Astrophysics in 2000

Among the entities to be welcomed this year were the remaining three mirrors, at 8 meters each, of the Very Large Telescope, the dedications of the Green Bank Telescope and LIGO (no published data from either yet), the opening of the new Rose Center and Hayden Planetarium at the American Museum of Natural History (which also seems to have begun a new custom of reviewing such exhibits as if they were books), the Gruber Prize in Cosmology (first two winners, P. James E. Peebles and Allan R. Sandage), and the (apparently successful) launches of Newton‐XMM (10 December), Cluster II (in two lots, 12 July and 9 August), HETE II (a second try at the High Energy Transient Explorer, the second adjective referring, we hope, to the events to be monitored, not to the mission itself), the Shenzhou "unoccupied orbiter" (21 hours in orbit on 26 November), the part of the International Space Station called Zvezda (on 12 July), and ACRIMSat (20 December). About 150 Ph.D.'s in astronomy and related subjects were awarded in the USA during the year.

During the year, there were some things that came and went, or went and came, so quickly or so often that it was hard to decide whether to say "hello" or "goodbye." Among these were the radio interferometry satellite HALCA (with about three near‐failures and rescues), NEAR (which came much NEARer to the asteroid Eros in February 2000 than in the previous year), the Pluto‐Kuiper Express (which was cancelled, but followed quickly by a new Plutonic announcement of opportunity), MIR, New Astronomy Reviews (a reincarnation of the old Vistas in Astronomy), and the Iridium set of communications satellites (which really does deserve to be called Dysprosium). A special award goes to Voyager 1, which has reached 76 AU from the Sun (the furthmost the mind of man has ever set foot, as it were) and still hasn't reached the heliopause.

Among the things we lost during the year were the Mars Polar Lander and its associated penetrators (declared dead in December), the Compton Gamma Ray Observatory (put into a geostationary orbit on the ocean floor on 4 June), the NRAO 12‐meter dish (though closing a ground‐based telescope is generally less irreversible than the space‐based case), Astro E (an unsuccessful launch on 10 February, continuing the Japanese tradition of never losing a named spacecraft), Comet LINEAR (whose brief appearance in July as six smaller comets showed that there is intermediate‐level structure between dust grains and the nucleus, to be called cometesimals, if you must), and the Irish Astronomical Journal, which ceased publication in mid‐volume after 50 years. It was a somewhat informal, chatty publication (long edited by Ernst Opik), and its loss is yet another in a sequence of relatively informal publications that have disappeared in recent years, including Quarterly Journal of the Royal Astronomical Society, Astrophysical Letters, and another journal edited for some years by the more blue‐pencilled author and not named here for fear of legal action. ROSAT collected a small amount of posthumous data between late October and late December 1998, spotting a z = 4.22 blazar and the highest redshift X‐ray absorption to date (Boller et al. 2000). And ASCA succumbed to atmospheric drag on 12 July.

Our human losses included five staff members of IRAM killed in an accident (to whom Cox et al. 2000 is dedicated), Dennis Sciama ("doktorgrossvater" and even gross gross vater to a number of US‐based astronomers), Colin Ronan (one of the great writers of popular astronomy), Abraham Taub (one of the last direct links to Einstein), and the following members of the American astronomical community, in roughly the order that the news reached us: Walter Wild, Billy McCormac, David Beard, John DeWitt, Sidney Kastner, Valentin Boriakoff, Charlene Heisler, Jan van Paradijs, Freeman Miller, John Evans, Gareth Jones, William Calder, Carol Rieke, Clinton Constant, Douglas Duke, Patricia Rogers Campbell, William Kaula, Gijsbert van Herk, John Wolbach, Philip Keenan, James Cuffey, Harrison Mendenhall, Samuel Goldstein, Jeffrey Willick, Jerome Korman, Frederick Hollander, Jean Heidemann, K. Narahari Rao, James W.‐K. Mark, Edward Dyer, Robert Hjellming, Henk van de Hulst, John Simpson, Bill Fastie, Donald Billing, John O'Keefe, Herbert Friedman, Frank Kerr, Raymond Grenchik, and Joseph Weber (on the 30th of September, last day of the index year and first day of year 5761).

2.1.1. Neutrinos—Do They Care about the Solar Magnetic Field?

2.1.2. The Third Era of Helioseismology

2.1.3. The Search for Gravity Waves (g‐Mode)

2.1.4. Solar Internal Rotation—A Matter of Anchor Depth

2.1.5. Getting in Touch with the Solar Core

2.1.6. Solar Internal Structure Inversion (p‐Mode)

2.1.7. Fundamental Global Oscillations (f‐Mode)

2.1.8. Ocean Waves on the Sun? (r‐Mode)

2.1.9. Local Helioseismology

2.1.10. Solar Dynamo

2.2.1. Dermatology of Granules, Mesogranules, and Supergranules

2.2.2. Dermatology of Intergranular Paths: The Network

2.2.3. Flux‐Tube Dynamics or Spaghetti Cooking

2.2.4. Sunspot Oscillations and Dynamics

2.2.5. Photospheric Abundances

2.2.6. Solar Rotation and Oblateness

2.3.1. Time‐averaged Structure

2.3.2. Convection‐driven Dynamics

2.3.3. Reconnection‐driven Dynamics

2.3.4. Magnetic Mapping

2.4.1. The Coronal Magnetic Field

2.4.2. The Quiet Sun

2.4.3. Coronal Loops

2.4.4. Loop Oscillations

2.4.5. The Coronal Heating Problem

2.4.6. Microflaring, Nanoflaring, Picoflaring, ...?

This section is mainly about eruptive phenomena in the solar corona, which includes eruptive filaments, prominences, flares, and coronal mass ejections. In large eruptive events one sees all of these phenomena performing in concert, but the conductor (always a magnetic destabilization process) may sometimes give any type of instrument a solo part, or may pick out any subset of instrument combinations. The only kind of performance we have not seen yet is whether the orchestra can perform without conductor (e.g., a flare without a magnetic driver).

2.5.1. Filaments

2.5.2. Prominences

2.5.3. Flares

2.5.4. Coronal Mass Ejections

2.6.1. Coronal Holes and Solar Wind Sources

2.6.2. Plumes, Rays, and Jets

2.6.3. Solar Wind Observations

2.6.4. Solar Wind Modeling

2.6.5. Solar Wind Composition

2.6.6. Solar Energetic Particle Events

Preliminary, nay even premature, reports of a transit of HD 209458 by its planet (radial velocity companion) just made the time cut for Ap99. Detections of its subsequent transits have become commonplace, with ground‐based photometry reported by Henry et al. (2000), Charbonneau et al. (2000), and Jha et al. (2000) and multiple events extracted from the Hipparcos data base by Castellano et al. (2000) and Robichon & Arenou (2000). Indeed, slightly out of period, amateur astronomers have begun monitoring the transits. Analysis by the observers and by Mazeh et al. (2000) reveals that the mass of the planet is about 2/3 that of Jupiter and the radius rather larger, so that the density is only 0.3 g cm^-3 (even more of a floater than Saturn, if you could find a big enough ocean). A harbinger of things to come is that the transit made a detectable change in the profiles of absorption lines in the spectrum of the parent star as recorded by ELODIE (Queloz et al. 2000b). This means that we can look forward some day to information on the composition of the planetary atmosphere, although separating the reflected planet light with a space‐based interferometer would yield a cleaner planetary spectrogram than subtraction techniques.

Transits also happen in the solar system, as watched from Earth, with the next Mercurial one due in May 2003 (these are not rare) and the next Cytherean pair in 2004 and 2012. These are rare, and even we have only vague memories of the previous pair in 1874 and 1882.

Radial velocity monitoring continues to yield more (Queloz et al. 2000a), and more (Udry et al. 2000), and more (Vogt et al. 2000) of the class loosely described as hot Jupiters, including two with masses a bit less than that of Saturn (Marcy et al. 2000) and one for which the best‐fit mass is about 10 times that of Jupiter (Korzenik et al. 2000, with at least one still heftier one in the press release stage. Overviews of the Lick survey (Cumming et al. 1999) and others (Marcy & Butler 2000) reveal that about 10% of solar‐type stars have at least one companion of planetary mass, that the mass distribution is something like dN/dM∝M^-1, and that masses larger than 3 M_J in orbits with semi‐major axes of 4–6 AU are genuinely rare.

Full details of the three‐planet family belonging to Upsilon And have appeared (Butler et al. 1999), and you have all heard hints of additional planet pairs and multiples, besides, of course, the companions of PSR 1257+12, which number at least three (Wolszczan et al. 2000b) and perhaps four (Wolszczan et al. 2000a). Pulsars in general do not display infrared‐emitting protoplanetary disks. Thus either they already have planets (one example) or they never will (the other 900 and 99; Greaves & Holland 2000).

Other ways in which planets might reveal themselves include (a) distortion of the times of eclipses in close binaries (a gravitational effect, not an occultation one), with a tentative case made for CM Dra (Deeg et al. 2000), (b) a bit of reflected light appearing as a bit of Doppler‐shifted spectrum superimposed on the main, stellar one (Cameron et al. 1999, with an "interesting if true" example of Tau Boo), (c) a little twitch in the light curve of a microlensing (MACHO) event, with two tentative candidates, 97‐BLG‐41 (Bennett et al. 1999, where the lens was already a close binary system) and 98‐BLG‐35 (Rhie et al. 2000, an otherwise‐single star lens).

And, dropping rapidly down the probability curve, we find (d) a CH₃OH maser which Slysh et al. (1999) interpret as belonging to a planet in orbit around an O‐type star, (e) a bunch of stars surrounded by disks whose structures include central holes or gaps that might be due to the tidal effects of planets (Jourdain et Muizon et al. 1999; Weinberger et al. 1999; Dent et al. 2000), and (f) dust enough to make a throng of Kuiper Belt Objects like ours around 55 Cnc, Rho CrB, and HD 210277, though not around 51 Peg, Ups And, or Gl 876 (Trilling et al. 2000; Jayawardhana et al. 2000). Note, however, that these stars are all already known to have radial‐velocity companions of planetary mass, so we are either (a) cheating or (b) moving on to properties of host stars.

All or nearly all planet‐blessed stars, including the Sun, are (as you have heard before) somewhat metal‐rich by the standards of their locations and ages (Gimenez 2000). There are also probably mild anomalies in the amounts of carbon and sodium (Gonzalez & Laws 2000) but not in lithium (Ryan 2000). Gimenez (2000) also reports that the hosts of the planets with the smallest masses may be the most metal‐rich. Yes, he, and you, and we can all think of physical reasons that this might be so but also of selection effects that would produce the same impression. Apparently other host stars do not share the Sun's relatively low speed relative to the local standard of rest (Gonzalez 1999).

Stellar activity, because it can be expected to jiggle line profiles around, is a traditional confounding variable in planet searches. This is not a falsifiable hypothesis as it stands, but Henry et al. (2000b) report that four of a sample of nine known hosts also display activity‐related variability in the H and K features of Ca ii (most for the youngest hosts; Kuerster et al. 2000).

