We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people.

Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house]

There are more things in heaven than were dreamed of in the philosophy I was taught at school

There has been quite a bit of press coverage in the last couple of weeks about the discovery of an object in the outer solar system, which has been given the astronomical name 2003 UB313, but which has been popularly dubbed “Xena”. In some circles this has been described as a new planet, and in others its discovery has been given as comprehensive proof that Pluto is not a planet and that there are only eight planets in the solar system. Personally I have two opinions here. Firstly, I think it should be “Rupert” and not “Xena”. And of the two viewpoints given, I tend to agree with the second, which is that the new discovery reduces the number of planets to eight. Although thinking about it some more, I am not sure that either viewpoint is right. A better interpretation might be that it reduces the number of planets to four. Or perhaps to zero. It all depends on your point of view.

Why do I think this? In order to properly understand the question, an astronomical primer is in order. Many of our readers will already know this stuff, but this is all quite interesting and is nice to put it all down in one place.

Let me describe the solar system. For the moment, I am going to leave Pluto out, as it does not fit into what I am to initially describe. The solar system is generally considered to contain two types of planet. One is the inner planets (Mercury, Venus, the Earth, and Mars). These orbit the sun at distances between 50 million kilometres and 250 million kilometres, and have radii of between 2500km and 6500km. They have surfaces made of solid materials (ie rock) . The second type of planet is the “Gas Giants” (Jupiter, Saturn, Uranus, and Neptune). These orbit the sun at distances of between 0.75 billion and 4.5 billion kilometres, and have radii of between 25000 and 70000 kilometres. They basically consist of atmospheres that get denser and denser as the altitude gets lower and lower, and which gradually thicken until at some rather indeterminate point they go from being a gas to a liquid to a solid or to even more exotic things that defy simple classification. These planets are orbited by many small rocky moons, and planetary rings. Three of them (Jupiter, Saturn, and Neptune) are also orbited by larger moons that would count as planets in their own right if they orbited the sun and were part of the inner solar system.

The orbits of these planets (of both kinds, but with one exception that we will get to) have two notable facts about them. First, they all circle the sun in approximately the same plane, known as the “plane of the ecliptic” or just “the ecliptic”. As a consequence, if there are a number of planets visible in the sky at the same time, they tend to be in a fairly straight line. Secondly, the orbits of the planets are approximately circular.

But they are not exactly circular, and they are not exactly in the same plane. Mathematicians have ways of quantifying both these things. The first of these is relatively simple. Simply measure the angle between the ecliptic and the plane of the planet’s orbit, and quote this number as the orbital inclination. With the exception that I will get to in a moment, the planets discussed already have small orbital inclinations of up to about three and a half degrees. See here for detailed planetary statistics of various kinds, including inclination.

While Copernicus was the first modern scientist to recognise that the Earth and other planets went around the sun, his theory did not quite successfully explain the movements of the planets in the heavens. That took someone with better mathematics. Johannes Kepler managed to fix this problem in around 1620 by explaining that the orbits of the planets are not precisely circular, but are ellipses. Mathematically, the shape of these ellipses are described by a number called the eccentricity of each ellipse. This number has a a precise mathematical definition that falls out of certain solutions to Newton’s equations of gravity and motion fairly naturally, but it is enough to explain that it is zero for an exact circle and it gets greater as the ellipse gets longer and thinner. When one end of the ellipse is stretched all the way to infinity the eccentricity becomes equal to one and the orbit becomes a parabola rather than an ellipse. (At this point we are digressing towards something called the mathematics of conic sections, which I will merely provide a link to rather than describe in detail).

So what is the exception? That would be Mercury, which has an orbital inclination of 7 degrees and an eccentricity of about 0.2. This makes it highly unusual, but as it is broadly similar in composition as the other inner planets and there are no other objects like it to be classified into a third class of planets, it is generally considered one of the inner planets.

