E-Letter responses to:

letters: Mark McCaughrean, Neill Reid, Chris Tinney, Davy Kirkpatrick, Lynne Hillenbrand, Adam Burgasser, John Gizis, Suzanne Hawley;, and Maria Rosa Zapatero Osorio What Is a Planet? Science 2001; 291: 1487b-1488b [Full text]

E-Letters: Submit a response to this article

Published E-Letter responses:

Black Sheep and Pre-paradigm Names

Nick Woolf   (2 April 2001)

Another vote for 'Rogue Planet'

Harold Nations   (2 April 2001)

Stars - Worlds - Rocks - and Rogue Planets

Larry J. Friesen   (2 April 2001)

Planetars

T. T. E.   (2 April 2001)

Planetary Erratics

Neil Ackerman   (22 March 2001)

Common usage is compatible with scientific value

George Wetherill   (22 March 2001)

Classification finds patterns

Michael F. A'Hearn   (1 March 2001)

What is not a planet

Chris Tinney   (27 February 2001)

This is an important question.

Dr. Ben R. Oppenheimer   (23 February 2001)

This is not an Important Question

Dr Dave Stevenson   (22 February 2001)

How should we define "planet"?

Prof. Gibor Basri   (22 February 2001)

Dual classification for Pluto and other bodies

Dr Brian G. Marsden   (22 February 2001)

What is a planet?

Dr Jack Lissauer   (22 February 2001)