The conclusion that a bunch of faint, red things in the direction of the Trapezium (Lucas & Roche 2000) are truly orphan planets is dependent on being sure that you have allowed correctly for field stars, conversion of colors into temperatures, and conversion of luminosity and temperature in turn into masses (Hillenbrand & Carpenter 2000).

Twenty or so papers dealt with planet formation, orbit evolution, and so forth. But, as a 19th century cookbook is supposed to have said, "first catch your rabbit." According to Heacox (1999) and to Stepinski & Black (2000) this has still to be done. The radial velocity companions, whether their masses are those traditionally called brown dwarfs or called planets, are, so these authors say, part of a single population, statistically like close binary star companions and formed in the same way, not real planets at all. They define real planets by formation mechanism (out of protoplanetary disks) and chemical differentiation. Both are difficult to observe. Co‐planarity of several planets might also count and could be observable.

Assuming, however, that "extrasolarsystem planets" is not an empty set, we can pick up and carry on from last year's slogan, "Make and then migrate." Data continue to support presence of an excess of large dust grains in disks of protostars, relative to the general interstellar medium (Bouwman et al. 2000) and the ability of grains to aggregate and grow rapidly (Blum et al. 2000, reporting results from a microgravity experiment on the Space Shuttle).

Should one worry that some sets of models and initial conditions, including close binaries, make no gas giants at all (Papalòizou & Larwood 2000: Nelson 2000) or make them only with considerable difficulty (Ikoma 2000)? Not necessarily, since 90% or so of stars so far surveyed have no gas giants, at least up close. Indeed one feels that only one paper in 10 on theory of planet formation should report success. Studies of the medical literature show, however, that negative results of trials are less likely to be reported than positive ones (but see the beginning of § 9).

Instead, successes outnumber failures. Sasselov & Lecar (2000) have moved the "snow line" in and can make Jupiters at 1 AU. Well, all right, only one per star, we suppose, but Armitage & Hansen (1999) write that the first big planet will actually trigger formation of more, so that the natural outcome is several Jupiter‐plus masses in highly eccentric orbits. Will the product last long enough to be seen? More details are needed, but the specific circumstances of the three planets of Ups And give it permission to stick around for a gigayear (Rivera & Lissauer 2000) or more (Laughlin & Adams 1999), whether or not the orbits can be described by Bode's Law (Laskar 2000).

Lubow et al. (1999) offer an explanation for an upper mass limit to real planets of 6–10 M_J. It arises from the increased difficulty gas finds in flowing tidily across the gap tidally created by the planet as the planet and the gap both grow. They assume that the planet remains a fixed distance from its star. If it does not, then the migration process should throw out little bits of sublimed planetesimals, which might show up as transient absorption lines (Quillen & Holman 2000). And, of course, you must stop the migration (by having the planetesimal disk all used up) before all the planets hit their stars, or we will not be here to talk about it (Kley 2000).

A theorist's work is, however, never done. If planets form and persist, then we want to be able to account for their properties. Burrows et al. (2000b) match the low density implied for the transiting planet of HD 209586 with a body that is both hydrogen‐rich and heated by its star. Sudarsky et al. (2000) propose a five‐stage cooling (or heating) sequence, from cold ammonia (like Jupiter) through water clouds, neutral alkali metals, and methane, to silicate grains as the dominant source of atmospheric opacity. Those with water clouds and silicate grains high in their atmospheres will have the largest optical albedos. A given planet might pass through two or three of the stages as its parent star evolves. Planets at different distances from their hosts will fill the full sequence.

Ford et al. (2000) predict that planets should be common even among old stars, based on their model for PSR 1620−26 in the globular cluster M4. Just possibly, this prediction has already been falsified by tight upper limits to numbers of planets in small orbits around the stars of 47 Tucanae set by absence of evidence for transits by them (Brown et al. 2000).

Also still in the realm of the theorist are the chemical and physical consequences to a star of swallowing one or more planets (Siess & Livio 1999 and a number of other papers). Somewhere between imaginative and convoluted are the thoughts (a) that a diet of planets increases lithium abundance, not because there was lithium in the planets, but because the capture triggers the nuclear reaction He³(α, γ)Be⁷(e^-, ν_e)Li⁷ in the star (Denissenkov & Weiss 2000) and (b) that a planet belonging to AA Dor tried to swallow its star during common envelope binary evolution, rather than the converse (Rauch 2000).

Pluto (notice that the combination of this name and section heading is already a political statement) has ethane as well as methane on its surface (Nakamura et al. 2000). It also has a moon, whose surface composition is more uniform than Pluto's own. Pluto is in a stable, 3:2 orbital resonance with Neptune (Varadi 1999, on the stability, not the resonance, which may well have been known to Tombaugh).

Neptune probably migrated outward to its present location, given the large number of small objects in various resonances with it (Ida et al. 2000b).

Uranus is not an X‐ray source (Ness & Schmidt 2000). Neither for that matter is Neptune. Saturn was seen for the first time in the observations reported by Ness & Schmidt, but is nowhere near so bright as Jupiter.

Saturn has seasons, revealed by infrared spectra of molecules in its stratosphere. (Ollivier et al. 2000). They are unpleasant even by Minnesota standards ("July and winter"). Saturn also, above all, has moons (30 in the latest out‐of‐period rumor), but even the old, familiar ones do not have densities known well enough to look for a correlation between density and distance from the planet (Dourneau & Baratchart 1999) of the sort found for Jovian inner satellites.

Jupiter has a Great Red Spot, which is some sort of Rossby wave (Li et al. 2000c). Encrenaz (1999), in a review of the GRS and other Jovian surface features, says that it was discovered by Robert Hooke in 1664. The story is, of course, enormously more complicated, and we won't spoil it by condensation from the tale as told by Hockey (1999), and quote only one line, from p. 142, "No one questioned Virtue" (meaning the Rev. James Virtue). Other Jovian traits include absolutely oodles of weather, driven by moist convection, like Earth's, but with an internal rather than solar heat source (Gierasch et al. 2000) and abundances of argon, krypton, and xenon relative to iron which equal the cosmic values. This means that the planet must have formed from cold planetesimals, able to retain even these extremely volatile elements (Owen et al. 1999). It would seem that the disk‐instability model for the formation of Jupiter and other massive planets (Boss 2000b) would equally well, perhaps better, account for the presence of noble gases in cosmic proportions.

Jupiter also has at least its fair share of moons. Most discussed were Europa, with evidence for transient subsurface brine, the evidence coming from an induced magnetic field (Kivelson et al. 2000) and the heating from tidal dissipation along faults (Gaidos & Nimmo 2000), and Io. The latter was imaged and otherwise autopsied by Galileo in October 1999, with results reported in six papers beginning with McEwan et al. (2000). Highlights of the results include all sorts of lava fountains and flows, mostly mafic (low viscosity), some rather hellish colors attributed to sulfur compounds, and sporadic melting of snowfields (not H₂O snow, of course; Kieffer et al. 2000). Io sheds so vigorously that it can support a variable exosphere (Russell & Kivelson 2000) and aurorae (Trafton 2000) and still keep up the Jovian dust supply (Graps et al. 2000).

Mars has rocks, both andesite (like the Andes) and basalt (like the bas of the ocean), but the relationships among rock type, age, and elevation are different from the terrestrial ones (Blandfield et al. 2000) for reasons that are perhaps understood (Zuber et al. 2000). Mars is also very widely credited with enough water in the past to flow, erode, gully, and all (Malin & Edgett 2000; Head et al. 2000) though the flat northern plain may well be tectonic rather than plutonic (Stevenson 1999). But the Martian heresy of the year is undoubtedly the suggestion (Hoffman 2000) that all the sinuous channels, gullies, and other features now generally blamed on ancient water flows are really the result of cold, dry eruptions of gas, dust, and rock, fueled by exploding liquid carbon dioxide. Thus the Mars of the past would always have been cold and dry, like Mars today. The liquid is necessarily stored underground, at 35 atm or more pressure, and liberated by crustal collapse and, perhaps, impacts. The terrestrial analogues are pyroclastic flows, like those triggered by volcanic eruptions in Montserrat in the 1990's and at Mt. St. Helens in May 1980.

Earth turned up in so many contests, most of them only very remotely astrophysical, that it gets its own section, and we note here only that it has been impact‐cratered at a rate of about 10^-14 km² yr^-1 over the past 125 million years (Hughes 2000) and has had fish for some 600 million years (Shu et al. 1999; Janvier 1999). Putting these together, we calculate that fish collectively could have seen the formation of a total cratered area of 6 m² and deduce that we have not properly interpreted at least one of the results.

Venus is skimmed by the crescent moon every synodic period of the pair (a moonth or so), but only rather rarely close enough and high enough in the sky to conduce to astrology, and even then you must be in the right time zone, weather conditions, and so forth. December 29, 2000 was roughly the third best the more Cytherean author has seen and reinforced the impression that such a conjunction is at least as likely a source of star‐and‐crescent images as supernova 1054.

Mercury was never completely mapped by Mariner 10. Ground‐based observers are now filling in the gaps with I‐band images taken at the Mt. Wilson 60^'' (Dantowitz et al. 2000; Baumgardner et al. 2000). The data are analyzed using "selective image reconstruction," which seems to mean using only the sharpest of many very short exposures. The features being found correspond reasonably well with those implied by radar mapping. The authors provide the additional information that the phases of Mercury were discovered in 1639 by Zupux and again in 1644 by Hevelius, while Mercury itself was discovered in 265 bce by Timocharis. Having always credited Og Troglodyteson, we are tempted to quote a cousin (Farmer 2000) who says that the ancient Greeks often claimed to be the first to have done something, when really they were just the first to write about it.

Another Planet X (this one must be about XVIII) has been deduced from comet orbits by Murray (1999). It has a retrograde orbit with a period of 5.8 Myr and must have been captured relatively recently (though presumably not since the last announcement of a 10th planet).

A whew of relief was heard all the way to Eros and back on 15 February 2000, when NEARShoemaker, back for second try (Ap99, § 3.3.3), swooped in close enough to learn that the asteroid is 34 × 11 × 11 km in size, essentially saturated with 1‐km craters (but no fish), non‐precessing, chondritic on its surface, not quite homogeneous but also not globally differentiated despite some melting, internally porous (10%–13% empty space, leading to an average density of 2.67 g cm^-3), and easy to get away from with escape velocities of 3–17 m s^-1. The set of encounter papers appears in the 21 September issue of Science, beginning with Yoemans et al. (2000).

Other near‐earth asteroids are only about half as common as previously advertised, with 500–1000 larger than 1 km in diameter versus 1000–2000 (Rabinowitz et al. 2000). Most should be found by 2020, and, by way of calibration, the object that hit Tunguska in 1908 (a very near earth asteroid) was about 0.07 km across. Bottke et al. (2000) point out that the asteroids so far uninventoried are likely to have orbits of large eccentricity, large inclination, or both, and will be hard to detect, but also somewhat less likely to hit Earth.

Trojan asteroids and quasi‐satellites are in quasi‐stable orbits. The former are known to exist, leading or trailing Jupiter around the Sun for a few million years and then departing to be replaced by others (Tsiganis et al. 2000; Christou et al. 2000). The latter librate around the longitudes of their associated planets in liver‐shaped orbits for thousands (near Jupiter) to hundreds of millions (near Neptune) of years (Wiegert et al. 2000). None are known, but we aren't all that fond of liver anyhow. In case it should ever come up, Thersites (1868) was the ugliest Trojan. It probably came from eating too much liver.

