By Phil Plait | July 27, 2012 10:00 am

Globular clusters are some of the most stunning objects in the sky. Composed of hundreds of thousands of stars, over 150 of these compact beehives orbit our Milky Way galaxy alone. Some are close enough that even through a small telescope they reveal a breathtaking beauty, individual stars sparsely distributed in their outskirts becoming more cramped and crowded until they blur into a generalized smear in the middle.

When you use a bigger telescope to look at them, you get wondrous beauty:

[Click to massively apiaryenate, and you really, really want to.]

This picture is from Adam Block, using the 0.8 meter Schulman Telescope on Mt. Lemmon in Arizona, and shows M5, one of my all-time favorite globulars (I posted a Hubble image of this cluster a while back, too). It’s located in the constellation of Serpens, visible over the entire populated region of Earth at this time of year. I spent many nights finding this cluster both when I was younger and also when I was in graduate school and teaching a lab class in observational astronomy. It’s a piece of cake to find since it’s bright and big. In fact from very dark skies it’s just visible to the naked eye, one of the few that are.

This picture from Adam reveals it in its glory. It’s roughly 150 – 200 light years across and 25,000 light years away, and you really get a sense of its hundred thousand stars (or more). Globular clusters like this are very old – 12 billion years or so. Any star in M5 with a mass of more than the Sun would have died long ago, turning into a red giant, blowing away its outer layers, and becoming a faint, hot, white dwarf. Even some lower mass stars have become red giants – you can pick them out pretty easily in the picture – and only stars with significantly lower mass are still chugging along, merrily fusing hydrogen into helium in their cores like the Sun does.

The cluster also contains over a hundred "blue straggler" stars – stars which are surprisingly blue given their age. These were a mystery for decades, but it’s now understood that they’re the result of binary stars, where one star is feeding off the material from another, and also from actual stellar collisions, where two stars physically collide and merge! Collisions like that are so rare they’re non-existent in the volume of space our Sun occupies, but globular clusters have a lot of stars tightly packed into a small region. Collisions are more common, and can result in a star that looks much younger than it really is.

There are so many reasons to love a good globular. I enjoy their relative simplicity – they’re only made of stars! – and structural symmetry, and how they can be observed with everything from your eyeballs to the most sophisticated telescopes on or above the Earth. And if you read about them, their physical history, composition, and orbits, you’ll find there is a vast amount to know about them, and a vast amount yet to learn. To my scientifically curious mind, that makes them just about the perfect object for study.

And, of course, they are surpassingly beautiful. That never hurts.

Image credit: Adam Block/Mount Lemmon SkyCenter/University of Arizona

Related Posts:

Desktop Project Part 12: The galactic deep diver M9
Desktop Project Part 3: The massive massiveness of M54
Gorgeous globular hides hundreds of rejuvenated stars
Lonely sentinel of the galaxy

CATEGORIZED UNDER: Astronomy, Pretty pictures

Comments (23)

  1. biologist

    How about renaming Cape Canaveral to Cape Woebegone until the budget is restored? That would mark the further decline: it was Cape Kennedy as long as the Apollo moon program lived, and then named Cape Canaveral. With these budget cuts, Cape Woebegone is the right name!

  2. biologist

    sorry, wrong comment section, can the Bad Astronomer delete this?

  3. KC

    M5 is not all that far from better known M13, another globular cluster in Hercules. M13 is fairly easy to find – its located in Hercules’ armpit!

  4. Would it be possible to search for small Kuiper belt objects and estimate their abundance and size distribution by pointing an array of several medium-sized telescopes at a suitably positioned globular cluster and observing occultations? You’d need multiple telescopes so you could confirm that a star blinking out was due to something in space and not a bird or something, and you’d need a lot of computing power and a lot of patience. Or are the individual stars in globular clusters too dim for this kind of observation?

  5. five.eagles

    Fantastic picture. I don’t know a whole lot about globular clusters, but I wonder, do they rotate at all around a galactic central point? Do the individual stars just move around randomly but are still tied together in a “glob” by gravity? (I suppose I could use Google to get the answer to these questions, but discussions can be fun!)

  6. Jaz

    Have any planet hunters pointed their gadgets at some clusters? I have a hunch Kepler might not be too useful for that, but what about the rest?

  7. jackd

    they’re only made of stars!

    I get that there aren’t any large gas or dust clouds in clusters, but are there planetary nebulae and supernova shells, just too faint to see at this distance?

  8. Funny… I’ll be TA’ing an Astro lab up here at ASU this semester … Sure we’ll be spying this one from the top of the physics building one night!

  9. James Evans

    That was my question too, jackd!