As it happens, there is a simplification in the above analysis. While the orbits of planets around the sun are almost ellipses, they are in fact not quite exact ellipses. If a planet was orbiting the sun and was not affected by the gravitational pull of any other object, the orbit would indeed be an ellipse. But as it happens the planets exert gravitational effects on one another, and these cause perturbations from exact elliptical orbits. (In a strict sense, the orbit of a given planet is influenced by the gravitational pull of every other object in the universe, but in practice it is the gravitational pull of the other planets that matters).

Prior to the eighteenth century, planets beyond Saturn were unknown to science. Calculations were made by various mathematicians as to the effects of the gravitational pulls of the planets of each other. (Without a computer this is very difficult to do accurately). However, it seemed that it was not possible to explain the orbit of Saturn. One potential explanation was that another unknown planet was influencing the orbit of Saturn. The approximate location of this planet was determined by further calculations. Astronomers started looking for it, and Uranus was discovered William Herschel in 1781.

Of course, mathematicians then did another series of calculations, and discovered that they couldn’t properly explain the orbit of the new planet Uranus. Once again the explanation was that there was another planet, and Neptune was discovered in 1846. By this point a pattern had been established. Calculations of the orbit of Neptune could not be properly explained and another planet further out was predicted. Astronomers looked, and in 1930 Clyde Tombaugh at the Lowell Observatory in Arizona discovered a ninth planet. The name it was given, Pluto, was coined by an eleven year old girl in Oxford named Venetia Burney.

Except that the calculations that the discovery of Pluto was based on turned out to be wrong. The mass of Neptune used in the relevant calculations were incorrect, and Neptune’s orbit could be adequately predicted without any need for an additional planet. It just happened that there was something where Tombaugh happened to be looking. That something was very peculiar by the standards of the other planets. Pluto has an eccentricity of 0.25 and an orbital inclination of 27 degrees. These are greater than any other planet, including Mercury. And Pluto has less than a tenth of the mass of any of the other planets, although it took a few decades for this to be determined.

Pluto was odd, yes. But this was not initially a huge issue. Mercury is odd, too. If there is an odd planet at each end of the solar system, so what. That just makes things interesting. But there was another possibility. Pluto was discovered because one small piece of the sky was being intensively examined. There were many other pieces of sky which were not being intensively examined. (Many pieces of sky a long way out of the ecliptic were and are seldom examined). What if there were other Pluto like objects, and rather than being lucky, Tombaugh had instead simply discovered one of many such objects. This does actually appear to be what happened, but we have only known this recently.

As well as the planets, the solar system contains many objects that orbit the sun that are too small to be classified as planets: these are known as asteroids and planetoids. The two terms are distinct, but overlap somewhat, and a third expression “minor planets” is sometimes used to describe them and other objects we will get to later. Most of the largest asteroids orbit between the inner and outer planets (ie between Mars and Jupiter).

But in recent years, a new class of minor planet has been predicted and discovered in the outer solar system

As knowledge of how the solar system formed has grown, then it has become steadily clearer that something called the Kuiper belt exists beyond Neptune. This consists of small rocky objects that have been ejected from the inner solar system by the gravity of Jupiter. These objects typically have high orbital inclination and eccentricity. Quite a lot of these objects have been discovered over the last fifteen years. It takes quite a lot of effort to find them because they are both quite small and distant and thus very feint, and also because their high orbital inclination means that they are further out of the ecliptic than are the planets. Interestingly enough, there appears to be an outer limit to the Kuiper belt, “the so called “Kuiper gap” or “Kuiper cliff”, beyond which there are no further such objects. It has been postulated that there is a larger object (perhaps the size of Mars or the Earth) orbiting at that distance which interferes with the orbits of Kuiper belt objects and does not allow them to remain in orbits beyond this point.

In any event, the mathematical analysis that predicts the Kuiper belt also allows for objects in it the size of Pluto, and larger, up to about the size of the earth. Thus it has been the opinion of many planetary scientists for a few years that Pluto was just another Kuiper belt object. While it would be okay to keep Pluto in the list of planets (despite being very dissimilar from all the rest) it would not be okay to do so if it was just one of a class of other objects. It would not be possible to include the rest of this class of objects in the list of planets, because many of them are smaller than other objects (eg some asteroids) that we do not count as planets. Other scientists and non-scientists had argued that as long as Pluto was larger than any other Kuiper belt object (and closer than the vast majority) it was okay to keep it on the planets list, largely for historical reasons.

But this position would go from eccentric to deeply misleading if Pluto was not the largest such object, and if there were other similar sized objects roughly the same distance from the earth. And over the last few years that position has got steadily shakier. More large Kuiper belt objects have been discovered, and Pluto’s status has become steadily less unique. In particular Quaoar was discovered in 2002, which was more than half the size of Pluto, although considerably further out. At the time there were a few calls to remove Pluto’s planetary status, which were listened to but generally ignored.

And on July 29 this year, the discovery of three new objects was announced. Two of these was a bit less than the size of Pluto, the other, 2003UB313, the aforementioned Xena, is clearly larger than Pluto. It is a little further out than Pluto, but in cosmic terms not dramatically so. Xena’s closest point to the sun is substantially further out than Pluto’s furthest point out, so in any meaningful way of looking at things they are neighbours. It is clearly an accident that Pluto and not Xena was discovered in the 1930s. (However, Xena has a very high orbital inclination – around 41 degrees). If Tombaugh had instead been looking in the direction of Xena, then he would have seen Xena. Xena can clearly be seen in many astronomical photographs going back decades now that people need to look at it. Incidentally, one of the other new objects announced last week, 2003 EL61 or “Santa”, has a moon, so attempting to use “Only planets have moons” as part of your definition of a planet doesn’t work. The truth is that we have reached the point where Pluto’s planetary status is not something that can be kept if we wish to keep any kind of consistency. It might be possible to give planetary status to Xena for now, but almost certainly another object of Xena’s size or larger will be found before long.

So, for now, we can keep some consistency by saying that Mercury is the smallest planet, and anything larger is a planet but that anything smaller is not. This is at least a consistent definition.

But even this presents us with problems, because there are clearly an awful lot of things that are orbiting the sun that we do not know about. What about that Mars or Earth sized object out at the far edge of the Kuiper belt? The chances are excellent that it is there, and I suspect that we shall actually see them before now. (Track the orbits of existing Kuiper belt objects, look at any large perturbations to their expected orbits, and narrow the positions down. Much more complicated forms of these calculations can be managed in these days of computers). Almost certainly such objects will be found that are larger than Mercury. At that point we either include these new objects as planets – something I think there would be a strong case for doing – or we change the definition of planets again, to something like, “objects that are orbiting the sun, that are the size of Mercury or larger, and that are no further from the sun than the orbit of Neptune. Or you could perhaps exclude Mercury, and work from “planets Mars size or larger, and….., and……

But it all perhaps gets down to a simple fact. Rather than just a few planets, there is in fact an immense amount of stuff orbiting the Sun. Any distinction between “planets” and “something else” is rather arbitrary. There are certainly distinct classes of object, including “Gas Giants”, “Inner planets”, “main belt asteroids”, “Kuiper Belt objects”. “comets” and more. Perhaps the very idea of a “planet”, which is a category that includes at least two of these distinct classes of object, is an idea that has outlived its usefulness. In any event, we have to add an awful lot of qualifications to the definition make it work. I suppose one could narrow down the definition further, and say that only the gas giants, “Jupiter, Saturn, Uranus, and Neptune”, are the only true planets, and that everything else merges into everything else with so little in the way of discontinuities that such a definition is the only one but makes sense.

But that leaves us with earth not being a planet, and as planets were originally considered to be “earth like objects”, then this is also a bit of a problem. It may be time to drop the idea of a planet entirely.

19 comments to There are more things in heaven than were dreamed of in the philosophy I was taught at school

  • I would use this definition: a “planet” is a body orbiting the Sun which owns one of the orbital radii which are consistent with Bode’s Law.

    By “own” what I mean is that any other large objects in the same orbit are held in gravitational embrace by the presumed planet (i.e. either in orbit around it or caught in one of the trojan points or in “catch-up/lose-ground” bistable orbits).

    By this definition the classic 8 would be planets, whereas Ceres and Pluto (and Xena) would not be.

    I’ve long felt that Pluto shouldn’t be considered a planet. Another strike against it was after Charon was found, and they were able to calculate Pluto’s mass, and thus its density. It turns out it’s a big iceball; it’s not rock/metal like the inner 4.

    Steven Den Beste

    (first posted in error at the posting above)

  • Verity

    Michael – what an absolutely fascinating narrative!

    I didn’t understand it all, but what an interesting and lucid introduction. I read every word – not all of which I could grasp (degrees, for example), but close enough to hold the attention.

  • Sorry, I meant to include a link to an explanation of Bode’s Law.

    I think that the existence of the Kuiper Belt objects pretty much proves that there aren’t any more planets out there (using my definition), for two reasons. First, they represent the mass that should have become such a planet (just as the asteroids represent at least some of the mass that should have made up the planet between Mars and Jupiter, but didn’t because of the effects of Jupiter’s gravity) and second because if there really were such a planet out there, it would long since have disrupted the orbits of the majority of the Kuiper Belt objects which we see today.

    Is it pedantic to point out that no matter how many there are, there has to be a “last” planet? After all, the Solar System cannot be infinitely large. The idea of an earth-sized object outside the Kuiper belt doesn’t appeal to me intuitively, in part because it seems that such an object that far out would have had too many opportunities over the last 4 billion years to be gravitationally captured by other stars passing by as Sol and everyone else orbited around the galactic core. I’m also not convinced that the “Kuiper Cliff” isn’t actually an observational artifact caused by a sensitivity floor on observational equipment, a limited number of scopes available for observation, and the relatively small number of years that people have been looking (and quite possibly a certain amount of confirmation bias, which has been known to happen in astronomy).

  • Thanks for that Michael. I’d always wondered why astronomers were so sniffy about Pluto and now for a short time (until I forget) I know.

  • Monsyne Dragon

    Hmm. The best non-arbitrary definition of a ‘planet’ that I’ve seen is “A non-stellar object orbiting a star, that is massive enought to shape itself into a sphere under it’s own gravity.”
    By this definition, Pluto is a planet. So is “xena”. So is anything in the solar system, besides the sun, that isn’t a moon, and is over 700-1000km in diameter.

  • Bill Price

    “planets were originally considered to be ‘earth like objects'”

    What is your basis for saying this? Surely, the ancient astrologies didn’t look at the glowing lights in the sky and say, “Hey, those must be just like the immobile, stationary, center-of-the-universe object that I’m standing on!”

    Until Galileo’s identification of Jupiter’s four largest moons and his observation of the rugged surface of our own moon (which was largely assumed to be a perfect sphere), I don’t believe anyone had any inkling that those moving, sparkling lights could ever be of similar complexity and structure to the Earth.

  • Monsyne, the problem with your definition is that it would also includes Ceres, and there’s long been consensus that the asteroids are not planets.

    Ceres is 950 km in diameter and is definitely a sphere. And it has a density of 2.05 gm/cm^3 compared to 1.75 gm/cm^3 for Pluto.

    (Compare Mercury at 5.4, Venus at 5.2, Earth at 5.5, Mars at 3.9, Jupiter at 1.32, Saturn at 0.68 gm/cm^3.)

  • My take on this controversy is that planets are like elephants. They are hard to define, but you know one when you see one.

  • RPW


    But basing your definition on Bode’s law (which is in any case almost certainly a coincidence) will also include Ceres – the thing was found by people looking for the “missing” fifth planet that Bode’s law predicted after all. For that matter, Bode’s law doesn’t work for Neptune, and nobody’s proposing to demote that.

    Monsyne’s definition is frankly the only one that makes any real-world sense (as distinct from history, tradition or whatnot) and if that means we have to retcon in Ceres into the planet list then so be it.