"Trans‐Neptunian Objects" tells you where they are. "Kuiper Belt Objects" tells you who thought about them (and perhaps the sort of axial ratio to expect, if you are old enough to remember Gerard P.). "Plutinos" tells you that some of them are in 3:2 orbital resonance with Neptune, as is Pluto (Yu & Tremaine 1999). "Centaurs" are another orbit family, not concentrated in the direction of the constellation Centaurus (or even Equuleus), but lying between Saturn and Neptune. KBOs are supposed to be the reservoir from which low inclination, short‐period comets arise (which in turn tells you that they are icy more than stony or metallic, like proper asteroids). Though the first one was found as recently as 1992, numbers have grown to permit statistical studies of both orbits and surfaces.

Orbit studies logically start with formation, and there is apparently no difficulty in making the (highly extrapolated) known numbers and sizes if you start with a protoplanetary disk mass a few times the minimum needed to make just the major planets (Kenyon & Luu 1999) The (again highly extrapolated) total mass of KBOs is comparable with that in the main asteroid belt (Trujillo et al. 2000, who also define some orbital subtypes without mythological names). The best estimate for the distribution of KBO sizes may come from craters on the surface of Triton. As you might expect, little ones are commoner than big ones, with dN/dr∝r^-3 (Stern & McKinnon 2000). Pluto continues to affect the orbits of the other Plutinos, introducing gaps in the distributions of orbital eccentricity and inclination at its own values (Nesvorný et al. 2000), from which you may deduce that Pluto itself is not just another Plutino.

KBO surfaces, as well as those of the smaller moons of the outer planets, Charon, and Chiron are frequently icy, but with a wide range of colors (visible to infrared) indicating a range of surface compositions and weathering patterns. There is no obvious bimodality or other discrete subset structure (Brown 2000; Luu et al. 2000; Brown & Calvin 1999; Barucci et al. 2000: Noll et al. 2000).

Ordinary main‐belt asteroids are the rocky and metallic ones (as their meteoritic fragments have been telling us for as long as it has been respectable to believe in "stones from the sky"—Jefferson did not). Surface studies and densities largely confirm what had been supposed. The bigger ones (Ceres, Pallas, Juno, and Vesta) have the infrared signatures of surface silicates (Dotto 2000). Of these, Ceres is the least dense, at 2.24 ± 0.04 g cm^-3 (Michalek 2000). Vesta and Pallas are both probably denser than 3 g cm^-3, but the masses are poorly determined because none of the largest asteroids has had a really close encounter. Hermione and Psyche did, and both claim to have densities near 1.8 g cm^-3 (Viateau 2000), reasonable for the rocky Hermione, less so for the supposedly metallic Psyche. A second asteroid with a moon, 45 Eugenia, revealed a density near 1.2 g cm^-3, close to the value found for the first case, Mathilde (Merline et al. 1999). The buzz word was "rubble pile", also mentioned by Ostro et al. (2000) for 216 Kleopatra (radar reflection data) and by Leinhardt et al. (2000), who note that such structure makes it easier to understand both how asteroids and planets were assembled in the first place and the characteristic "dog bone" shape, like that suspected for Kleopatra. The proper phrase might be soggy rubble pile, according to Zolensky et al. (2000), describing one meteorite that seems to be briny. Analysis of the dissipation of wobbles, on the contrary, shows asteroids to be drier than the Earth (Efroimsky & Lazarian 2000).

Asteroid orbits are subjected to a great many random perturbations. Vozikis et al. (2000) have applied a chaos detector to some of them but have not actually detected any chaos in their "power spectrum of geodesic divergence". In case you might have forgotten, Piazzi was the first to spot an asteroid, Ceres, on 1 January 1801 (sounds of centennial trumpets and krummhorns), and Gauss is generally said to have invented the method of least squares in order to extrapolate its orbit from inconsistent observations.

The asteroid‐comet distinction is not a clean one. Comet 107P Wilson‐Harrington is also minor planet 4015, and comet 95P Chiron is also minor planet 2060. Chiron was initially an asteroid (Centaur) that later displayed cometary activity, while Wilson‐Harrington, a catalogued comet, was later independently discovered temporarily embarassed by loss of its coma. Other examples exist (Toth 2000). The topic merited a brief, cogent review (Yoemans 2000), whose discussion of objects with the orbit of one type and the morphology of the other is a grave temptation to citing of James (1611).

Comet inventories are not so complete as we have been accustomed to suppose. SoHO found one by chance, post perihelion, that probably peaked near m = +11 (just a tad faint for amateur discovery), wasting its photons on the desert air. Discoverer (Makinen et al. 2000) and commentator (A'Hearn 2000) suggest that similar comets may be about as common as near‐earth asteroids, but that search strategies are much less advanced, and many are being missed, especially those with out‐of‐ecliptic orbits.

A number of dynamical issues remain of interest, including non‐gravitational effects on comet orbits (Festou & Barade 2000) and families of short‐lived orbits tied to planets (Drobyshevski 2000b) in somewhat the same way that the Trojan asteroids are tied to Jupiter. But if you had a comet in your lab, what you would really want to ask it is, "How far has your chemistry gone towards complex molecules?" A recently‐selected NASA mission, DeepImpact, won't quite bring the samples home, but it is scheduled to plunge into a comet and kick stuff out of the interior on Fourth of July some years downstream. Meanwhile, both Hale‐Bopp (Ziurys 1999) and several others (Wyckoff et al. 2000) have revealed carbon and nitrogen with isotopic ratios very close to the solar system average, implying that comets can provide information about the very early initial conditions for prebiological chemical evolution.

Hale‐Bopp, by the way, really had a double nucleus (Marchis et al. 1999), Hale and Bopp we presume, but are not sure which is which. Mini‐comets, made mostly of ice, have a long and bumpy history (of which Frank et al. 1986 is an intermediate protruberance). Bronshten (2000) has in mind a different, less volatile class of mini‐comets, whose members disrupt lower in the atmosphere (50–120 km up) and are the cause of fireballs. Tunguska, at 80 tons, was, in his view, one of the larger ones. This is, you will note, a good deal less massive than you would expect for the 70 m diameter mentioned above for Tunguska as an asteroid shard.

Many comets have dust tails wagging years behind them in their orbits, even poor old comet 15P/Finley, which has been turning into an asteroid for the past 300 years (Beech et al. 1999). These make meteor showers if we happen to cross through them (Jenniskens & Betlem 2000 on 55P/Tempel‐Tuttle and 109P/Swift‐Tuttle) and lead us inevitably to the next topic (Hughes & Williams 2000 on comet and meteoroid stream orbits).

The Leonids of November 17–18 were not only a better show in 1999 than in 1998 (Watanabe et al. 2000 vs. Ap99, § 3.3.2), they were better than can be entirely understood, with light trails up to 2 km wide (LeBlanc et al. 2000). Also more and better than expected were the flashes when some hit the Moon (Ortiz et al. 2000, who also use the absence of such flashes at other times to say that there are not a lot of mini‐comets wandering around our neck of the solar system). Some commentators have expressed reservations about the nature of these flashes (e.g., Weissman 1999, who suggested space debris or satellites reflecting sunlight). With data in hand from 1998, 1999, and by now 2000, predictions for the new few passages through the Leonid stream have naturally improved. One change is an increase in the best value for density of the fragments from 1 to 4 g cm^-3, and it will hardly be worth going outdoors in November 2001, at least for the meteors (Goeckel & Jehn 2000).

Some other meteors and, especially, meteorites constitute samples of which we can again ask, "How far has your chemistry gone?" Far enough that you need both grain‐surface and aqueous reactions to make everything you see in Murchison and Orgueil etc., say Sephton & Gilmour (2000).

We live, it seems, only 0.05 pc from the edge of a typical (T ≈ 7000 K, n_H ≈ n_e ≈ 0.1 cm^-3), small (93 pc³), warm interstellar cloud, and will depart from it in about 3000 years (Redfield & Linsky 2000). The authors also raise the question of whether such mean cloudlets are permanent entities or transient shock features or turbulent eddies (Linsky et al. 2000). The next cloud we enter will be the one now inhabited by Alpha Cen A and B.

Not much closer to home is the heliopause, toward which Voyager 1 and 2 struggle ever onward (most recent ETA 2020; Ferlet 1999). This is, we keep telling ourselves, no further ahead than the Patras IAU (which was just yesterday) is behind. What is more, models of the interface between solar wind and local interstellar medium are rapidly acquiring the status of testable predictions, at least relative to the first one (Munch & Unsold 1962, according to Linsky et al.). In a recent model, Fahr et al. (2000) find they need to include effects of five fluids, solar wind, local ISM, pick‐up ions, Galactic cosmic rays, and anomalous cosmic rays. Data on the interface seen as Lyman alpha absorption lines in stars yield a temperature for the neutral hydrogen wall of 38,000 K (Wood et al. 2000), which is a funny temperature for neutral H!

Cosmogony for many decades meant specifically the effort to understand formation of the solar system. This task has, in recent years, been largely subsumed as part of the study of star and planet formation in general. Thus we find "make and then migrate" scenarios for our system as well as others (Thommes et al. 1999).

Unique, however, to the solar system is the evidence provided by fossil or extinct radioactivities. These are the decay products of finite‐lifetime nuclides that are found in chemical contexts indicating that they must have been live when a dust grain or larger entity solidified. Chemically‐distinct inclusions in meteorites, especially carbonaceous chondrites, are the prototype sites, and Al²⁶ (now Mg²⁶) and I¹²⁹ (now Xe¹²⁹) are two of a dozen or more examples. Because some of the half‐lives are quite short, the implication is either that stuff that was gaseous when the solar system acquired its identity solidified in a hurry (e.g., less than a million years) or that many "pre‐solar" grains survived to be incorporated whole and unmelted into the meteorites, or, of course, both. The present interplanetary dust includes (still, or again, if knocked off asteroids and all) grains with isotopic anomalies in hydrogen and nitrogen, suggesting pre‐solar origin (Messenger 2000). Because these are not decay products, no particular time scale is implied.

The author whose father was a chemist has been predicting for several years that eventually this whole topic will be understood well enough that you could remember what all the various radioactivities and anomalies and their meanings are just by looking at a periodic table. This has not so far happened (see Arnould & Prantzos 1999 for an overview). What index year 2000 did see was (1) evidence for short‐lived radioactivities and grain heating being quite inhomogeneous through the solar nebula (Meibon et al. 2000 on heating; Guan et al. 2000 on chemical inhomogeneities; Lugaro et al. 1999), and (2) evidence suggesting that the various fossil radioactivities originated in more than one sort of site and process. Contributors include (a) Type II supernovae, the only place you are likely to get the p‐process nuclide Nb⁹² (Yin et al. 2000) and two other p‐process chronometers, all of which suggest 10 Myr of decay between synthesis and solidification, and Mo^97,98 incorporated into grains immediately after their synthesis in the r‐process (Meyer et al. 2000a), (b) asymptotic giant branch stars with a range of masses so as to make SiC grains with a range of ratios of Si²⁸:Si²⁹:Si³⁰ (Lugaro et al. 1999), (c) irradiation of the solar nebula itself, since Be¹⁰ can be made only by spallation (McKeegan et al. 2000) and it was live when Allende formed; this irradiation must have come from the Sun itself, as opposed, for instance to a nearby supernova (Hua et al. 2000), and (d) dust melting provided by a nearby gamma‐ray burst close to the time of solar system formation (McBreen & Hanlon 1999).

