I always see galaxies in "disk" shape in pictures. It is like, there is a rotational movement on a plane and the system is balanced by the pulling-in gravitational force which prevents the stars to leap out of the galaxy. What is the physical reason behind this? I have never seen a star in disk shape. A star is nothing but a much much smaller non-solid mass revolving around an axis, just like the galaxies. Does size matter? What makes the difference? Why doesn't a typical galaxy form in a spherical shape rather than a disk shape?

enter image description here

  • 2
    $\begingroup$ Please note that even if stars and galaxies are not solid there is a enormeusly big difference: the star is a compact body of plasma, with pressure, while the galaxy is a big void of very sparse atoms with some clouds and very small stars, so it is nothing like compact. $\endgroup$
    – Envite
    Dec 12 '13 at 23:11
  • $\begingroup$ Many duplicates on Physics SE - this may be the best one physics.stackexchange.com/questions/93830/… $\endgroup$
    – ProfRob
    Mar 4 '15 at 11:01

Spiral galaxies may be the most commonly depicted probably because we find ourselves in one, however, galaxies come in many sizes and shapes. Wikipedia's page on Galaxy morphological classification explains the most common classifications.

Answering the question on the title:

The basic answer to your question is Conservation of Angular Momentum. Astronomers have been interested in this question for a long time and are still writing papers on the subject. Even if you assume that at the time of the big bang there was no net angular momentum in the universe, one would expect that there were local fluctuations. When galaxies initially formed by gravitational attraction these net local amounts remained. Now angular momentum is conserved so when the initial collection of stars and gas in one galactic volume start to collapse the net angular momentum has to stay the same. Since angular momentum $\vec{L} =m\vec{v}\times\vec{r}$, for a given $v$ its easier to conserve $L$ if the $r$ is large. In the orthogonal direction (along the axis of the accidental angular momentum) there is no such hindrance to collapse and so the collection turns out to be disk shaped. By the way there has been several studies on the net angular momentum of a large collection of galaxies coming from the Sloan Digital Sky Survey and the answer is pretty close to but not quite zero. Stay tuned.

Extracted from an University of Phoenix Department of Physics article

  • $\begingroup$ I would argue one has to also refer to shocks which cancel out momentum of the gas perpendicular to the plane; the conservation of angular momentum does not explain alone discs galaxies because elliptical galaxies also conserve angular momentum. $\endgroup$
    – chris
    Mar 15 '14 at 21:06
  • 1
    $\begingroup$ @chris I agree, this doesn't answer the question at all. The presence of gas and dissipational processes is mandatory. Many elliptical galaxies rotate. $\endgroup$
    – ProfRob
    Mar 5 '15 at 9:19
  • 1
    $\begingroup$ Much better explained on this Physics.SE answer $\endgroup$
    – kubanczyk
    Mar 17 '17 at 7:43

The mass of the galaxy (mostly in the form of dark matter) is in a roughly spherical blob. So if you look at mass, the galaxy isn't a disc, it is a spheroid. But Dark Matter is invisible, and what we can see (stars, gas etc) is in a disk.

The reason that Dark matter and the normal matter behave differently is that when gas flows there is "friction" (Dark matter doesn't interact with itself or normal matter). This causes the gas to heat up, and that heat energy is then released (as infra-red, light and so on) This means that over time the gas in the galaxy will tend to fall to a lower level. However the gas also has angular momentum (it is rotating), and angular momentum must be conserved (it can't be radiated away like energy). So the gas will try to fall into a low energy configuration that can maintain angular momentum. The shape that achieves this is a disc.

Any gas clouds that are not orbiting in the plane of the disc will hit it, and over time they will be pulled into the same disc.

Gas clouds produce stars, and so most stars will also be in the plane of the disc. Very old clusters of stars in globular clusters however can be found in a spherical pattern around the disc.

So galaxies form disc shapes because the gas that makes stars falls into a disc shape.