Black Sheep and Pre-paradigm Names 2 April 2001
Nick Woolf, Astronomer Steward Observatory Respond to this E-Letter: Re: Black Sheep and Pre-paradigm Names	The reality is that we have seen a single sheep, one side of which is black, yet our response is to report "There are black sheep here." The conclusion cannot be drawn from the observation. We have objects like the companion to Sirius, Gliese 229B, and Jupiter, all of which receive less radiation from their brighter companion than they put out, yet we suspect that none of them are producing any major fraction of that radiated energy from nuclear reactions, and we give them three different names that result from inferences about their history. And Jupiter is named a planet to be the same as objects which mainly re-radiate energy from the Sun! We can afford to leave white and brown dwarf designations. But the lower mass objects surely deserve a better place-holding name. The current reality is that we have no observations that suggest that any system is substantially like our own. And the word "planet" has developed a suite of meanings historically connected with our own system, so that to give it to other potentially quite different objects is guarateed to produce confusion. Indeed, as our methods of finding objects of low mass improve, we are also going to be trying to match occultation observations, gravitational lens observations, astrometric observations and imaging observations. I suggest that while the confusion rages, and until a paradigm seems to be gathering consensus, we give all these objects that we consider to have a likely close relationship to planets a single name - "PMOs", planetary mass objects, pronounced "peemose". This name would be restricted to objects which seem likely to be no more than an order of magnitude more massive than Jupiter, and no less than an order of magnitude below Mercury. Clearly there will be some fringe objects, where we are uncertain what the mass is, and these can be called probable or possible PMOs. I too reject the Deuterium burning limit as being an appropriate boundary. It seems unlikely to play any significant role in the formation or evolution of these objects. Certainly, when we get a better model that explains the likely histories of these objects and systems, we may find that we will reclassify a few objects. But the nomenclature proposed does allow us a substantial amount of breathing space until we converge on a paradigm.
Another vote for 'Rogue Planet' 2 April 2001
Harold Nations, astronomer Nightwatch Observatory Respond to this E-Letter: Re: Another vote for 'Rogue Planet'	I find myself agreeing with Dr. Wetherill's proposals with one modest change: why not call the free-floating, supposedly planetary mass objects "rogue planets"? This would capture the basic idea that they seem to overlap the mass range of objects we all accept to be "planets" while conveying the major observable difference between them and similar mass objects in bound orbits.
Stars - Worlds - Rocks - and Rogue Planets 2 April 2001
Larry J. Friesen, Adjunct Instructor University of Houston at Clear Lake Respond to this E-Letter: Re: Stars - Worlds - Rocks - and Rogue Planets	I’d prefer to be cautious about using formation processes to define categories of objects. There may be more than one way to build the same kind of object, and I’d rather not have different names for them on that basis. Several years ago, Alan Stern and I batted around ideas for categorizing astronomical objects. Objects massive enough to radiate, powered by nuclear fusion at some stage in their history, would be “stars”. Objects too small for fusion but massive enough that their overall shapes are determined by their self-gravitation, their rotation, and possibly tidal forces of nearby massive bodies, would be “worlds”. Objects too small to be worlds, whose shape was determined by the strengths of their materials and their collision history, would be “rocks”. At the time of our original discussion, the concept of brown dwarf was not clearly formulated, and the “rubble pile” model for asteroids had not yet been put forward. So we’d probably want to include “brown dwarfs” as a class between stars and worlds, as objects massive enough to get deuterium fusion going, but not massive enough for proton-proton fusion, thus never achieving a main sequence lifetime as a star. And we’d want to include “rubble piles” along with “rocks” at the low end of the mass scale. Internal structure, not mass or composition, would determine which objects fit into the “rubble pile” class and which were “rocks”. Alan wanted the definition of a “planet” to include both mass and orbit. A world in orbit around a planet would be a “moon”. Thus some moons are worlds, and some are rocks (or rubble piles). I argued for extending the definition of planet to include worlds not in orbit around stars, but moving independently in the galaxy, to allow for the possibility of “rogue planets”, a term used by science fiction writers in the past. If an object has the mass of a world, and is not a moon, I fundamentally don’t care how it formed, I’d like to call it a planet. If the manner of formation leads to fundamentally different types of structure, I’d set up various CLASSES of planets. At the lower end of the mass scale, most asteroids are rocks or rubble piles, but Ceres, Pallas, and Vesta are clearly both worlds, and thus (minor) planets and asteroids (because they are members of a SYSTEM of objects), in this scheme. In the same way, Pluto is a world in orbit around the sun and thus a planet, or perhaps Pluto and Charon are a double planet, even though Pluto and Charon are also members of a system of Kuiper belt objects. I’ve read of a recent discovery of an exceptionally large Kuiper belt object which is probably of world size. If it is, it should also be considered a planet.