Was there a particular AGB star, supernova, or even gamma‐ray burst whose impact triggered the collapse of the cloud that became our solar system? Stock in this hypothesis has risen and fallen many times since Opik (1953) first said yes. Vanhala & Boss (2000) conclude that such a scheme remains possible, with the newly‐made atoms carried in on the wing of a Rayleigh‐Taylor instability, but the triggering event must have happened closer to us than 0.1 pc (for an AGN star) or a few parsecs (for a supernova).

Most distracted of all are our own globe and its inhabitants. This section contains some items published during the year about the Earth as a terrestrial a planet and as an (the?) abode of life and some otherwise unclassifiable papers that transcended even the elastic limits of § 13.

3.6.1. Earth, Water, and Fire

3.6.2. Air

Asymptotic giant branch stars are to horizontal branch (or clump) stars as ordinary red giants are to main sequence members. That is, (a) they come after, (b) they last about 10% as long, and (c) they are fueled largely by fusion in a shell of the same stuff that fused in the core before (helium or hydrogen). In addition, their luminosity and other properties are mostly a function of the mass of the core inside the fusion shell (Paczynski 1970; Refsdal & Weigert 1970; Uus 1970). Unique to the AGB is the extreme thinness of the burning shells, which causes them to flash on and off, driving convection zones that chase each other back and forth across the star and eventually pollute the surface with production of the interior reactions (Iben & Renzini 1983), including nitrogen, carbon, s‐process products (Lebzelter & Hron 1999 on Tc), and lithium, to the Galactic supply of which they make a significant contribution according to Abia et al. (1999), Ventura et al. (1999), and Polosukhina et al. (1999).

Within living memory, the rest was simple. You blinked briefly while the extended, cool envelope lifted off (its total energy is positive anyhow, allowing for ionization), and soon you had a spherical or ringlike shell of planetary nebula expanding at 10–30 km/sec around, and ionized by, a pre‐WD core. And the core cooled down toward the white dwarf regime on about the same 10⁴ yr time scale as the nebula dissipated.

When the curtain rises on the new, more complex Y2K scenario, AGB stars already have winds blowing, as indeed do stars all their lives. The AGB ones are cool (≲100–200 K) and slow, as you would expect from atmospheric conditions (Crawford & Barlow 2000), and carry off material at rates that are largest for the most massive stars at the latest phases, up to 10^-3 M_⊙ yr^-1 (van Loon et al. 1999). Since there is something like a solar mass to be got rid of, we are not surprised to hear that statistics imply an average duration for this superwind phase of 3700 years, about half of which is spent as an OH/IR star (that is, with OH maser emission; the IR part the poor star cannot help; Lewis 2000).

But mass loss is by no means steady. First, these stars are all pulsators (Miras, semi‐regulars, RV Tauri stars and all). The pulsation is a wind‐driver (Schroeder et al. 1999) with the superwind turning on (for LMC composition and masses) as the period lengthens to about 500 days (Nishida et al. 2000), though of course the period shortens again (and mass loss continues) as we see down to hotter photospheres (Barthes et al. 2000 on HD 56126). The pulsation (dynamical) times are too short to show directly in later nebular structure.

Remember, however, those shell flashes. They too contribute to mass ejection, and the times between them are long enough to yield discrete shells visible later on in the PNe. U Cas probably had its first flash about 800 years ago (Lindqvist et al. 1999). The post‐AGB star IRC +10216 has lots of shells, implying mass loss modulated at 200–800 yr intervals (Mauron & Huggins 1999). TT Ari is somewhere in between, still visible in the visible (you can tell from the name) but with multiple CO (cold) shells around. One implication is that there should already be lots of gas and dust outside those shells, left from an earlier slow wind (Steffen & Schoenberner 2000).

Multiple‐shelled PN NGC 6891 is a fairly typical end‐product, having kinematic ages consistent with interpulse times (Guerrero et al. 2000a; Phillips 2000b on other late He‐flash double shells and halos). Soker (2000) has advanced a competing hypothesis, that multi‐shelled PNe and proto‐planetaries are evidence for solar‐type cycles with periods of 200–1000 years.

Early shell flashes will change the outer star only slightly, for instance the recent period decrease in the Mira S Sex (Merchan Benitez & Jurado Vargas 2000). But a very late, last thermal pulse can return a blue star to the AGB, drastically modify its surface composition, and set the stage for R CrB‐type fadings caused by carbon dust (Herwig et al. 1999). The prototype is FG Sge (Ap99, § 5.4) and the most recent example is V4334 Sgr (Sakurai's object). V4334 is extreme by any standard—its evolution more rapid than that of FG Sge and its first fading (11 magnitudes in V) large even by R CrB standards, though naturally the energy eventually sneaks out as infrared (Kerber et al. 1999; Duerbeck et al. 2000; Tyne et al. 2000; Kipper 1999; Pavlenko & Yakovina 2000). Before looking again at the ejecta we can follow at least a subset of the stars evolving through a series of phases defined by their prototypes, FG Sge, RY Sgr, R CrB, UV Cam (which has stopped drastic fadings since the early 20th century) to XX Cam, which no longer even sports an IR excess (Doroshenko et al. 2000).

Now about those nebulae. A proper PN is, of course, an emission line source, meaning that the gas has to be ionized. Since the ejected shells contain CO, the gas must first be dissociated. That process can be tracked in the C i/CO ratio (Knapp et al. 2000). By the time H₂ molecules are hot enough to be seen in emission, non‐spherical shapes have already been established (Hrivnack et al. 1999; Garcia‐Lario et al. 1999). Ionization, of course, comes afterwards, but so quickly that H₂ remains just outside the ionized region, tracking the twisted and bipolar morphology of the ionized part (Guerrero et al. 2000b).

Bipolar or the opposite is as simple as PN shapes get. Ueta et al. (2000) report that zero of 21 proto‐PNe imaged with HST are spherical. They deduce a correlation with progenitor mass, in the sense that low mass cores have slow enough mass loss that we can see the central star and elongated emission from an equatorially‐concentrated superwind, while higher mass ones cloak their central stars in ejecta of sufficient mass that the fast stuff can get out only along the polar directions and we see bipolar morphology (an astrophysical shape common enough that it desperately needs a less pompous name; there will be a small prize for the best suggestion). This mass correlation has a nice, logical sound to it, and Stanghellini et al. (2000) concur for an LMC sample that bipolar shape is a signature of a massive progenitor as deduced from the N/C ratio and closeness to the Galactic plane. Phillips (2000a) disagrees, saying that the axial ratios of PNe do not much depend on initial mass, or, indeed, on interactions with interstellar material.

Wind composition also matters, because carbon dust increases the efficiency with which radiation pressure pushes on the ejecta, leading to larger outflow velocities for carbon‐rich Miras than for oxygen‐rich ones (Groenewegen et al. 1999). Remember, however, that it is easier to make a carbon star if total metallicity is small to start with (because there is not so much oxygen to swamp with carbon made from the triple‐alpha reaction). Indeed, it is possible to reach the proto‐planetary stage with O/C still >1 only for initial metallicities larger than those characteristic of the solar neighborhood (Leahy et al. 2000). Not surprisingly, larger initial metallicity eventually results in a larger dust/gas ratio in AGB winds, whatever the dust is made of (van Loon 2000). The assortment of s‐process products seen at the post‐AGB stage also depends on initial metallicity, but this is part of a different story (van Winckel & Reyniers 2000).

As you would expect, when hotter layers with larger escape velocities are uncovered, wind speed increases, leading, inevitably to two‐speed models and colliding winds. BD +30°1639 is the best case so far for X‐rays coming from slow wind material shocked by the fast wind, as opposed to a coronal or jet source or a very hot star (other reasons for PNe to be X‐ray sources; Leahy et al. 2000).

The prize for maximum complexity during the transition from AGB to PN goes, uncontested, to V Hya, still near the beginning of the process at T_e = 2650 K. It is a binary (P = 17 yr) with both fast and slow winds, and the cool star is also a carbon‐rich Mira (P = 530 days; Knapp et al. 1999).

Even it will have to work hard to match up to the multiplicity of structures seen in more highly evolved objects, including M2‐9, which is a rotating corkscrew with a period of 20 yr, according to a data stream begun by Minkowski (1947) and completed by Doyle et al. (2000). K4‐47, which appears to consist entirely of core and jets and blobs, perhaps arising from a binary plus precession (Corradi et al. 2000) and He 2‐47 and M1‐37, each of which has seven or eight dramatically young blobs or multiple jets (Sahai 2000), add to a picture in which an AGB star can become any of a wide variety of shapes and sizes of nebula as it dies.

If there is a white dwarf angel when we get wherever we are going, she is going to have to answer a hand's worth of questions. (1) What is the maximum main sequence mass that leaves a white dwarf, and how does this depend on initial composition, rotation, and whatever else? (2) How do mass loss, gravitational settling, dredge‐up, accretion, and residual nuclear reactions combine to yield the full range of surface compositions found, and how many trajectories are there through "surface composition space?" (3) Are present calculations of cooling rates good enough that the sparcity of very faint disk WDs tells about the onset of star formation in the Galactic disk, and can this be extended to globular cluster and field halo stars to set limits to the age of the universe? (4) Do all white dwarfs rotate as slowly as the subset for which we have rotation periods from narrow line profiles or shifting magnetic patterns (e.g., more than 100 years for GD 209 and G840‐23, Berdyugin & Piirola 1999; 18 yr for R Aqr, Hollis & Koupelis 2000)? and (5) What is the physics responsible for the very wide range of WD magnetic fields (measured at 3 × 10⁴–10⁹ G for 25% of single stars, with the rest smaller), and why is the range smaller, only 10⁷–3 × 10⁸ G in close (cataclysmic) binaries?

None of the more than 50 indexed white dwarf papers entirely answered any of these questions, so here are some partial answers, followed by sometimes more complete answers to questions we hadn't actually asked.

Starting, back to front, with magnetic fields and rotation, Wickramasinghe & Ferrario (2000) conclude (a) that the distribution of rotation periods is bimodal, with merger products responsible for the fast ones, and (b) that strong fields in single stars descend from both Ap/Bp stars and mergers. Incidentally, the stars with large fields are more likely than others to have atmospheres with hydrogen and helium mixed and to have larger masses than the weak‐field stars. The commonest topology is an off‐center dipole.

Cooling calculations cannot yet be subjected to the test of whether they predict the correct changes for pulsation periods, and though they will be, only the hot, young end of the track will be thus checked (Kepler et al. 2000). Meanwhile, isochrones are getting better. Richer et al. (2000) include the effects of non‐gray atmospheres, and some extra ultraviolet opacity in hot DA atmospheres comes from iron (Chayer et al. 2000). Where the iron comes from is left as an exercise for the reader. Down inside the WD, a traditional uncertainty is how much energy is released (and, therefore, how much the life is extended) by separation of carbon and oxygen during core crystallization. Cyrstallization itself locks up energy in zero‐point fluctuations and so shortens cooling times. The answer seems to be that phase separation is a 10%–20% effect (Isern et al. 2000) delaying coolth by not much more than a Gyr at most (Montgomery et al. 1999). Small amounts of residual fusion will, of course, also extend cooling times, and are not as easy to rule out as you might think (Napiwotzki 1999; Sarna et al. 2000).

White dwarfs with iron cores cool much faster than the CO ones that are the baseline models, but unless they are a good deal more numerous than the three found among Hipparcos targets, the total effect on luminosity distributions and age estimates will be modest (Panei et al. 2000). The author who has done more ironing remains surprised that there are any of these at all. The three are GC 140, EG 50, and Procyon B, with a mass of 0.65 ± 0.15 M_⊙ determined independently from its orbit (Girard et al. 2000) and from its spectroscopic surface gravity.

The ultimate test of white dwarf cooling theory will, of course, be the faintest stars in the globular clusters. But, until a larger, spacier telescope than HST comes along, M67 remains the best check on whether white dwarf cooling ages equal main sequence turnoff ages. "More or less" say Richer et al. (2000), 3–4 Gyr from the WDs versus 4 Gyr from the MS.

The answer to the questions about surface composition seems to be that both the stars and the theorists have to work fairly hard. A quick recap: DO's are very hot and display He ii features. DA's are pure hydrogen (at least until you look very carefully). DB's display neutral helium. And anybody with a Z in his name also has heavy elements in the atmosphere. DC means continuum only (not carbon) and is thought to imply a cool helium atmosphere. OK.

Gravitational settling begins and He/H starts to fall when the star fades below about 3000 L_⊙, while a planetary nebula is likely still to be visible (Driebe et al. 1999). Complete elimination of helium from the photosphere (assuming there is hydrogen to float) is inhibited by wind as weak as 10^-12 M_⊙/yr and is completed near log g = 7 (Unglaub & Bues 2000, who also conclude that there must be at least two separate trajectories through the white dwarf composition space from the beginning). That is, DAO and DA stars must come from precursors that have hydrogen in their photospheres at all stages, while the very hot PG 1159 stars (with log g already = 7.5–8.0) feed into the DO and DB territories. The separation is, it seems, normally complete before the star is compact enough to count as a real white dwarf, in the sense that 5 DBAs have recently been reclassified as various sorts of subdwarfs (Bergeron et al. 2000).

Accretion of interstellar material is now part of the received doctrine for elements heavier than carbon and oxygen in white dwarf atmospheres (Friedrich et al. 1999 on three DBAZs where all but a millionth of the hydrogen has already settled out of the atmosphere, despite the accretion being recent and rapid, and Koester & Wolff 2000 on a couple of DZAs for which log H/He is still as large as −3). The metals, however, still do not look quite like what one would have expected from accretion. The spotty distribution (in both angle and height) of silicon in the hot DAZ GD 394 (Dupuis et al. 2000) naturally prompts one to say "magnetic fields of complex topology," while the excess of nickel over iron in the DO REJ 0503−289 (Barstow et al. 2000) brings an equally immediate "ummm?"

Coming at last back to the issues of mass and statistics, we recall that previous editions of Ap9x have caught assorted calibrations of the WD/NS mass cut and apparent statistical discrepancies among the death rates of AGB stars, the birth rates of WD, and the throughput of PNe. This year we noted only (a) that there are a lot of white dwarfs, 121 within 20 pc, which is still incomplete beyond 15 or 16 pc (Tat & Terzian 1999), (b) that the average (commonest, mode) mass is 0.58 M_⊙, but quite a few reach to 1.1 M_⊙ and beyond (Vennes 1999), (c) that the average mass of the nuclei of planetary nebulae in the bulge is, at 0.61 M_⊙, a tad larger than you would expect for the population (Gesicke & Zijlstra 2000, and no, you cannot yet see the white dwarfs there), and (d) that Lambda Sco, the most massive star (B1.5 IV) with a white dwarf companion, must be pushing upward on the WD/NS dividing line as it goes around every 5.96 days (Berghoefer et al. 2000b). Three less extreme BV+WD systems are currently known (Burleigh & Barstow 2000).

Now a few questions that we didn't think to ask. Are white dwarfs the accretors in anomalous X‐ray pulsar binaries? Maybe (Geroyannis & Papasotiriou 2000). Are they the accretors in supersoft X‐ray binaries? Yes (Reinsch et al. 2000), if you don't insist on a perfect fit to the spectrum (Shimura 2000), and provided that you define the class to exclude X‐ray sources with spectra that you might call supersoft but that are known to be other things (Greiner 2000, a catalog).

And finally, how many white dwarfs does it take to change a light bulb to make up a triple? Three, and they are called 1704+481, Sanduleak B, and GR 577 (Maxted et al. 2000).

Not being sure what the important questions were, we sorted through the 100 or so indexed neutron star papers looking for what colleagues had said were interesting answers, and came up with the following.

Magnetic fields. Well, we aren't any of us getting any younger, and there was general agreement that neutron star fields do eventually die away, get expelled, or whatever, though it may take 10⁸ yr (Konar & Bhattacharya 1999; Makishima et al. 1999; Konenkov & Geppert 2000; Jahan‐Miri 2000). It took only 10⁶–10⁷ yr when we too were younger. The process apparently starts with the field re‐orienting to become more nearly perpendicular to the rotation axis or with non‐perpendicular moments dying off first, as the period slows to 0.1 sec (Malov 1999). And, as you drivers of previously‐owned pulsars are aware, recycled ones are just not the same (Possenti et al. 1999).

In mild dispute, however, is the prevalence of fields in excess of something × 10¹⁴ G among the soft gamma repeaters and anomalous X‐ray pulsars (Ap99, § 7). To continue the discussion, Harding et al. (1999) find that a strong field model of SGR 1806–20 is not self‐consistent unless the emission and spin‐down are both episodic, and Chatterjee et al. (2000) have again presented a fossil accretion disk model as an alternative for the X‐ray pulsars with P = 6–12 sec . On the strong magnetic field side of the court we find Gogus et al. (1999 concerning SGRs) and Kaspi et al. (1999 on the timing of anomalous X‐ray pulsars). Association of both sorts of neutron stars with young massive stars (Vrba et al. 2000) and supernova remnants (Gaensler et al. 1999; Corbel et al. 1999) are, we think, just evidence that, whatever they are doing, they haven't been doing it for very long.

Radio‐quiet, unpulsed, and other oxymoronic pulsars. No, we can't claim that these make up a well‐defined set or that, even if they did, we could put them into an evolutionary sequence. The game plan, however, is Cas A first, with a few other young supernova remnants, a few strange non‐pulsars which arguably are neutron stars, and finally what are called pulsar wind nebulae (the words are equally informative in any order, with nouns and adjective appropriately adjusted to nebular pulsar winds, windy nebular pulsars, and so forth).

Non‐pulsar of the year was the compact X‐ray source imaged by Chandra near the center of supernova remnant Cas A, whose discovery just made the cut for Ap99 (§ 5.2). The initial theoretical gloss was that the emission was difficult to understand as that of a neutron star dating from 1650–90. Not surprisingly, ingenuity has triumphed, and emission either from a cooling neutron star or from a residual accretion disk around a black hole now seems possible (Pavlov et al. 2000; Umeda et al. 2000).

Hughes et al. (2000) show the Chandra image and point out that, not only are there iron‐rich knots of stuff from the silicon burning shell of the progenitor star, but they are outside the silicon‐rich material, overturned by the action of neutrino‐driven plumes of material rich in (then) Ni⁵⁶, as must also have been the case in SN 1987A for the gamma rays to get out as quickly as they did. (See Lawrence et al. 2000 for an update on the ejecta of 1987A and Middleditch et al. 2000 for a report, not the first, that its remnant is no longer a radio‐silent pulsar.) Returning to Cas A, data from the Infrared Space Observatory also show that the mixing of the layers of the parent star occurred on a macroscopic scale, leaving knots of O‐burning stuff, C‐burning stuff, and so forth (Douvion et al. 1999). Hwang et al. (2000b) show maps in the X‐ray lines of Si, S, Ar, Co, and Fe. All but the iron one look a lot like the optical image you are used to. Other relatively young supernova remnants with evidence for element segregation include Vela (Tsunemi et al. 1999), the Cygnus Loop (Miyata & Tsunemi 1999), and SN 1006 (Vink et al. 2000).

A last‐Cas update on the progenitor: Sad to say, the copy of Flamsteed's Atlas Coelestis in the Zinner collection at San Diego State University does not show his rogue star labeled, "This otherwise unknown star may have been the supernova progenitor of radio source Cas A." Indeed it doesn't even show Tycho's star, B Cas, progenitor of Tycho's SNR, though other atlases do, including Tycho's own posthumous Astronomiae Instauratae Orogymnasmata. Incidentally, Tycho numbered the stars neither from bright to faint, east to west, nor north to south, but from important to unimportant body parts. Thus 1 and 2 are head and heart, 4 the private parts, 6–10 hands and feet, 21 the belly button, and 23–26 the Arundinis in her hand. (It means reed, but we had to look it up, so why shouldn't you?)

The X‐ray compact core of SNR G347.4−6.5 is unpulsed (Slane et al. 1999) and so presents the same choices as the core of Cas A, though at less good angular resolution. The core of radio SNR PKS 1209−51/52 appears compact in the Chandra image. It is radio‐silent, but X‐ray pulsed at P = 0.424 sec (Zavlin et al. 2000). Whatever the energy source for these (residual heat, left‐over accretion disks, or rotational kinetic energy), when they fade, they fade very thoroughly. That is, even deep surveys reveal very few old, isolated neutron stars as X‐ray sources (Treves et al. 2000). One might have expected at least a bit of accretion luminosity when they pass through denser bits of interstellar medium, but apparently they are rolling too fast to gather much moss (Popov et al. 2000).

Two more classes that are not pulsars: (1) any of the first 92 FIRST sources with linear polarization in excess of 5% (Crawford et al. 2000) and (2) a bunch of unidentified EGRET sources, although the argument is largely statistical, since EGRET error boxes are large (Oka et al. 1999; Zhang et al. 2000a). The unidentified gamma‐ray sources with variability on time scales near one day are probably not pulsars either (Wallace et al. 2000).

This brings us to "pulsar wind nebulae," meaning filled‐center sources of synchrotron radio and/or X‐rays resulting from the confinement of a relativistic pulsar wind by external pressure (Gaensler et al. 2000). These do not require the pulsar to be aimed at you and so, if you look really hard and see no nebula, then there is no pulsar or no confining material. A surprising absence is that of the X‐ray core of SNR Pup A, which is an X‐ray pulsator at 75 msec (Zavlin et al. 1999), and has plenty of stuff around, but no radio synchrotron (Gaensler et al. 2000).

Older pulsar wind nebulae are also rarer than expected. Stappers et al. (1999) looked at five pulsars with P = 0.06–0.2 sec, rediscovered one PWN, did not confirm a second, and set three tight upper limits. Becker et al. (1999) similarly report that several ASCA candidates (including Geminga) for PWNe were false alarms.

Masses. The smallest you are likely to be able to form in a Type II supernova is 0.9 M_⊙ (Strobel et al. 1999). The largest we spotted was 2 M_⊙ for the accretor in a low mass X‐ray binary, derived from the assumption that its quasi‐periodic oscillations mark the period of the innermost stable orbit (Schaab & Weigel 1999). The polite way of expressing any reservations you might have about this is "model dependent" (Semerak 1999a, 1999b). If 2 M_⊙ isn't enough, strange quark stars will take you to somewhat higher masses (2.6 M_⊙ according to Zdunik 2000—a truly random sample of a dozen papers mentioning these hypothetical entities). And on beyond, one finds perhaps still more compact stable families of stars with still more generic equations of state (Glendenning & Kettner 2000). Only recently have we suspected that, when Fritz Zwicky called these more compact structures Object Hades, he was really saying, "Oh, H‐ ‐ ‐!".

Pulsation and precession. The former remains unseen after 32 years, and at least the r‐modes are perhaps damped away by magnetic field (Rezzolla et al. 2000). The sparcity of astronomical and planetary objects known to exhibit free precession relative to the number of lectures devoted to the phenomenon in upper division mechanics courses was noted several years ago. Stairs et al. (2000) have added to the inventory the 250 second and 1000 day oscillations in pulse shape and derivative of PSR B1828−11, while Melatos (2000) notes that such things are not free anyhow (pause for suitable feeble joke about the costs of computing and observing time) but are coupled to radiative torques, leading to a time scale of a few years for, e.g., the Crab Nebula pulsar, where precession (free or coupled) is not seen anyhow.

Pulsars. Having reported last year that an observed pulsar, J2144−3933 with P = 8.5 sec, ought to have died when the most productive astronomers were pterodactyles, we are happy to note a reprieve, in the form of a (take a deep breath) space‐charge‐limited‐flow‐polar‐cap‐acceleration model (Zhang, Harding, & Muslimov 2000).

Binaries and multiples. It is possible for a pulsar to be younger than (that is, to have collapsed to a neutron star since the death of) its white dwarf companion say Tauris & Sennels (2000), mentioning two examples by their ZIP+4 codes. Their model began with 7 + 5 M_⊙ main sequence stars and a period of 10 days. Merging pairs of neutron stars can produce all of the heavy (A>130) r‐process abundance peak (Freiberghaus et al. 1999). And the binary millisecond pulsar 1620−26 in the globular cluster M4 has, according to timing measurements reported by Thorsett et al. (1999), a third star of only 0.012 M_⊙ in a 100 yr (25 AU) orbit.

Kick velocities. The speed at which a pulsar begins its life is completely dominated by asymmetry of the supernova explosion for both binary and single stars said Donder & Vanbeveren (1999). With so firm a statement in hand, we almost gave up reading about the subject and so nearly missed the largest asymmetry to date, which imparted a velocity of 740 km/sec (with large error bars) to the X‐ray binary Cir X‐1 (Tauris et al. 1999). No kick at all was needed for the low mass XRBs with black holes (Nelemans et al. 1999), and space velocities are in general a good deal smaller for systems that remain together long enough to recycle their pulsars (Gothoskar & Gupta 2000).

Pulsars. They radiate, a few over a very wide range of wavelengths (Malov 1999). This is not a new discovery. The radiation is polarized, perhaps always 100%, but with gremlins adding up modes to conceal this (McKinnon & Stinebring 2000). Still under discussion are (a) the area from which the photons come (Gwinn et al. 2000), (b) the emission process that sends them on their way (Xu et al. 2000), and (c) how the charged particles that do the radiating are accelerated and transported to get where they need to be to do their job (Beskin & Rafikov 2000), and we are neither repudiating nor endorsing the specific suggestions made in the cited papers. Under the circumstances, the now‐long‐established thought that millisecond pulsars do something different from the others (Kuzmin & Losovski 2000) is not as informative as it might be.

Pulsars also glitch, being glitchiest when they rotate fastest (Lyne et al. 2000b, who specifically vote against a decrease in moment of inertia, and hence for an increase in magnetic moment, as part of the process). Bastrukov et al. (1999), however, ask to have starquakes taken back into consideration after many years of neglect, at least for binary pulsars in eccentric orbits. What Sedrakyan & Shahabasyan (1999) are asking to have taken into consideration is less clear.²

Et Cetera. Here live all sorts of interesting papers concerning X‐ray binaries, quasi‐periodic‐oscillations, and such that are this year's data drop. A few appear in § 5.5.

The spectacular Chandra image of the Crab Nebula just missed the temporal cut for Ap99 and is now to be found many places (Weisskopf & Hester 1999; Weisskopf et al. 2000). It shows jets roughly parallel to the long axis of the nebula and a torus and inner ring that presumably tell the jets which way to go. The data also reveal where energy is being deposited by the pulsar wind, since the brighter bits also have harder spectra. In a bit of tidying up, Wallace et al. (1999) conclude that the nebula resides inside an H i bubble, that is, in a low density region, and does not sport an outer emission line coat. The extended H‐beta emission reported by Murdin (1994) apparently belongs to a background source. Thus about half the mass of the progenitor star remains unaccounted for. The fastest moving gas reaches −2500 km/sec in C iv emission (a STIS result; Sollerman et al. 2000b). The previous record was −2200 km/sec in ground‐based, photographic spectra of Hα and [N ii] exposed long ago by Rudolph Minkowski.

The less‐euphoniously named SNR 0540−69, in the Large Magellanic Cloud, once called more Crab‐like than the Crab (not by us), shows hints of a similar torus‐jet structure in its Chandra image (Gotthelf & Wang 2000). The Tycho remnant, unlike the Crab, apparently does have a faint, outer halo of optical emission, possibly nitrogen‐rich (Ghavamian et al. 2000). This presumably consists of gas shed by the parent star before 1572 (well, long before 1572, allowing for light travel, but you know what we mean).

The Tycho and Kepler remnants (but not the Crab) have in common with Cas A the curious trait of X‐ray nebulae expanding faster than radio nebulae, although they are currently the same size in the two wavelength regimes (Hughes 1999 on Kepler). Could this just be a phase that every (non‐pulsar) SNR goes through at age 400 or thereabouts? In order to get bright X‐rays from the region around a pulsar, its electron‐positron wind must cool directly by synchrotron radiation and not waste its energy making shocks and such (Chevalier 2000). This happens for the Crab and 0540−69, but not for Vela and CTB 80.

Supernova remnants come and go. 3C 397, for instance, is only about 1000 years old (Dyer & Reynolds 1999) and far enough north (δ ≈ + 7°) and close enough that someone might well have seen it from China or Europe. SN 1988Z was undoubtedly seen, and it has just been promoted from elderly SN to young SNR in the last year or two (Aretxaga et al. 1999). The Gum Nebula, on the other hand, should probably be removed from your SNR handbook, not that it has faded away, but because a study of its radio recombination lines suggests that it is a purely thermal source (Woermann et al. 2000a).

Finally, last year we mentioned a bunch of SNRs that are at present colliding with molecular clouds or other bits of densish interstellar matter, so this year we will note only that Pup A is not doing so (Woermann et al. 2000b).

Horizon‐enshrouded objects of planetary and globular cluster mass may belong in the basket of dark matter candidates (§ 12). The sort found in Galactic centers appear as part of the discussion of starburst versus active galaxies (§ 11). And here are a few other topics that have not yet disappeared behind R = 2GM/Rc² (R = 2M in relativists' units), though some may deserve it.

The proceedings of a December, 1997 conference on black holes appeared belatedly as Nos. 3 and 4 of Journal of Astrophysics and Astronomy. X‐ray binaries, cosmic censorship, and other traditional topics are addressed, but the favorite of the more redshifted author is the conclusion of Frolov & Forsaev (1999) that the entropy of a black hole is derivable from work published by Sakharov in 1968. This is not only pre‐Hawking but pre‐Bekenstein. The name Hawking radiation nevertheless stands in accordance with well‐established principles of eponymy (Stigler 1999). The radiation has not yet been seen, but if it were, it should not have a blackbody spectrum, owing to von Zeipel's theorem (de Vries & Schmidt‐Kaler 2000). It should be mostly on branes if there are extra dimensions (meaning more than four, not just more than you feel you need; Emparan et al. 2000). And (take a deep breath) the "coincidence" that we live at a time when a primordial black hole now boiling away (10¹⁵ g) has classical and quantum entropies roughly equal incorporates the same physics as the "coincidence" that we live when the Hubble time is roughly the lifetime of main sequence stars (Barrow 1999).

We think that the intersection of the papers of Fabbri et al. (2000) and Anderson et al. (2000) is that Reissner‐Nordstrom black holes take longer to evaporate than Kerr or Schwarzschild BHs, but not, literally, forever. The Kerr ones, however, require an infinite number of oscillations (which Zeno, at least, thinks means forever) to reach an infinite value of the curvature scalar at their centers (Ori 1999).

A double handful of other papers dealt with spectral and energetic signatures that might allow us to distinguish (nearly) maximally rotating Kerr black holes from Schwarzschild or near‐Schwarzschild ones. A quick and biased overview suggests that the most readily observable should be spectral (Kurpiewski & Jaroszynski 2000; Martocchia et al. 2000) and the least observable the particles that achieve negative energy before falling back in (Li 2000). This, however, allows energy extraction to be as efficient as that from merging BH pairs (Li & Paczynski 2000), thereby matching the fluxes for the brightest X‐ray sources in other spiral galaxies (Makishima et al. 2000).

And the rest is binaries, which undoubtedly make it more difficult to produce a black hole than if you start with a single star of the same mass (Wellstein & Langer 1999). Nevertheless, X‐ray binaries with black holes (meaning things with sizes like Schwarzschild horizons and masses too large for neutron stars, whether or not they turn your watch into a ruler) have become so numerous that we decided unilaterally last year to call them just plain old BHXRBs, and not "black hole candidates." This year we note a suggestion of fractal magnetic structure in their accretion disks (Kawaguchi et al. 2000) and some new orbital data for an old friend, Nova 1938 = Ginga X‐ray nova 1989 (Herczeg & Maloney 1999). Yes, it is still way over the Oppenheimer‐Volkoff limit for any orientation you might choose (M₂sin²i = 6.7 M_⊙). A special adverbial award goes to Uemura et al. (2000) writing on XTE J1118+480, "possibly the first firmly identified black hole candidate X‐ray transient in the halo."

Intermediate mass black holes of 10–100 M_⊙ have decorated several press releases during the year and are just creeping into the archival literature. The "what has changed" here is improved angular resolution of the current crop of X‐ray satellites, which allows resolving what used to be multiple, confused sources in external galaxies into individual sources whose luminosities and spectra can be determined (Ward 2000). Being older than the satellites, we feel free, however, to sound a warning (you can imitate the sound by tearing this page out of PASP, but only if it is not a library copy) derived from the history of "the most massive single star," whether R136 in the LMC or some other. The point is that successive generations of improved angular resolution have persisted in separating off more and more components, until the brightest was probably not much more massive than 100 M_⊙. One can imagine the same thing happening for the X‐ray sources, until individual masses are about what you would expect from the core of a 100 M_⊙ star, the canonical 6–10 M_⊙ of the known XRBs.

The Higgs particle has (probably) been seen (Janot 2000). This was the last missing particle in the standard model of particle physics (following the production and recognition of the tau neutrino earlier in the year) and is the one that allows all the others to have non‐zero masses. Without it, there would be no demand for Weightwatchers, Metracal, or diagonally seamed garments. The discovery follows a traditional pattern whereby bosons (the Higgs) are found at CERN and fermions (the tau neutrino) at Fermilab. Since the Higgs completes the standard model, it is appropriate that its rumor should have been followed soon (though out of period) by one for a measurement of the dipole moment of the muon which disagrees with the standard model prediction and perhaps foreshadows the coming of supersymmetry, technicolor, or other new physics (as opposed to systematic errors). See Wilczek (1999, on CP violation) for what it means for something to be evidence of "Beyond the standard model."

A quark‐gluon plasma has, perhaps, been made at CERN (Antinori et al. 2000). Whether this is the first since the big bang depends on the level of technology achieved on other planets.

Holmlid (2000) says that he has seen the de‐excitation of Rydberg matter in unidentified infrared bands from the interstellar medium. Notice that the "IR" part of this rules out one's first guess that Rydberg matter does something at 13.6 eV.

Somebody (besides us) once said that the only absolutely conserved quantities in the universe are those (charge, angular momentum, and mass‐energy) for which the cosmic total is zero. Lepton number is not conserved "beyond the standard model" and McDonald (2000) has revived the idea that there might be an excess of neutrinos over anti‐neutrinos that is very large (10⁸–10⁹) compared to the baryon excess over anti‐baryons, though still small (1% or thereabouts) compared to the number density of photons. This sort of imbalance is too small to affect the products of nuclear reactions in the early universe by more than the error bars due to other factors.

Antimatter that persists as the universe cools down below a few MeV (z = 10¹⁰, T ≈ 3 × 10¹⁰ K) is another idea that has been in and out of fashion with the regularity of wide lapels. Kurki‐Suonio & Sihvola (2000) have again explored the use of persistent antimatter to increase the amount of deuterium and decrease the amount of helium coming out of the big bang, thus giving better agreement with the observations, even if the baryon content of the universe is larger than last year's best‐buy value.

The chirality (handedness) of life could be a reflection of fundamental physics (§ 3.6.1), a result of ultraviolet irradiation of interstellar molecules (Chrysostomou et al. 2000, showing that the dust in OMC‐1 does indeed see the ultraviolet from the central source), or merely a small error in coordinate systems (Mignard & Froeschle 2000). Actually, this last merely points out that the fundamental coordinate system FK5 has global rotation, which is unlikely to be responsible for either cosmic or local handedness. But if some bit of fundamental physics begins to sound a little too bottom heavy, you can always close the discussion by saying, 'Well, if it weren't like that, we wouldn't be here to talk about it.' This may well be true for the ratio of carbon to oxygen made in stars and its dependence on the balance between coulomb (electromagnetic) and strong (nuclear) forces (Oberhummer et al. 2000). If you take this sort of argument seriously, you generally call it the anthropic principle.

Newton's constant of gravity, G, has been the least precise of the important constants for decades. It may still be, but G = 6.674215 ± 0.00092 × 10^-x (x = 8 or 11 in some popular unit systems) is a very considerable improvement, and probably a change in the 4th digit from what you learned in school. The new value comes from a new sort of torsional balance, owned and operated by Gundlach & Merkowicz (2000). Our knowledge of the masses of the Sun and Earth were thereby also much improved to 1.988425 and 5.972245 (± errors of similar small fractions and in suitable units). The expected lifetime of the Sun is thereby decreased by about 0.07% or 6 million years. Oh dear. We were counting on that time to finish this review.³

Notoriously, if you build a better mousetrap, then (a) nature will build a better mouse and (b) you will lose your shirt trying to market it, but some venture capitalist will make a fortune selling mousetrap pads. We hope that neither of these will be the fate of the imaginative new devices and observatories pandected here.

Images and spectra from the X‐ray satellites Chandra and XMM‐Newton appeared in many colleagues' lists of highlights for the year. Chandra images are mentioned in § 4.4 on supernova remnants and elsewhere. In addition, there are some very interesting early results on clusters of galaxies, for instance the X‐ray hole coinciding with the location of the radio lobes in Hydra A, and a point source at the radio core (McNamara et al. 2000). Most of the three dozen or so X‐ray cluster papers we caught during the year, however, still were reporting data from ASCA, BeppoSAX, or even older missions. And XMM‐Newton made the archival literature only through its successful launch (§ 1.2).

Optical interferometry and adaptive optics were the second‐most‐mentioned technology during the year. Saha (1999) provided a very comprehensive review of interferometric techniques and their application. The nulling interferometer (Wallace et al. 2000a) can suppress a central point source by a factor up to 10⁴ to facilitate examination of the surrounding sky. Planet searches are, of course, the driver for this. Keck I has been used as a Michelson interferometer by masking part of the aperture to yield a dynamic range of 200 (Tuthill et al. 2000b). The range sounds larger when you say it in magnitudes (5.75), and the ESO New‐Technology Telescope has been pushed to a dynamic range of 9 mag in the K band (Neuhauser et al. 2000).

Adaptive optics is for when you have only one telescope, though if it happens to be Keck II, images of 002 are achievable first time out (Wizinowich et al. 2000). Even the comparatively petite SOR reaches 005 resolution, though its users have had lots of practice (Angel & Fugate 2000, a minireview). Given how hard it is to find even a single guide star for A0, to the point where people make them with lasers, you might suppose that asking for multiple guide stars would render the whole thing impossible. In fact, multiple guide star adaptive optics can apparently be made to work over the entire sky (Ragazzoni et al. 2000, who note that the scheme was first proposed by Beckers 1989). Several other papers in this territory noted that some of their intellectual infrastructure had come from other papers by the same author. Under the circumstances, we think it ominous that he has moved to Chicago, where there are hardly enough guide stars to find your way around the Loop. Some additional papers reporting results achieved with adaptive optics and interferometry appear with their astronomical subjects, as someday clearly all such papers will.

More than 40 other named (or more often acronymed) devices, algorithms, and sites appeared in the year 2000 literature. It is not necessarily undesirable to have been omitted from this list of favorites.

• •  
  The Very Small Array is being built to look for Galactic foreground noise that impedes our view of the cosmic microwave background (Giardini et al. 2000). The name and acronym were, however, already used some years ago for a pair of small radio telescopes operated at Sonoma State College by Lynn Cominski and her students, and we have the T‐shirt to prove it (but are not quite sure how to cite a T‐shirt in PASP house style—the style of the T‐shirt is large, sloppy, and magenta).
• •  
  GRAPE4, which was used to show that warps in disks of spiral galaxies are more stable if the halo is prolate instead of oblate (Ideta et al. 2000) has been succeeded by GRAPE5 (Kawai et al. 2000), whose purpose in life is to calculate gravitational forces in N‐body problems, while a regular computer takes care of the orbits of the particles.
• •  
  NBODY1 to NBODY6 was a series of (strangely enough 6) algorithms used to study multi‐object systems in which gravity is the dominant force, from planetesimals to cosmology. Aarseth (1999) is the text of the Brouwer Prize Lecture by the developer, who once ran out of his office on a Saturday morning when very few people were around, asking, "Virginia, how many planets are there?" "Nine," said she (this being in the days before nasturtiums had been cast upon Pluto). "I've made eight," said Sverre, about 10 years into this 40 year effort.
• •  
  Last year we asked what immersed gratings were immersed in, doubting that it was either water or wine. That of Lee & Allington‐Smith (2000) swims in glycerol, of which there seem to be several dozens kinds, with indices of refraction between 1.403 and 1.478. It's amazing what you can find in the Handbook of Chemistry and Physics (though a speaker at a recent AAS meeting, unable to locate the coefficient of thermal expansion of an AAS president, was forced to use that of rock salt).
• •  
  AGAPEROS (Melchior et al. 2000b) is a new automated search for gravitational lensing by stars (etc.) whose first yield is 632 variable stars in the Large Magellanic Cloud, noise to some observers, signal to others.
• •  
  Diabolo is a high throughput, dual‐channel, millimeter photometer (Benoit et al. 2000). If the first is an acroynm for the second, it is not in any of the languages we know.
• •  
  PHEBUS was a gamma‐ray detector (Barat et al. 2000), while Phoebus is a collaboration to study solar oscillations (Appourchaux et al. 2000).
• •  
  Mt. Graham is the home of (at least) three telescopes and a much larger number of Mt. Graham red squirrels. The latter apparently don't mind the company of the former as much as had been feared (Young & Smith 2000). They eat very different sorts of nuts.
• •  
  Biosphere 2 wasn't really home to anybody, but some of its temporary residents managed to collect observations of V803 Cen, a dwarf nova with a helium star donor (Patterson et al. 2000).
• •  
  Milagrito is a prototype of the Milagro array to detect very high energy gamma rays by counting muons and such, so that you don't have to wait for dark clear nights to see flashes of Cerenkov radiation. Atkins et al. (1999) reported its first scientific results in a paper whose 44 authors may have outnumbered the photons they counted. The flux found for Mrk 501 was consistent with earlier Cerenkov measurements.
• •  
  Queue observing is the idea that projects accepted for a particular (ground‐based) telescope are carried out with the proposer not present, in an order intended to make most efficient use of changing sky and weather conditions. Tried at WIYN for several years, it was less productive of "scientific papers per observing night" than standard PI observing by a factor about 1.4 (Massey et al. 2000). The sample of papers is small, and no citation data were compiled so that the 9 "queue" papers could conceivably be much more valuable than the 19 "non‐queue" papers (though we doubt it). Space‐based observatories nearly all operate in a queue‐like mode, owing to the difficulty of getting observers there and back in a timely fashion.
• •  
  X‐ray focusing optics push ever onward to harder photons. Pieces of an LP record work at 23 keV (Cederstroem et al. 2000), providing a focal length of 22 cm when the pieces are 180 μm apart. This innovative use for what is now nearly a waste product gives renewed hope to those of us who have saved our hula hoops, frisbees, nerf balls, and pet rocks.
• •  
  X‐ray grazing incidence interferometry appears to require something more than four LP records, but Cash et al. (2000) are thinking about it as a way to resolve the microarcsecond structure of shadows cast by black holes, death inspirals, and other cheerful topics.
• •  
  Fast Intermediate‐band Lightgathering Medium is still being used, along with what seems to be the world's largest interference filter (35.6 × 35.6 cm, centered at Hα) to look for Herbig‐Haro objects (Mader et al. 1999) and to survey the Milky Way (Georgelin et al. 2000). Though it has been only a little more than a century since there was real disagreement among variable star astronomers about whether visual or photographic methods were best (Saladyga 1999), the acronym film has been in use for so long that the full name has nearly been forgotten. This topic was originally indexed under "making a virtue out of necessity," along with the use of a transit circle in Washington, DC (Rafferty & Holdenried 2000), the measurement of radio pulsar polarization using a single feed (Ramkumar & Deshpande 1999), and the biomechanical servo of Kraan‐Korteweg (2000, a catalogue of 3279 galaxies in a small portion of the zone of avoidance covered by the ESO/SRC Schmidt survey plates, found by eye examination).
• •  
  WISP has recorded the first wide‐field UV polarimetric image of anything. Cole et al. (1999b) chose the LMC and concluded that many of the UV photons that eventually make their way out have been scattered.
• •  
  CDS and ADS are the two giant astronomical data bases located (in so far as anything largely electronic can be said to be located) at Strasbourg and NASA. Genova et al. (2000) and the following three papers and Kurtz et al. (2000) and the three papers after that explain what is in them.