  10. Jess Tauber

    Any ancient civilizations (is there even enough ‘metal’ lying around for planets, and life to form here?) in one of these things might either have to hightail it to the main galaxy, or perhaps they are the ones making blue stragglers, like adding wood to the campfire.

  11. Steve Sittig

    I set up a calculation assuming a diameter of 150 light years and a star count of 200,000, and found that there would be about 2.56 light years between stars if evenly distributed. Of course, the stars are definitely NOT evenly distributed, getting much more dense toward the center of the glob, but I thought it interesting that the average separation is not much different than ours from Proxima Centauri.

    Assuming 200 ly diameter of glob and 100,000 stars, I calculate a separation of 4.3 ly, right on the Sol-Proxima Centauri distance.

    I’d love for someone to confirm or correct my calcs. Also, I’ve always wanted to see an artist’s conception of the night sky from a planet deep in a glob. Of course, it may be too turbulent for solar systems to exist in there.

  12. @ ^ JessTauber asked : “is there even enough ‘metal’ lying around for planets, and life to form here?”

    Sadly, that’s very unlikely since these Globular clusters have some of the lowest metallicities in the Galaxy because they’re amongst the most ancient metal-poorest stars. :-(

    @6. Jaz :

    Have any planet hunters pointed their gadgets at some clusters? I have a hunch Kepler might not be too useful for that, but what about the rest?

    Yes. I’m pretty sure there’s been several searches of globular clusters for planets that have been unsuccessful – although one very strange planet – PSR B1620-26 b a.k.a. the Genesis* or Methuselah planet – was discovered orbiting a pulsar / white dwarf binary in Messier 4.

    @7. jackd :

    I get that there aren’t any large gas or dust clouds in clusters, but are there planetary nebulae and supernova shells, just too faint to see at this distance?

    Yes. Planetary nebulae almost certainly given stellar evolution and the number and age of stars in the globs. Supernova shells from white dwarf superniovae type Ia seem likely although I’m not aware of any supernovae occurring in globulars that we’ve seen. No supernovae shells from SN types Ib, Ic (stripped core collapse from Wolf-Rayet stars) or type II supergiant supernovae at least none for a very lo-oong time indeed because the high mass stars here are very long gone.

    Great photo. :-)

    * Not to be confused with the Genesis planet from Star Trek esp. the movies II, III! 😉

  13. For more info on exoplanet searches and planetary nebulae in globular clusters see :

    Discussing an HST search for Hot Jupiters in 47 Tucane which came up empty.

    It seems from the Centauri Dreams’ article linked to my name here that the same
    team also searched the huge globular cluster (and probable ex-dwarf galaxy!) Omega Centuari with identical negative results.

    As for planetary nebulae in globulars, see :

    Good online article with photos about the Pease1 Planetary nebula in the globular M15. This notes :

    The pinkish object to the upper left of the cluster’s core is a gas cloud surrounding a dying star. Known as Kuestner 648, this was the first planetary nebula to be identified in a globular cluster. In 1928, F. G. Pease, working at the 100-inch telescope of California’s Mount Wilson Observatory, photographed the spectrum of K 648 and discovered the telltale bright emission of a nebular gas cloud rather than a normal star. In the ensuing 70 years, only three more planetary nebulae have been discovered in globular clusters.

    Plus there’s this :

    with a great astrophoto of a planetary nebulae in NGC 1846 although it isn’t 100% sure – or I gather from the comments- whether this planetary nebulae is an actual cluster member or not.

    Web searching has also noted that Planetary Nebula GJJC1 in Globular Cluster M22 is apparently a very tough challenge for amateur astronomers to observe.

  14. Jess Tauber

    Perhaps, then, these globulars might be good hideouts, holes in the wall, as it were, for aliens who have gotten on the bad side of some of their neighbors. If it were me, though, and I had the kind of advanced tech needed for such an endeavor, I’d probably just give my species a thorough biomolecular and morphological makeover so nobody recognized us. A good memory wipe as well and you’re home free.

  15. BigBob

    “Collisions like that are so rare they’re non-existent in the volume of space our Sun occupies”

    I wonder if you meant space our galaxy occupies?


  16. @ ^ BigBob : I reckon so! Otherwise we’re in an improbable amount of trouble! 😮

    Incidentally, does anyone else think Anser et vulpeculae – “the Fox and the Goose” – is the closest thing among the 88 constellations to a pub name? 😉

    @13. Jess Tauber : “Perhaps, then, these globulars might be good hideouts, holes in the wall, as it were, for aliens who have gotten on the bad side of some of their neighbors?”