Planetars 2 April 2001
T. T. E., Student Respond to this E-Letter: Re: Planetars	Long ago people used the word planet to describe the tiny points of light that moved against the background of "fixed" stars. Once adequate telescopes were invented planets were found to have other smaller objects orbiting them. So why didn't the early astronomers call the smaller objects planets as well? Because such labeling would have been redundant and confusing. Instead of being repetitive, the new objects were called moons. So why not call the planets moons of the sun? Well, because they were already classified as planets. As for the current debate over wether the free-floating "planets" are planets, it is hard to say if they are planets or something else so it would make more sense to call them something new. For example, they could be called planetars.
Planetary Erratics 22 March 2001
Neil Ackerman, Attorney none Respond to this E-Letter: Re: Planetary Erratics	In the ever increasing realm of scientific classification, it appears that definitions need to be concise and well formulated in order to promote consistent debate and growth. With regards to celestial masses without the requisite mass required to achieve sustained fusion and without the "parent" body around which they orbit, the term "planet" appears inapplicable. "Planets" have necessary attributes. One of these attributes is the fact that a "planet" is a satellite of a larger body. However, this raises some definitional problems of its own. What attributes of a planet differentiate it from other satellites such as comets and asteroids? Atmosphere cannot be a distinguishing factor as Mercury, for example, has none. Likewise, size should not be the distinguisher unless one would want to declassify Pluto as a planet (assuming that bodies larger than Eros in the asteroid belt would still be considered asteroids). And speaking of Pluto, a regular orbital path, denoting formation of the planet during the condensation of the gas cloud, cannot be the determination as Pluto has an erratic path and is possibly a visitor from space that got itself caught in Sol's gravitational field. What, then, are the attributes of a Planet sufficient to support a well defined definition? If the definition does not require that a planet have the attributes of a satellite, then the so called "brown dwarfs", or "grey dwarfs", may well qualify as planets. I suggest taking an example from geology and the definition given to large boulders that are out of place due to glacial activity. These are call "glacial erratics" to denote the fact that they did not appear naturally in the setting where they exist. Since we expect planets to exhibit some attributes of a satellite, perhaps the best way to describe a free floating, self formed body with insufficient mass to generate sustained fusion and without a larger body controlling its path in an orbit could be as a "planetary erratic".
Common usage is compatible with scientific value 22 March 2001
George Wetherill, Planetary Scientist DTM,Carnegie Institution of Washington Respond to this E-Letter: Re: Common usage is compatible with scientific value	As others have recognized, the goal of this discussion is to maintain scientific usefulness without confusing the issue by tampering with historically well established usage. Even though I am not generally a supporter of positivism, I think the easiest way to accomplish this is by using operational definitions, and not be overly concerned by differences in formation models, especially since they are not sufficiently well understood to assume that role, even for our own planetary system. 1. Distinguish between a planet and a star by deuterium burning. Thus brown dwarfs are a kind of star. There will be borderline cases, but so what? Some day, when we understand formation of these bodies much better, finer distinctions can be made, but we don't know enough yet to make definitions based on theories of origin. 2. Don't downgrade Pluto, regardless of the reasons given by others. It can be called "the largest plutino" as well, by those who make use of this characteristic in their work, without removing its status as the smallest planet. How can that hurt anything? It doesn't imply that all planets formed in the same way. That's something we don't know yet and may not know for a long time. 3. There are just plain "planets", the size of Pluto or larger, and smaller bodies called "minor planets". The boundary is arbitrary, so why not just leave it where it is? The distinction between different kinds of minor planets, (asteroid, comets, etc.) can also be made in a way that does not change traditional usages, but still be scientifically appropriate. 4. Planets (and minor planets as well) are defined by being gravitationally bound to one or more stars. 5. Bodies that would otherwise be classified as planets, but are not gravitationally bound to a star, can be called "free planets" (including free minor planets, that certainly exist, even though they haven't be found yet.) Again, this doesn't imply any assumptions of their origin. 6. Similar distinctions can be made regarding "satellites", but that is not under discussion here. 7. When other distinctions absolutely need to be made by individual workers for valid scientific reasons, this can be explained by them on an ad hoc basis, without suggesting that these standard operational definitions be changed.