    If people really have a problem, then I suppose we could revise the definition to something like “A non-stellar object orbiting a star, that is not part of a larger population of similar objects and is massive enough to shape itself into a sphere under it’s own gravity” might do, but this would certainly mean demoting pluto, which gets us back to where we started.

  • John East

    RPW, I think compliance with Bodes’ Law should be used to support the identification of an object orbiting the sun as a planet, although I suppose orbital radii could change over time due to close encounters with other objects. I cannot dismiss this law as you do as “a coincidence”. The odds against coincidence must be immense.

    If you consider the asteroid belt (along with Ceres) to be a fifth planet which either didn’t assemble itself because of Jupiters gravitation, or maybe was destroyed by a collision with some other large object, then Bodes’ Law holds up well. As for Neptune failing to comply, an orbital radius of 38.8 AU rather than the predicted 30.1 AU means that it is still closer to its alloted slot than any of the other planets. Maybe its orbit was pertubed when the axis of Uranus was tipped over?

  • Ceres doesn’t “own” that orbit. I included that clause in my definition precisely so as to exclude Ceres.

    No one knows the reason for Bode’s law, but it is highly unlikely that it’s coincidence. It is more likely that it represents some sort of gravitational resonance with Jupiter. It may be related to the effect which causes the “grooves” in Saturn’s rings.

    The real problem with my definition is that it doesn’t seem to apply to the extra-solar planets which have been found so far. (Sigh)

  • John East

    Steven, you make an excellent point. Extra-solar planets don’t conform to Bodes’ Law, so we may as well ditch it as a general rule.
    I’ll just have to cop out, and fall back on the idea that the billions of big and small lumps of stuff orbiting stars do not care how we classify them so why should I lose sleep over it.

  • RPW

    What does “own” mean in this context? Ceres on it’s own comprises at least a third of the mass of the asteroid belt – which we now know is a far greater fraction than what Pluto represents of the Kuiper belt. It really is difficult to come up with a definition of a planet (other than history and tradition, anyway) that includes Pluto and excludes Ceres.

    Using a definition of “owning” can exclude other bodies we are used to thinking of as planets – Earth, for example. Astronomers tend to view the Earth and Moon as two planets orbiting the Sun in the same orbit whose paths are each heavily perturbed by the gravitational pull of the other rather than a planet and a satellite in the same sense as, say, Jupiter and Ganymede. Which body owns the orbit?

  • Findlay Dunachie

    Dear RPW

    A personal message: I have always wanted to thank you for your Comment on my review of Melvyn Bragg’s “The Adventure of English”, in which you give the relevant reference about the Y chromosome similarity from Frisia right across the English Midlands to the Welsh border but the e-mail address you give just gives a bounce-back.

    It was extremely interesting, and you may have noticed that Bragg says that Frisian is the Germanic dialect closest to the “Anglo-Saxon” base of English.


    Sorry, everyone else. It’s a great post, isn’t it? It needs printing out, of course, to appreciate it fully.

  • Kristopher

    “Planet” means wanderer … ancient astronomers noted that some of those bright stars moved about.

    Every planet since the original five was simply given that status by astonomers … including the Earth.

    What consitutes “planethood” is whatever the current consensus says it is.

  • Richard Thomas

    Pah, you libertarians always miss the obvious answer. Clearly a commision must be established (under the auspices of the UN of course but funded by the UK and US) to study the problem of such assignments. Fact finding trips to the Carribean and unlimited expense accounts for all. Hurrah!

  • Strophyx

    “Clearly a commision must be established (under the auspices of the UN of course but funded by the UK and US) to study the problem of such assignments. Fact finding trips to the Carribean and unlimited expense accounts for all. Hurrah!”

    You overlooked the requisite specialization for such bodies. While the English-speaking countries will fill their usual role of funding the activities, the official name of the body and the primary language of any reports issued will, as usual, be French.

  • If “orbiting a star” means “attracted more strongly to a star than to anything else”, the Moon is orbiting the Sun.

  • Dillon

    if bodes law were followed how many planets could be fit in the area affected by the suns gravitational pull