    Hmm .. yeah perhaps. If they can find or make themselves a world to hide in there! 😉

    @10. Jess Tauber : Now where did that second comment of yours suddenly come from interrupting my ^ symbol dangnabbit! Sigh.

    Oh well, guess it all make sense kinda anyhow.

    As for your suggestion there, well it sounds like a Stephen Baxter / Isaac Asimov / Brinford type idea. Dunno if you plan on writing it but you could well do. Neat thought.

  17. D’oh! Wrong thread. Thought we were talking about this one :

    That & my overtiredness explains it. Sorry. :-(

  18. Artor

    I’ve always wondered about the gravitational dynamics in a globular cluster. It’s easy to see how a spiral galaxy can maintain a stable configuration for billions & billions of years, but how does a glob stay globular without collapsing into a common center? Or it it doing just that, hence the prevalence of “blue stragglers.” For that matter, how does an elliptical galaxy keep it’s figure?

  19. James Evans

    …because they’re amongst the most ancient metal-poorest stars.

    Once we perfect ion drives, or plasma thrusters, or wormholes, or whatever, we’ll have to send Sepultura there for the M5 Globular Cluster Tour. That should raise the level of m/etal in the galactic neighborhood sufficiently.

  20. Jon Hanford

    @7 jackd:

    “I get that there aren’t any large gas or dust clouds in clusters, but are there planetary nebulae and supernova shells, just too faint to see at this distance?”

    There are four planetary nebulae known to be associated with globular clusters in our galaxy –

    Pease 1 in M 15:

    GJJC 1 in M 22:

    JaFu 1 in Palomar 6 &
    JaFu 2 in NGC 6441:

    (more info here: )

    I don’t know of any supernova remnants (aside from pulsars) associated with globular clusters, but as MTU notes above, they along with their massive progenitors are likely long gone.

  21. TMB

    @Artor: In general, the random velocities of the stars are enough for them to maintain the shape… you could imagine that you take a rotating disk galaxy and tilt the orbit of each star by a random amount. Each star is then still on a stable orbit, but instead of a flat disk you have a spherically-symmetric cluster. (that’s not exactly what happens – the stars aren’t typically on circular orbits – but it gives you kind of an idea).

  22. Regner Trampedach

    TMB @ 21 and Artor @ 18: Most of the stars are on highly elliptical orbits, randomly oriented in space, giving the spherical distribution of stars. The orbits going almost radially towards the core are called plunging orbits. So each star in its orbit, feels the total gravitational potential from all the other stars. There are enough stars to make this potential smooth and very close to constant in time. In such a situation the stellar orbits would all be stable, in that each star cannot change its total energy (kinetic + potential energy).
    You need close encounters between stars, where they can feel each others gravity, more than they can the total cluster field. But just two stars colliding (in physics, objects don’t need to touch to have undergone a collision – as long as their mutual forces [gravity or electric] change their course, they are said to have collided) won’t change much. That is because the stars can’t exchange angular momentum, L, since they are spherical (L = vector{r}dotvector{v}= the dot-product of the stars position relative to the center of the cluster, with its velocity – in circular orbit it is just radius of orbit times orbital velocity. L is a conserved quantity, just like energy).
    In a collision between one star and a binary system, on the other hand, angular momentum can be exchanged between the orbit of the binary and the cluster orbit of the intruder – sometimes resulting in a closer (harder) binary system and the intruder in a higher energy orbit – sometimes thrown out of the cluster. In this way a cluster can slowly evaporate, while its core gets denser.
    As the cluster orbits the galaxy (highly elliptical randomly oriented – just like the stars in the cluster. That’s why the Milkyway’s halo of globular clusters is spherical like the clusters themselves) it will most likely pass through the plane of the galaxy (where gravity is large and changes sign), stripping the most loosely bound stars by tidal forces (different gravity felt on opposite sides of the cluster).
    BigBob @ 15: “Collisions like that are so rare they’re non-existent in the volume of space our Sun occupies”
    The BA was right here. Our neck of the Milkyway has a density of stars that is so low that collisions are exceedingly rare – you’d be better of getting a lottery ticket than waiting for two stars to collide the next billion years… Space is big!
    Why did you think the BA meant “…space our galaxy occupies.”?

    I hope that answered some questions – and sorry it got so long…

  23. amphiox

    It was my understanding that planetary nebulae and supernova remnants don’t last very long on astronomical time scales, before the gases making them up disperse so much that they are no longer recognizable. So one would expect them to be quite rare in things as old as globular clusters?

    Supernova remnants especially so, since, the stars originally in the cluster big enough to produce supernovae would surely have all exploded billions of years ago.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar