back to article You wait ages for a sun, then two come along at once: All stars have twins, say astroboffins

Nearly all stars, including our Sun, are born from hot, dense molecular clouds and come in pairs, according to a paper to be published in the Monthly Notices of the Royal Astronomical Society. Binary star systems are common in space. It was estimated that up to 80 per cent of massive, bright O‑type and B‑type stars are locked …


  1. Your alien overlord - fear me

    Nice stating a ;fact which can't be disproven - if Sol had a twin (and they've already named it after a Star Trek movie), we should see it as a big hunka light, even if it's 1,000,000 AU away. We haven't spotted it so I would guess either (a) it never existed or (b) see a !!!

    1. Anonymous Coward

      Is ";fact" some way of trying to say "assertion"?

      Sorry, I just can't let the supreme ignorance of attributing the name to some movie pass. At least try googling before posting such drivel.

      I happen to share at least some of your scepticism, but I wouldn't be so arrogant as to rubbish the paper without reading it first. Meaningful criticism would require some insight into scientific method, and differences between evidence, proof, and the role of mathematical models in describing physical systems.

    2. Flocke Kroes Silver badge

      How much can you get wrong in one post?

      The idea of a companion star for sol named Nemesis dates back to 1984. Star Trek: Nemesis came out in 2002. I think the hypothetical star was named after the Greek goddess especially as ST:TNG only started in 1987 (Planet Vulcan predated ST:TOS by over a century).

      Astronomers have already found about 50 stars within 1,000,000 AU (15.8ly). Stars move relative to each other. Scholz's star (currently 20ly away) came within 52,000 AU (0.8ly) of Sol only 70,000 years ago. The problem of identifying Nemesis is more likely to be that astronomers have already found many stars the right age and composition, but they have little idea where they were 4.7 billion years ago.

    3. Anonymous Coward
      Anonymous Coward

      Binary/Twin Star?

      The problems I have with this theory/article are in the use of the words 'binary' & 'twin'.

      A binary system is one in which two stars are gravitationally bound and orbit each other about their common barycenter. There isn't an astronomical definition of a 'twin' star.

      If Sol formed in a binary system, where it was gravitationally bound with another star, then something very energetic indeed must have happened to the other star to expel it from the pairing: it wouldn't have just drifted off, as almost implied in the article; it would still be gravitationally bound to Sol.

      That Sol formed in relatively close proximity to other stars is pretty certain; molecular clouds are big - even the little ones contain around 100 solar masses, in which several stars would have probably formed.

      As for 'twin' - well, we've no way of knowing how similar to Sol other stars in our particular nursery were, but it's almost certain that Sol had several 'siblings'.

      1. Rol Silver badge

        Re: Binary/Twin Star?

        It's not beyond logic to assume the forces necessary to unbundle Sol from Nemesis, came about at ignition, and likely as not given the other gas giant/protostar a push toward ignition itself.

        Thinking as to why stars like our Sun appear to come in pairs:-

        Perhaps as a body travels through a nebula the material that falls into it's gravitational influence collides together many many miles behind it and thus coalesces into a substantive body, which in turn would have a similar wake of material falling into the recently travelled path.

        I guess a reasonable way of demonstrating this would be to scatter iron filings onto a smooth surface and then race a very strong magnetic ball across the surface.

        I would expect the filings would be drawn to the magnet, but due to the initial inertia and the speed of the ball, they would mostly end up in a line along the path the ball took. And once we factor in the likelihood that gravity would coral these individual particles together, and in a frictionless environment still be travelling in the balls general direction would eventually become a significant brake on the initial ball, and thus they end up in orbit.

      2. james 68

        Re: Binary/Twin Star?

        As both stars would be losing mass then it stands to reason that their gravitational pull would weaken to the point where they would no longer be bound, sending them off rather quickly in opposite directions. Depending on specific mass, orbital distance and speed, this could have happened very early in the life of the stars.

        1. Anonymous Coward
          Anonymous Coward

          Re: Binary/Twin Star?