Classification finds patterns 1 March 2001
Michael F. A'Hearn University of Maryland Respond to this E-Letter: Re: Classification finds patterns	Why do we, as scientists, care how Pluto (or anything else) is classified? This question must be answered before we can intelligently deal with how to present the result of a classification to the public. Scientists put things into groups, the members of which share common properties, in order to find patterns that will enable s to better understand how the bodies work or how they became what they are. If we are interested in origins, then it is clear with our present understanding (which might change in the future) that free-floating bodies of mass comparable to Jupiter are not in the same category as Jupiter itself. Similarly, it is clear that Pluto is not a planet like Jupiter but rather a planet like the numerous Plutinos in the 3:2 libration with Neptune. Thus Pluto should be classified as the largest Plutino. On the other hand, if we want to know how gravitatinoally dominated bodies with a lot of ice work, the Pluto is the class of frammises, together with Europa, Titan, and Triton.
What is not a planet 27 February 2001
Chris Tinney, astronomer Anglo-Australian Observatory Respond to this E-Letter: Re: What is not a planet	First, let me make one thing perfectly clear. The the low-mass brown dwarfs discovered by Zapatero Osorio et al, Lucas & Roche and other astronomers (1) in young star forming regions ARE exciting and interesting. However, I cannot agree with any of Dr Zapatero Osorio's arguments for naming these objects "planets". It is argued (2) that "The definition of the word "planet" has been modified several times in the last three millennia on the basis of an increasing scientific insight." Whether this is true is debatable - one could argue that though we may have added more planets, our idea of what a planet is, has not really changed for at least 500 years. What is certainly true is that the recent discoveries of "planetary mass objects" in star forming regions, adds only one new item of understanding to our concept of "What is a planet?". Which is that objects with similar masses to other objects - about which we have no uncertainty about applying the name planet - clearly form in star forming regions. These same regions show a continuum of objects forming at masses from >100 solar masses, down to and accross the brown dwarf boundary at 0.072 solar masses, and down to and across the deuterium burning boundary at ~13 jupiter masses. Occam's Razor clearly suggests all these objects should receive a similar classification, unless we can provide evidence they are different. No such evidence has yet been provided. There ARE theoretical arguments which suggest that there is a minimum mass for the formation of objects by "star formation" processes. However, I would reiterate there is NO observational evidence to inidcate this. Indeed, the one thing we have learned from the detection of these new objects is to the contrary - a 'break' in star formation processes does NOT appear to take place near the deuterium burning boundary. The suggestions that a 'boundary' between planets and brown dwarfs, or a definition of what is a planet, should be based purely on an observable (eg. mass or deuterium fusion) is completely disingenuous. Measuring mass for these objects is a difficult, indirect and HIGHLY model dependent process. It is NOT a clean or direct measurement. Measuring the presence or absence of deuterium burning is currently completely impossible (except via indirect theoretical arguments leading us back to mass measurement). As such, in the absence of an (admittedly diffocult to acquire) understanding of formation processes and evolution, neither is worth adopting as a new "definition". If there is any uncertainty about the classification of these new objects, it is clearly beholden on us as cientists to make that clear, and to apply a name to these objects which reflects that status. What we shouldn't do is rush to apply the sexiest name possible. Or the name which will give our papers, press releases and grant applications the highest possible profile. Or the name which will gain us the most international recognition in future job searches. Or the name which will confuse and eventually anger the public, when we have to clarify ourselves. In short, we shouldn't call them planets. (1) P. W. Lucas and P. F. Roche, Mon. Not. R. Astron. Soc. 314, 858 (2000); J. Najita, G. P. Tiede, J. S. Carr, Astrophys. J. 541, 977 (2000); M. R. Zapatero Osorio et al., Science 290, 103 (2000). (2) Letter Response, M. R. Zapatero Osorio, Science 291, 1487 (2001) www.sciencemag.org/cgi/content/full/291/5508/1487b (3) Conflict of interest statement : I am a co-author on the original submission.
This is an important question. 23 February 2001
Dr. Ben R. Oppenheimer, Hubble Postdoctoral Research Fellow Univ. California-Berkeley Respond to this E-Letter: Re: This is an important question.	I think that Prof. Stevenson's remarks are somewhat misguided. While I understand his point, it is a good example of science dissociating itself from lay people, the very people who not only fund, but also provide the huge impetus for the search for planets outside the solar system. First, Prof. Stevenson suggests that we wait until a large sample of these objects is found and that the definition and the appropriate term will then simply fall into place. Yet he works in a field where such developments hardly, if ever, happen. Take for example the use of the word quasar. The term comes from quasi-stellar. No one working in astronomy today would be caught dead suggesting that quasars are stars, or even distant relatives of stars, although until quite recently they remained unresolved in astronomical images, point sources, like stars. Yet the name has stuck and we use it day in and day out in the field, even though the name has nothing to do with the physics of the object. There are countless examples like this, particularly in astronomy. Furthermore, if one were to take Prof. Stevenson's suggestion seriously, we would have to rename many of the objects in astronomy: Galaxy originally meant milk, and while there may be milk in some galaxies, we know they are not composed primarily of milk, and brown dwarfs are not brown at all, in fact perhaps purple is a better description since they are red in the optical bands and blue in the infrared. The names of objects rarely change based on our knowledge of them. Nor should they, continuity in the literature is useful. Second, to suggest that the actual words we use in our research are not important is absurd. In order for people to dicuss ideas such as those in this debate, we must have words that are clearly defined. In the case of reporting on planets this issue is of paramount importance because of a vast public interest in the subject. (An interest which is warranted because we actually live on a planet and any discussion of extrasolar planets necessarily has, and should have, implications for the possibility of life similar to what we know here on Earth.) When an astronomer talks about planets (however you care to define them) to a gathering of lay- people or a reporter, it is imperative that he understand the audience. That audience does not think of the term planet as undefined, or even questionably defined. The term is defined today by the examples of the planets in our solar system, by the images of those planets from telescopes, beginning in 1609 with Galileo's first drawings of the craters of the moon and the "starlets" orbiting the banded Jupiter, to the space probe images brought to us by NASA through missions such as Voyager and Galileo. It is further defined by pictures of the Earth from space. These are not nebulous and blurry images that are conjured into people's minds when the word planet is mentioned. To ignore this important question is to remain aloof of one's responsibilities as a scientist to report and explain what we are researching. I think the discussion is excellent, and in addition to focusing on the scientific definition, we must necessarily consider the broader impact of the use of the word planet.
This is not an Important Question 22 February 2001
Dr Dave Stevenson Caltech, Pasadena California Respond to this E-Letter: Re: This is not an Important Question	In this period of exciting new discoveries of low-mass bodies, some in orbit about a star and some not, the important goal is to determine what's there. The categories and names we give to these objects do not matter. The nature and origin of these objects matter a lot. It does not help to debate terminology, the usefulness of which can only emerge after a more complete inventory is made. We can and should debate their nature and origin and can do so without needing to decide whether they merit the designation "planet".
How should we define "planet"? 22 February 2001
Prof. Gibor Basri, Dept. of Astronomy Univ. of California at Berkeley Respond to this E-Letter: Re: How should we define "planet"?	The current controversies over the astrophysical definition of "planet" arise from the conflict between three different arenas in which the discussion might take place. I'll call them (i) characteristics (the physical properties of the object); (ii) circumstance (the circumstances and environment in which the object is found); and (iii) cosmogony (the way in which the object was formed). Essentially all the controversies come from a disagreement on how much weight to give arguments in each of these arenas (and lesser disagreements about where to draw the line within each of them). For example, the debate over "free-floating planets" is over the use of characteristics vs. cosmogony and circumstance, and the debate over Pluto is between characteristics and circumstance. The primary characteristic on which the border between planets and brown dwarfs may be drawn is mass. There are two obvious effects of mass one might use to make the distinction - source of luminosity and source of pressure support. The former suggests making the dividing line at about a mass of 13 jupiters (where deuterium fusion becomes possible); this has been preferred. It reflects the cultural impression that "stars shine by their own light, but planets don't". We