However, not all galaxies are discs. When disc-shaped galaxies collide, this can disturb the orbits of the stars, and you get a galaxy which is "blob" shaped, these are called elliptical galaxies, and are very common. Small galaxies also often don't have a disc structure either. These are called irregular galaxies.


Galaxies are disc shaped because they are gas rich and dynamically young. Stars are also gas rich but they are dynamically old so they have had time to rid themselves of their discs. Young protostars (which are dynamically young) are surrounded by proto-stellar discs. The reason many young gas rich objects are disc shaped has to do the fact that circular orbits are compatible with no orbit crossing, hence no shocks. In some sense, many young astronomical objects are disc shaped because they contain(ed) gas which can radiate away non circular motion.

But a disc is not the most likely state of a gravitational system: given time, torquing, instabilities or viscous processes it will tend towards a more likely compact state, where mass flows inwards and angular momentum outwards. This is why proto-stellar discs become stars. Galaxies on the other hand have not had time to turn themselves into gigantic black holes, or been given the opportunity to do so via torquing with their environment.

When two gas poor disc galaxies collide they produce an elliptical which is not disc like. When two gas rich disc galaxies collide they produce a disc like galaxy with a bulge.


Another way to think about this is to consider the ratio of angular momentum to total energy. Through shocks, if gas is present the system can radiate away some of its energy so that this ratio becomes larger. A high ratio will typically correspond to a disc-like system.

  • $\begingroup$ This is not correct. Disk galaxies stay disky unless perturbed by a major merger. You can easily have disk galaxies that are 10 Gyr old, whereas stars form on time scales of 10 kyr, i.e. 1e6 times faster. $\endgroup$
    – pela
    Mar 16 '17 at 20:45
  • $\begingroup$ @pela so stars are dynamically old. What is not correct? $\endgroup$
    – chris
    Mar 17 '17 at 6:23
  • $\begingroup$ Okay, maybe I misunderstand your term "dynamically old". If you mean "has had the time to dissipate away so much energy that it has collapsed to form a spherical object", then you're of course right (by definition). But disk galaxies don't dissipate their energy away and become spherical (unless perhaps if you wait some 1e14 yr when all stars have died. But maybe this was your point?). Mergers may make disk galaxies into sphericals, but you could still have two disks merging to produce another disk. $\endgroup$
    – pela
    Mar 17 '17 at 9:33

It's all about whether the gas (or whatever other dynamical object) can cool effectively while collapsing. If it can, you get a disk oriented perpendicular to the average angular momentum, while if it cannot, you get a spherical object.

Disks: spiral galaxies, black hole accretion disks, protostellar disks

spheres: stars (dense enough to be opaque to cooling radiation), star clusters & elliptical galaxies (stars mostly don't collide, so cannot cool), gas in galaxy clusters (so low density that the atoms don't collide often enough to cool effectively)

  • 1
    $\begingroup$ Why the heck was this downvoted? $\endgroup$
    – ProfRob
    Apr 7 '19 at 17:31
  • $\begingroup$ I would have said that stars/planets assume a spherical shape because of forces that resist gravitational collapse. It is possible that the above post is a clever physics-y way of saying the same thing, but i'm not enough of a physicist to know. It's possible the downvoters had similar thoughts. $\endgroup$
    – Ingolifs
    Apr 8 '19 at 0:02
  • $\begingroup$ Indeed, pressure resists gravity. Cooling makes pressure go away, leaving centrifugal support as the only remaining support, which gives a disk rather than a sphere $\endgroup$ Apr 8 '19 at 21:39
  • $\begingroup$ This answer is excellent and concise! I really don't get the downvotes. If it's too physics-y, they can just ask for clarification. Also, welcome Mordecai! I really hope this is not your last answer :) $\endgroup$
    – pela
    Apr 9 '19 at 10:35

Not the answer you're looking for? Browse other questions tagged or ask your own question.