          @james 68: there are a couple of problems with your idea of stellar gravity weakening due to mass-loss. The first problem is that as you increase the orbital distance between the two stars, due to the reduction of gravity due, in turn, to stellar mass-loss, you also decrease their orbital velocity. Thus as Sol and its partner drew further apart due to stellar mass-loss they would be moving more and more slowly; they would never reach "the point where they would no longer be bound, sending them off rather quickly in opposite directions" - they would just drift apart more and more slowly.

          The second problem is that if Sol's gravity weakened to the extent that its partner was able to drift away then all of Sol's planets etc. would also have drifted away.

      3. MyffyW Silver badge

        Re: Binary/Twin Star?

        Stars in binary systems can be very different from each other. Witness Sirius A, a white main-sequence star and it's companion, Sirius B, a faint white dwarf.

    4. a_yank_lurker Silver badge

      "Nice stating a ;fact which can't be disproven" - Not having read the paper, but I suspect the model (or an improved version) would give a clue as to where to look for the twin. If the twin is there, then the theory is on much better footing. Valid theories make predictions that can be verified and explain known observations. This theory offers a possible explanation for the number of binary systems which is a known observation.

      1. Pompous Git Silver badge

        "Not having read the paper, but I suspect the model (or an improved version) would give a clue as to where to look for the twin."
        Not really. The paper is about star formation and QI IFF you're an astronomer. Sol is ~4.6 billion years old. Tracing even an approximate trajectory of a star over that period of time is fraught. See: n-body problem

        The author of this paper proposes further tests that have the potential to falsify the theory presented.

    5. Michael H.F. Wilkinson Silver badge

      Just a little mathematics

      1,000,000 AU is about 4.8 parsec. At this distance, a M-type dwarf like Sholz's star which has an absolute magnitude of 19.4 would have a visual magnitude of 17.8, well beyond my 8" SCT. Even when Scholz's star passed at 0.8 ly, it would be a puny mag 11.3 star, roughly 250x to faint to see with the naked eye. There are some very, very faint stars out there. In fact, they outnumber the stars of the sun's brightness or brighter by a huge margin.

      It is well known that although many stars form in open clusters, these drift apart, mainly due to gravitational disruption by other objects passing by in the fairly dense galactic plane. NGC 188 (also known as Caldwell 1) is an exception, in that it is a very old open cluster, which probably wasn't disrupted, because it is some way away from the dense traffic in the plane. It is therefore easy to imagine that stars forming as a wide binary system could be disrupted, and the two would ultimately drift apart.

  2. John Smith 19 Gold badge

    So the 2 suns don't orbit each other and started drifting apart shortly after forming.

    Since this all happened several billion years ago the twin will be far away by now.

    Potentially excellent work.

  3. David Roberts Silver badge

    Identical or non-identical?

    Makes a big difference.

    1. This post has been deleted by its author

    2. MrZoolook

      Re: Identical or non-identical?

      Before I accept any candidates as being our sun's twin, I'll need some pretty hard evidence.

      I have Jeremy Kyle on speed-dial to do a DNA test when required.

  4. Pascal Monett Silver badge
    Thumb Down

    Seems difficult to accept

    Our closest neighbor is the Centauri system, which happens to be a multi-star system. I don't think it likely that our Sun's hypothetical twin can be found there.

    The second closest is Bernard's star, which is over 6 ly away. Other star systems are at over 8 light years away, but they are all multiple systems.

    So, for our Sun to have a twin, it would have to be Bernard's star, but apart from the distance, one would also have to explain how it could be a twin of our Sun when it is over 2 billion years older.

    The next closest single star is not even in the 26 closest list linked above, so it is over 12 ly away and that's getting ridiculously far for a so-called twin.

    So, for our Sun to have a twin, it would also have to have formed with our Sun, failed to ignite, got ejected from our system and cooled down sufficiently fast to not show up on infrared satellite surveys.

    That's starting to be a bit much to swallow.

    1. Christoph Silver badge

      Re: Seems difficult to accept

      Why do you assume that a star that has been drifting away from us for billions of years must be a close neighbour? It's not only been moving away that long, it's been pulled differently by other stars by being in a different place. Add that all up and it might be on the other side of the galaxy by now. If someone drove away from your house ten years ago, would you expect them to still be in the next street?

      1. Anonymous Coward
        Anonymous Coward

        Re: Seems difficult to accept

        If it had been ejected only at the speed of Voyager1 - which is about 1/20,000 c - then after 2 billion years it would be 100,000 light years away. That also happens to be the diameter of our galaxy.

        This means it could be any of the 100 billion stars in the Milky Way.

        1. DropBear Silver badge
          Thumb Up

          Re: Seems difficult to accept

          "If it had been ejected only at the speed of Voyager1..."

          This is a particularly startling instance of being "ninja'd" - I came to post the exact same reasoning starting from the exact same data point. I guess the problem with this sort of thing is that our intuition is worse than useless when it comes to the kind of numbers involved in astronomy. Fermi order-of-magnitude estimation method for the win, I suppose...?

    2. Anonymous Coward
      Anonymous Coward


      "So, for our Sun to have a twin, it would have to be Bernard's star, but apart from the distance, one would also have to explain how it could be a twin of our Sun when it is over 2 billion years older."

      Apart from what Christoph said there's another scenario which, so far, seems to be getting ignored: apparently we're all also assuming that this sun is still alive. Why?

      For all we know it could have collapsed in the mean time and is now one of the many black holes out there. Which would make it harder to spot.

      1. Rol Silver badge

        Re: @Pascal

        It could just have easily fallen into another star cluster.

      2. Brewster's Angle Grinder Silver badge


        "For all we know it could have collapsed in the mean time and is now one of the many black holes out there."

        The lifetime of a star is dependent upon its mass and, to a lesser extent, upon its chemical composition. The chemical composition of the sun and its companion will be identical, as they were formed from the same material at the same time. A massive disparity in mass seems unlikely, too; maybe it could be twice as big, but not ten times. So I would expect it still to be burning.

        1. IT Poser

          Re: @ShelLuser

          F class stars, solar mass of 1.04-1.40, have a lifespan of 2-10 billion years. A class stars, solar mass of 1.40-2.10, have expected lifespans of 2.2 billion years or less. Since you are willing to accept Sol's former companion was potential twice as big, I don't see why it is a stretch to say that star hasn't gone poof 2+ billion years ago.

          While I agree that, most likely, Sol's former companion is still burning, I still must downvote your downvote explanation.

    3. AndrueC Silver badge

      Re: Seems difficult to accept

      Who is 'Bernard' ?

      The name is Barnard's Star

      But it's an interesting possibility I suppose.

    4. This post has been deleted by its author

      1. John Brown (no body) Silver badge

        Re: Seems difficult to accept

        Yeah, and has President May done anything about it? Nooooooo, it's something that will happen after the next election!

    5. John Brown (no body) Silver badge

      Re: Seems difficult to accept

      The second closest is Bernard's star, "

      Oh FFS, I get that enough from my wife who can't manage to say "Barnard Castle", without having to put up with it here too!

    6. Alan Brown Silver badge

      Re: Seems difficult to accept

      "So, for our Sun to have a twin, it would have to be Bernard's star,"

      Our sun is several hundred light years (at least) from the neighbourhood where it formed and the odds of _any_ nearby star being related are slim to exceedingly negligable.

      1. Esme

        Re: Seems difficult to accept

        The Sun has made about 20 orbits of the galactic core in its lifetime, travelling a total of about 3 million LY in the process - if one assumes that its orbit has been relatively unperturbed, which may not have been the case. That said, even minor perturbations over the 4 billion-odd years of the Sun's existence by the very large number of other bodies in the Galaxy could cause the Sun and any star that formed nearby to be in radically differnt parts of the galaxy by now. Note COULD. As stars out here are quite sparse though, it is still possible that an initially gravitaionally bound association of stars could remain reasonably close together. The Sun's current velocity relative to the galactic core as a fixed point is about 220 km/s, and relative to the average velocity of other stars in the neighbourhood is about 20 km/s.