might call objects that are ever capable of fusion "fusors". The circumstance important in defining planets is whether the object is in orbit (and what sort of orbit, around what). Is being in orbit (now, or at least during formation) an essential property of planets? Definitions in the popular culture generally insist on this. The "Pluto question" asks whether one can call the object a planet if there are a lot of similar objects in similar orbits (Ceres lost on this count). A final (defunct) criterion required the orbits to be fairly circular. This was actually part discussion in the third arena: cosmogony (since circular orbits were thought to imply formation in a protoplanetary disk). We now understand that formation in disks is generic to fusors and non-fusors. The extrasolar planetary systems that have been discovered show that we know too little about planetary formation to resolve the many questions that have arisen. It is natural, therefore, that any definitions of "planet" that rely predominantly on cosmogony produce currently unresolvable debates, and more confusion than enlightenment. From an astrophysical point of view, it is cleanest to take a purely mass-based definition. I don't agree that the presence or absence of fusion is a "minor phenomenon" qualitatively; it is not remarkable that at the boundary between classes it becomes so quantitatively. Since, however, the word "planet" is used broadly in our culture, it is reasonable to take account of its cultural meanings. I propose a definition that captures the essential astrophysical and cultural imperatives: "A planet is a non-fusor born in orbit around a fusor". This makes it sensible to call single non- fusors something like "grey dwarfs". The system recently found with a 10 jupiter-mass object at 0.3 AU from a solar-type star, and a 25 jupiter- mass object at 3 AU, is a stellar/brown dwarf binary system (albeit with an intriguing cosmogony) in which the star has a close giant planetary companion. My thesis is developed in much greater detail at: http://astro.berkeley.edu/~basri/whatsaplanet.htm .
Dual classification for Pluto and other bodies 22 February 2001
Dr Brian G. Marsden, Associate Director for Planetary Sciences Harvard-Smithsonian Center for Astrophysics Respond to this E-Letter: Re: Dual classification for Pluto and other bodies	Though seemingly evoking a very basic astronomical concept, the ancient word "planet" may have been doomed as soon as it ceased to apply to the seven traditional sky wanderers and acknowledged instead the place of the earth in the Copernican revolution and Galileo's recognition of scarcely smaller objects themselves in orbit about Jupiter. Whether one is talking about a "terrestrial" or a "giant" planet in our solar system, or a "free-floating" planet in a young star cluster, it has rarely been scientifically useful to use the word without at least some qualification. Nowhere has this been more evident than in connection with the smaller sun-orbiting bodies to which one could not obviously instead apply the word "comet". After its initial classification as the "eighth" planet, Ceres was catalogued as the first "minor" planet, following the recognition of other small bodies in the cisjovian belt. Several of the denizens of the transneptunian belt found in recent years have also now been added to this catalogue, including very recently, as No. 20000, an object the size of Ceres. Its popular affection as the ninth "major", or "principal" planet (given that Neptune understandably replaced Ceres as the eighth) has kept Pluto out of the small-bodies catalogue, even though it has a diameter little more than twice that of No. 20000 and the listing contains a number of companion "plutinos" with orbital periods precisely half as long again as that of Neptune. Many astronomers think it only a matter of time before another, more distant, Pluto-sized transneptunian is found and included in the compilation. It is a pity that Pluto would not be listed first. Some thinking would have it that a gravitationally collapsed, and hence approximately spherical body is a "planet" (if it isn't a star or a brown dwarf), with no further qualification needed. Such a definition would include Ceres, Pluto, the moon and a number of other small bodies. But small members of the solar system are very obviously classifiable by the relationship of their orbital characteristics to those of larger bodies, and the sheer numbers make organized tabulations of small bodies necessary. Consideration of both the physical and the dynamical situation therefore suggests that an appropriate compromise might be to describe some bodies as having "dual status". Indeed, a few objects have already been classified as both planetary and cometary.
What is a planet? 22 February 2001
Dr Jack Lissauer NASA-Ames Research Center Respond to this E-Letter: Re: What is a planet?	Thousands of objects can be seen in a clear night sky. Apart from the Moon and the occasional comet, all of these objects appear point-like to the unaided eye. Most of these "stars" remain fixed relative to one another; the five exceptions, which move relative to other objects in the sky, were called "planets" or wanderers. The Copernican revolution showed that Earth orbits the Sun as do the celestial planets. Two additional large planets and tens of thousands of smaller planets have subsequently been discovered. The five wanderers known to the ancients, plus the Earth and the three largest bodies that orbit the Sun are known as planets, whereas smaller bodies orbiting the Sun are referred to as "minor planets" or "asteroids". The Earth and the other planets, major and minor, as well as their moons, are now known to shine by reflecting sunlight, whereas the fixed stars glow from internally produced energy, as does our Sun. Stars derive their energy primarily from sustained fusion of ordinary hydrogen, which requires a mass of at least ~ 8% that of the Sun. In contrast, Jupiter, the most massive planet orbiting the Sun, is just under 0.1% of the Sun's mass. Substellar objects are faint and difficult to detect beyond our Solar System, and the gap between stars and the planets in the Solar System is large enough that two distinct distributions are clearly defined. In the 1970's, the name "brown dwarfs" was given to a then hypothetical class of objects intermediate between planets and stars (1). Since the mid-1990's hundreds of substellar objects of various masses have been discovered outside of our Solar System, filling in the gap between stars and the largest planets within our Solar System (2). Additionally, many large minor planets are being discovered beyond the orbit of Neptune. Various definitions of a planet, based upon mass, origin and/or orbital circumstance have been proposed. These definitions generally place objects known prior to a decade ago into their traditional classification. However, new discoveries of intermediate objects require us to consider what properties should be used to define a planet. I believe that the definition of a planet should be based primarily if not exclusively on mass, with the boundary between planets and brown dwarfs being the mass at which an object fuses half of its deuterium (about 13 times the mass of Jupiter). Alternative proposed definitions based on origin suffer two major problems: (i) It is extremely difficult to determine the origin of a body, even with extensive observations. Jupiter is composed primarily of hydrogen and helium, yet has several times the fraction of heavier elements compared to the Sun. Most researchers believe that Jupiter formed from a predominantly heavy element core that accumulated solar-composition gas (3), but a solar-composition body that subsequently accreted planetesimals would look very similar. (ii) There may not be a clean dividing line between objects formed from direct collapse/fragmentation and those formed from instabilities in disks. For example, objects resulting from instabilities when a clump hits a disk would be intermediate in nature. The dividing line between core nucleated accretion and gas instability is probably cleaner, but some researchers (4) would then consider Jupiter a brown dwarf. Stars are defined by their ability to sustain quasi-equilibrium fusion of ordinary hydrogen at some stage in their evolution, regardless of whether they orbit a more massive star or not. Analogously, brown dwarfs can be defined as objects massive enough to fuse a majority of their deuterium, but not sustain fusion of hydrogen. The term "planet" should apply only to bodies too low in mass for substantial fusion of any type. However, to correspond with the Solar System, very low-mass objects would not qualify as planets (they would be referred to as minor planets or asteroids), and objects orbiting a planet would be classified as moons rather than planets. Moons must be less massive than the planet they orbit, but two moons in our Solar System are larger than the planet Mercury, and seven (one of which, Triton, is believed by many to have formed in heliocentic orbit (5)) are larger and more massive than the planet Pluto. Finally, we come to free-floating objects of planetary mass. Surely some of these formed like stars - even if a more massive cloud core is required to initiate collapse (6), occasionally the action of nearby massive stars will intervene and leave a gravitationally bound object of less than 13 Jupiter masses. Equally evident is that some such objects will have formed in disks and then been ejected by encounters with other planets (7), a companion star, or an interloping star (8). For objects of a given mass, one formation mechanism may well dominate. Nevertheless, in most cases, it will be impossible to know the origin of a particular object with great confidence. Thus, we should avoid using origin hypotheses as part of the definition of "planet". However, in analogy to moons, which must orbit a planet, one might consider using "free-floating planetary mass objects" (and ultimately a new term) for small interstellar objects and reserve "planet" for objects orbiting a star. 1. J. C. Tarter, Astrophysics of Brown Dwarfs, (Cambridge, Cambridge Univ. Press, 1986), p. 121. 2. www.exoplanets.org 3. J. B. Pollack, O. Hubickyj, P. Bodenheimer, J.J. Lissauer, M. Podolak and Y. Greenzweig, Icarus 124, 62 (1996). 4. A. P. Boss, Science 276, 1836 (1997). 5. P. Goldreich, N. Murray, P. Y. Longaretti and D. Banfield, Science 245, 500 (1989). 6. P. Bodenheimer, A. Burkert, R. Klein, and A. Boss, Protostars and Planets IV (Tucson: University of Arizona Press; eds Mannings, V., Boss, A.P., Russell, S. S.), p. 675 (2000). 7. H. F. Levison, J. J. Lissauer and M. J. Duncan, Astron. J. 116, 1998 (1998). 8. G. Laughlin, and F. Adams, Astrophys. J. Lett. 108, 171 (1998).