Must a star belong to a galaxy, or could it be completely isolated?

In case it can be isolated (not belong to a galaxy), could it have a planet orbiting around it?

  • $\begingroup$ There could even be a planet without a star If there are Earth size bodies drifting in space could we identify them? $\endgroup$ Feb 19, 2019 at 18:48
  • $\begingroup$ No, @JamesJenkins, it couldn't. But this is nothing to do with the underlying physics or history of the universe. It is simply a matter of definition: Wikipedia says A planet is an astronomical body orbiting a star or stellar remnant that is [...]. It has a discussion and links to official definitions. The original use of the word did not even include Halley's Comet. This can only be because its orbit was too long or eccentric as it would have met the other criteria at the time (that covered everything from the sun to Saturn) $\endgroup$ Feb 19, 2019 at 22:39
  • 3
    $\begingroup$ @DavidRobinson en.wikipedia.org/wiki/Rogue_planet $\endgroup$ Feb 20, 2019 at 6:54
  • $\begingroup$ Not sure why you ask this in physics -- astronomy had the same (main) question not too long ago, with an intriguing image attached to this answer. But it's noteworthy that these events are closely linked to galaxies because you do need more matter than usually present in intergalactic space, and that comes from galaxies. $\endgroup$ Feb 20, 2019 at 13:00
  • $\begingroup$ @anand_v.singh Your point is so pertinent that I have raised a new question: What orbit does a planet have to have to be a planet? $\endgroup$ Feb 20, 2019 at 15:27

2 Answers 2


No, stars do not need to be inside a galaxy. It is estimated that about 10% of stars do not belong to a galaxy [1]. While most intergalactic stars formed inside a galaxy and were ejected by gravitational interactions, stars can form outside of galaxies as well [2].

We assume that such stars could have planets, just like stars in a galaxy, although no specific examples have been detected yet.

[1] "Detection of intergalactic red-giant-branch stars in the Virgo cluster", Ferguson et al. Nature 391.6666 (1998): 461.

[2] "Polychromatic view of intergalactic star formation in NGC 5291", M. Boquien et al. A&A, 467 1 (2007) 93-106.

  • 3
    $\begingroup$ I like how this shows both other answers are wrong. I would have thought star formation only happened in galaxies. $\endgroup$
    – JollyJoker
    Feb 19, 2019 at 10:24
  • 2
    $\begingroup$ If a galaxy would be a prerequisite of star formation, how would the first stars formed before galaxies existed? $\endgroup$
    – mg30rg
    Feb 19, 2019 at 13:36
  • 4
    $\begingroup$ It's worth noting that the example cited, [en.wikipedia.org/wiki/NGC_5291](NGC 5291) is a pair of interacting galaxies. During such events large amounts of galactic gas/dust is drawn out of galaxies and compressed via tidal interaction triggering star formation. That can be considered intergalactic in that such clouds of gas and stars formed in them can be given sufficient velocity to never rejoin the merging galaxies; but also isn't in the sense that it's something that only happens in the immediate vicinity of galaxies. $\endgroup$ Feb 19, 2019 at 15:55
  • 3
    $\begingroup$ I'm not aware of any star formation occurring in deep intergalactic space without any galaxies in the area to serve as a trigger. $\endgroup$ Feb 19, 2019 at 15:55
  • 3
    $\begingroup$ Possibly relevant: D100 gas tail $\endgroup$ Feb 19, 2019 at 21:00

They probably have to be formed in a galaxy. But they can escape. There are many rogue planets that are found outside of solar systems. NASA has observed many rogue stars as well. This usually happens when galaxies collide and they throw out some of the stars. And yes, the rogue stars can have planets orbiting them

  • 2
    $\begingroup$ Looking at the other answer (the one with sources), stars probably don't have to be formed in a galaxy. $\endgroup$
    – hyde
    Feb 19, 2019 at 7:23
  • $\begingroup$ @hyde it goes against Kaiser bias, i.e. the likelihood of passing a threshold for gravitational collapse is enhanced if you are sitting on top of a lower density region (e.g. a proto galaxy). So stars probably form in the vicinity of other stars = a galaxy. $\endgroup$
    – chris
    Feb 19, 2019 at 12:56
  • 4
    $\begingroup$ @chris "Any given star was probably formed inside a galaxy" is a much weaker claim than "Stars probably have to be formed inside a galaxy" $\endgroup$
    – Taemyr
    Feb 19, 2019 at 13:57
  • $\begingroup$ @Taemyr stars like to form together, and in sufficient number this defines what a galaxy is. So yes stars like to form in galaxies. Doesn't mean its impossible for a star to form alone in the middle of nowhere. Its probably just exponentially rare. Having said that I don't think we know the initial power spectrum properties on stellar scales so this is hypothetical. $\endgroup$
    – chris
    Feb 19, 2019 at 14:01
  • $\begingroup$ @chris: Exponentially rare doesn't jibe with the 10% figure given elsewhere $\endgroup$ Feb 19, 2019 at 15:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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