        Good luck trying to find any solar 'twin' out there! 8-}

    7. Winkypop Silver badge

      Re: Seems difficult to accept

      I'm just going outside and I may be some time....

  5. Zog_but_not_the_first Silver badge

    You know how it is with stars...

    Difficult second album, creative differences, the band splits.

  6. Aqua Marina Silver badge


    1. wolfetone Silver badge

      +1 for the truth

  7. Paul Herber

    You know how it is with twins, one of them is always evil.

    1. Brewster's Angle Grinder Silver badge

      With a name like Nemesis, it's hard not to be.

      1. Hero Protagonist

        "With a name like Nemesis, it's hard not to be."

        How do you it's not Sol that's the evil one?

        1. John F***ing Stepp


          No goatee?

          1. cray74

            Re: Well,

            No goatee?

            Sol went for a more heavy handed boogie man look, the Demonic Jack-o-Lantern

        2. This post has been deleted by its author

        3. Thesheep

          As the bard put it...

          Dr. Hibbert: That means the evil twin is and always has been...Sol.

          (They all turn around and stare at Sol.)

          Sol: Oh, don't look so shocked.

        4. DropBear Silver badge

          "How do you it's not Sol that's the evil one?"

          Easy. "We" is by definition "the good guys". Hollywood taught me so.

  8. This post has been deleted by a moderator

  9. Brewster's Angle Grinder Silver badge
    Thumb Up

    It was noticed we got the arxiv link. Thanks, Katyanna.

  10. Chris Gray 1


    You're all Sol-centric here, i.e. you are assuming that Sol formed as one of a pair, and then the other flew off. Isn't it equally likely that Sol is the one that "flew off"? In that case, the original system could have had more than 2 new-born stars, and so Sol's closest siblings are still in a multi-star group.

    Also, depending on the exact mechanism that results in multiple stars forming in a close group, does it follow that the stars are very close in initial size and density? I would think it would follow that the composition of the stars would be quite similar, so perhaps what astronomers can do is look for nearby stars with similar composition, backdated to the time of formation (i.e. ignoring expected composition changes since then, which might vary depending on the kind of star that they started as).

    1. IT Poser

      Re: Sol-centric

      Alpha Centauri A and B have noticeably different metallicities, 0.20 versus 0.23,and masses, 1.100M☉ versus 0.907M☉. I can think of several reasons that would explain the difference without requiring the two stars to form from different clouds.

      Over the course of 4.6 billion years Sol's companion could easily have traveled hundreds of light years in a different direction. It could even be one of the stars that has been shot out of the galaxy.

      I think your search area is too limited, based on the little we know.

    2. Jeffrey Nonken Silver badge

      Re: Sol-centric

      Yeah, it's like my marriage. She keeps telling people it was a mutual decision, but the fact is, she left me. I just didn't fight it.

      ... Oh sorry, you were saying something about stars?

      1. Graham Jordan

        Re: Sol-centric

        My hero.

  11. chivo243 Silver badge


    With the number of stars in the universe, I have a hard time believing that every star has a twin. Maybe they had one like the Miradorns, but collisions/black holes etc would surely have eliminated some of the twins...

    1. Pompous Git Silver badge

      Re: Statistics

      "With the number of stars in the universe, I have a hard time believing that every star has a twin."
      You won't be surprised to learn then that the paper proposes no such thing. Indeed, the word "twin" occurs exactly zero times in the paper. What Sadavoy and Stahler propose is a solution to the problem of binary star formation.

    2. TechnicalBen Silver badge

      Re: Statistics

      Multiples/doubles. Like matter/antimatter creation. There are natural mechanics and forces that cause a large percentage of duplications/doubles.

      Take for example the wake from a boat travelling in water. You get one on one side, the other on the other side. Though a nebular collapsing would have a 3 dimensional makeup, they can no doubt test to see if it would collapse in a way (or be effected to collapse) that causes two and more dens spots for stellar formation.


POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019