactic gas with carbon, oxygen, and iron. They are running on hydrogen and helium plus a sprinkle of everything else, and their spectra look correspondingly clean — fewer absorption lines, less metal "blanketing" of the blue end of the continuum.

That single number, [Fe/H], does several jobs at once.

• It dates the cluster. Low metallicity is a chemical fossil signature. The metal-poorest globulars formed when the universe itself had barely been enriched — within the first billion years after the Big Bang. The richer the cluster, the more recent the gas it formed from.
• It tells you where the cluster was born. Milky Way globulars split into two clean populations: a metal-poor halo set ([Fe/H] < −1, the classical ancient halo objects) and a metal-rich bulge/thick-disk set ([Fe/H] ≈ −0.5 to solar) that formed later in the denser inner galaxy. M13, M3, M5, M15, M92 are firmly metal-poor halo. 47 Tucanae at [Fe/H] ≈ −0.7 is the showpiece of the metal-rich population.
• It tells you whether the cluster is even ours. Several Milky Way globulars are chemical outliers — wrong metallicity for their orbit, wrong age. They are almost certainly accreted from dwarf galaxies the Milky Way ate. Omega Centauri's enormous internal metallicity spread (from [Fe/H] = −2.0 to −0.5) is one of the strongest pieces of evidence that it is the surviving nucleus of a captured dwarf rather than a real globular at all.

There is one more chemical surprise hiding inside almost every massive globular: multiple populations. High-resolution spectroscopy in the last twenty years has shown that the stars in a single cluster do not all share identical light-element abundances — most clusters host two or three sub-populations with anti-correlated sodium and oxygen, suggesting a second generation of stars formed from gas polluted by the first. The exact mechanism is still debated, but the implication is striking: a globular cluster is not a single starburst but a brief, internal multi-generational story playing out before the gas ran dry.

For most of history, astronomers assumed the Sun sat near the center of the Milky Way — partly out of cosmic vanity, partly because the disk's dust hides the true galactic core from view. In 1918, Harlow Shapley, working at Mount Wilson Observatory, broke that assumption with one elegant idea.

He noticed that globular clusters are not evenly distributed around the sky. They cluster heavily in the direction of Sagittarius. Then he used the RR Lyrae stars in the globulars as standard candles — measure the period, you know the absolute magnitude; compare with apparent magnitude, you get the distance. He worked out the distance to dozens of globulars and found that they form a roughly spherical cloud whose center is not on the Sun, but tens of thousands of light-years away in Sagittarius.

That offset is the distance from the Sun to the center of the Milky Way. Shapley overshot the modern figure by a factor of two (he didn't yet know about interstellar dust dimming his stars), but his fundamental answer was right: we live in the suburbs.

Around 150 globulars are known in our galaxy, but a small number do almost all the heavy lifting at the eyepiece. Here are the ones every observer should learn by heart.

Northern hemisphere headline acts

M13 — the Great Hercules Cluster is the northern showpiece. Magnitude 5.8, 22,000 light-years away, hosting roughly 300,000 stars. Naked-eye from a dark site as a fuzzy spot between η and ζ Herculis. In a 100 mm refractor it is a granular powderpuff. In a 250 mm Newtonian under good seeing it dissolves into thousands of pinpricks across roughly 20 arcminutes — a sight worth a clear night by itself.

M5 in Serpens — at magnitude 5.6 it is brighter than M13 and arguably more beautiful, with an asymmetric, slightly ragged outer halo and a tight, brilliant core. Sits high in northern summer skies. Many seasoned observers prefer it to M13.

M3 in Canes Venatici — magnitude 6.2, similar brightness to M13, with an even denser core. Famous for its rich population of more than 200 RR Lyrae variables — if you photograph M3 several nights running, individual stars in the cluster visibly pulse.

M15 in Pegasus — magnitude 6.2, one of the densest globulars known, with a textbook core-collapsed center. Look for the bright stellar pinpoint in the middle — that is not a foreground star, that is the cluster's collapsed core. M15 also harbors Pease 1, a small planetary nebula embedded in the cluster: the only easy planetary inside a globular and a serious challenge target for large amateur scopes.

M92 in Hercules — magnitude 6.4, often unfairly overlooked because M13 is right next door. Tighter and more concentrated than M13, with a strikingly compact core that resolves later as you increase aperture.

Southern hemisphere giants

If you can travel south, the prize globulars of the entire sky await.

Omega Centauri (NGC 5139) is in a class of its own. Magnitude 3.7 — easily naked-eye, even from a suburban site, looking like a slightly fuzzy fourth-magnitude "star." In binoculars it is unmistakable; in any telescope it is jaw-dropping. Ten million stars, three times the mass of the next-largest Milky Way globular, almost certainly the surviving nucleus of a galaxy our Milky Way ate. The full angular size is 36 arcminutes — bigger than the full Moon.

47 Tucanae (NGC 104) — magnitude 4.0, the second-best globular in the sky, only narrowly behind Omega Cen and arguably more elegant. Sits in the foreground of the Small Magellanic Cloud, an absurd photographic juxtaposition. A core so bright that an 80 mm refractor shows a sharp, almost stellar nucleus.

M22 in Sagittarius — magnitude 5.1, the brightest globular accessible from mid-northern latitudes (it just clears the southern horizon from central Europe). Looser than M13, with stars that begin resolving at very modest apertures. The closest bright globular at 10,400 light-years.

NGC 6752 in Pavo — magnitude 5.4, the fourth-brightest globular overall. Easily resolved, with elegant chains of stars curving away from a bright core.

The most common question observers ask about globulars is: what aperture do I need to "resolve" the cluster into stars? It depends on three things at once: the cluster's distance, its concentration class, and your seeing.

A useful rule of thumb:

Three observing tips that genuinely help:

1. Use medium-to-high magnification. Globulars look "fuzzy" at low power because the stars overlap into a glow. Crank to 150–250× and the halo stars suddenly snap into individual points — the cluster goes from a smudge to a city. The Difficulty Matrix on each catalog page reports detection thresholds, but resolution is a separate threshold and almost always benefits from more magnification.
2. Wait for the cluster to be high. A globular at 20° altitude is being seen through three times more atmosphere than one at zenith. Sagittarius globulars from northern latitudes never get high; the same M22 that struggles to resolve from Munich is a postcard from southern Spain.
3. Use averted vision on the core. In any cluster denser than Class V, the inner core is sometimes invisible to direct vision but pops into stars when you look slightly off-axis. The fovea is mostly cones; the rod-rich outskirts of the retina see the dim individual core stars better.

A globular cluster does not look like much in a single photograph compared with a galaxy or a glowing nebula. But hold the picture in your head: a sphere of three hundred thousand stars, older than almost everything, plunging on a 100-million-year orbit through a galaxy that did not exist when it formed. The same swarm wheeled past a different sun while the dinosaurs were dying. It will still be wheeling past, mostly unchanged, when the Sun is a white dwarf.

Charles Messier cataloged the bright globulars in the 1770s as smudges to be ignored — non-cometary nuisances on the way to real discoveries. We have since discovered that those smudges are the best fossils we have of the early universe. Every clear summer night, the oldest stars in our galaxy are above your roof, waiting for you to look.