Laboratory measurements continue to be crucial for many parts of astrophysics. Familiar cases include nuclear reaction rates (Rayet & Hashimoto 2000 on what is needed for better calculation of yields from the s‐process), properties of dust (Wurm & Blum 2000 on grain shapes), atomic data (Nahar et al. 2000, paper XLIII from the Iron Project), and molecular properties (Arnoult et al. 2000 on PAHs, which, we have just learned, are to be regarded as dangerous pollutants on Earth, although a good thing to have in interstellar space). Other, less usual examples of laboratory astrophysics include:

• •  
  A dynamo made of rotating liquid sodium (Gailitis et al. 2000), which is another thing you might not want in your backyard, though very useful at the center of the Earth (Jackson 2000).
• •  
  A radiative shock (Shigemori et al. 2000) with the velocity of the blast wave less than the Sedov‐Taylor value (but in agreement with better calculations).
• •  
  Masing HCN, the only molecule known to be excited by the same mechanisms in laboratory and ISM (chemical pumping; Schilke et al. 2000).
• •  
  A laser‐driven blast wave of eventual relevance to supernovae and gamma‐ray bursts (Keilty et al. 2000).
• •  
  The proceedings of an entire (Spring 1999) conference devoted to laboratory astrophysics with intense lasers (Remington et al. 2000).

There are, of course, many other astrophysical processes for which observations and the best available theory disagree, and one only wishes one could make laboratory measurements, for instance the broadening of hydrogen lines arising from levels with n = 200 and higher in H ii regions (Bell et al. 2000).

Light echoes began with Nova Persei 1901 (Kapteyn 1907) and continued with Nova Sgr 1936 (Swope 1940). The case of SN 1987A has been discussed many places, but SN 1991T (Sparks et al. 1999) is a recent addition to the club. Light echoes from dust could contribute to the late optical glows from gamma‐ray bursters (Esin & Blandford 2000). And it is just barely possible that we might still see that of Tycho's 1572 supernova (Maslov 2000). Highly polarized, near infrared in a ring about 5^'' across is the best bet. This comes so close to time travel ("see what Tycho saw") that it feels a bit spooky. Except, of course, that Tycho did very little near infrared observing.

Detonation (supersonic) versus deflagration (subsonic) has been an issue in models of Type Ia supernovae for years. Half a dozen or more papers discussed it, making a further distinction between delayed and convective deflagration (Dupke & White 2000) and noting a pathological case (Sharpe 1999), chosen from among self‐sustaining, supported, cellular, and double detonation (this last means helium ignites first, then carbon and oxygen). Timmes & Niemeyer (2000) discuss the somewhat similar dichotomy for helium explosions in X‐ray bursters, invoking a process triply eponymed to Zeldovich, von Neumann, and Döring. Richard Armour described a triumvirate as an arrangement in which, when one of them left the room, the other two made uncomplimentary remarks about him. One suspects it would have been Döring in this case.

What pumps OH masers? Different processes at high density (star formation regions) and low (supernova remnants) says Caswell (1999).

You have to add energy to ionize or excite atoms and to dissociate molecules (indeed sometimes to make them if there is a barrier to get over). Is this done by photon absorption or particle impacts (shocks)? Yet another example of "both please" as you might expect, illustrated by Quillen et al. (1999 on excitation of H₂ molecules in Seyfert galaxies), Sugai & Malkan (2000 on ionization in LINERs), Muthu et al. (2000 on shock ionization in some planetary nebulae, where we have known that there is photoionization since the time of Zanstra or thereabouts), and, among a number of others, Palumbo et al. (2000 on the chemistry to make the carriers of the 4.62 μm band, which can be promoted by ion bombardment as an alternative to ultraviolet photolysis).

Inverse Compton scattering was, in its day, a mechanism looking for a source to operate in. This year there were at least 15 places, or anyhow papers, where it happened (from Giommi et al. 1999 on the blazar SJ 0716+714 to Sazonov & Sunyaev 2000 on the preferential backscattering of soft photons) and at least six where it did not (from Blasi 2000 on making the hard X‐ray tail in clusters of galaxies by distorting the Maxwell‐Boltzmann distribution to Berghoefer et al. 2000a on the soft X‐ray and extreme UV excesses of M87 and the Virgo cluster as synchrotron), where we mention only the first and last papers from each point of view that appeared during the year. Notice that "both please" is not a possible answer for the C^-1 or not C^-1 case, unless you are willing to accept synchrotron‐self‐Compton as an example (Tagliaferri et al. 2000).

Where is the return current? In the outer collar of the jets from young stellar objects, where the current goes out in the core, say Lery & Frank (2000).

Is the Jeans mass good for anything? Certainly not for earning citations for Jeans (not even from us). But does it have anything to do with star formation? More than you might think in recent, magnetically dominated years (Indebetouw & Zweibel 2000) according to Coppin et al. (2000 reporting SCUBA data on OMC‐1 and emphasizing fragmentation processes) and Curry & McKee (2000 on models of Jeans mass cores inside much larger, stable clouds).

Turbulence and star formation are two things we have been describing for years as "not very well understood" (compared, say, to hydrostatic equilibrium and the structure of main sequence stars). It is, therefore, altogether fitting and proper that a connection between the two should be coming into play. First, it seems that turbulent flow in H i speeds up the formation of molecular clouds and their first stars (Ballesteros‐Paredos et al. 1999). The associated magnetic field should also be turbulent or stochastic (Parker & Jokipii 2000). Second, the turbulent structures within the resulting molecular clouds may translate directly into the initial mass function, N(M), of the stars formed (Myers 2000; Muench et al. 2000). Simplifying nearly beyond recognition, stars form in the small, still pools, close to the Jeans mass, in between turbulent vorticies (Elmegreen 1999; Kornreich & Scalo 2000). Such turbulence in molecular clouds dissipates in much less than the free fall time and must be continuously driven (Mac Low 1999). Shocks from many sources (Kornreich & Scalo 2000) and thermal instabilities (Burkert & Lin 2000) are among the drivers.

Not quite everyone regards turbulence as a good description of star formation regions. Gosachinskii & Morozova (1999) find that the internal velocity dispersions of H i clouds as a function of their size are not described by a turbulent spectrum, while Lazarian & Pogosyan (2000) say that the description is OK, but that a Kolmogorov turbulent cascade is not the cause.

Convection and turbulence remain important after stars form, and the year witnessed many papers concerning convective overshoot, semi‐convection, and their observable consequences for stellar evolution and surface abundances (e.g., Herwig et al. 1999; Bruntt et al. 1999). There also appeared paper M in a series of N (N≥M) by Canuto (1999) in which the author, sometimes with colleagues, struggles onward toward a unified theory of stellar turbulence, including rotation and so forth. This one introduces a post‐Schwarzschild, post‐Ledoux criterion for convective instability, turbulent kinetic energy >0.

Stars and the ISM are not the only places where convection happens. On Earth we call it weather. They also increasingly do so on Jupiter, though "they" is, unfortunately, not the Jovians, but only terrestrials interested in Jovian surface phenomena (Gierasch et al. 2000; Ingersoll et al. 2000; Seiff 2000). The correlations of storms and lightning imply moist convection (like on Earth), but driven by heat from inside rather than from the Sun. Gu et al. (2000) applied mixing length theory to the convective transport of energy in accretion disks (for cataclysmic variables, active galactic nuclei, and X‐ray binaries), but did not mention weather, let alone trying to do anything about it.

What drives stellar winds? All the things you've ever heard of (Ap93, § 4) from MHD waves (Airapetian et al. 2000) to radiation pressure in the absorption lines. And if the momentum in the outgoing radiation isn't as large as the momentum in the wind, then all the photons just have to work twice as hard (Herald et al. 2000) or even six times as hard (Nugis & Lamers 2000). The proper name for this is "multiple scattering." Much the same can be said for winds from disks in various contexts (Pereyra et al. 2000 on cataclysmic variables).

Missing opacity, meaning that to make your model stellar atmospheres match the colors (etc.) of real ones, you must use optical (etc.) opacities larger than the sum of all the causes you know about, continues to be missed (Cassisi et al. 1999 on globular cluster stars).

High latitude (halo) B stars are puzzling if they are massive young objects rather than evolved (post‐AGB) low mass ones. Lehner et al. (2000) draw attention to a particularly anomalous one, with Population I composition and surface gravity, but very small rotation velocity (vsin i<5 km/sec) and microturbulence less than 2 km/sec, compared to the OB norm of 15 km/sec (Villamariz & Herrero 2000). Well, there are also some people who seem to have been born middle‐aged.

Dwarf carbon stars supposedly got that way by being dumped on by an asymptotic giant branch companion (now a white dwarf), and they ought to exist if you want to blame binary mass transfer for composition anomalies in more evolved stars. They do (Ap93, § 5.3), but one wonders whether 0 out of 48 in a high latitute sample of carbon stars with upper limits to their proper motions is really enough (Totten et al. 2000). It is easier to make a carbon star if metallicity is small to begin with, but Zacs et al. (2000) say that BD +75°348 is the old disk analog of the halo CH (giant) stars.

Supermetalrich stars continue to exist, an analysis of Mu Leo (K2 iii) yielding [Fe/H], [Ca/H], and [Mg/H] = 0.3 ± 0.04 according to Smith & Ruck (2000), who find the expected, lower metallicities for other K giants in the solar neighborhood and and in the Hyades using the same techniques.

Dynamical screening is, um, er ... . Well, ordinary screening is that two protons (etc.) attempting to get over their mutual coulomb barrier so as to embrace each other's nuclear potentials will, on average, find some portion of an electron between them, which lowers the barrier a bit (remember electrons are bigger than protons in a quantum mechanical sense). Dynamical screening is a further lowering of the barrier when you remember that the particles are all running around madly. Round one in whether this is an important effect was fought several years ago. We missed citing the TKO arguments of Gruzinov (1998), but Opher & Opher (2000) now say that those arguments are invalid (without voting on whether or not dynamical screening is important). Shaviv & Shaviv (2000) say again that it is important. (The family aspect of the discussion reminds one a bit of the Hatfields and the McCoys.) And, just in case you think that, at least, the sign of the effect was agreed upon, if not the amplitude, Tsytovich (2000) has weighed in with a calculation of screening that has static effects cancelling out, so that only dynamical screening remains and, he says, raises rather than lowering the coulomb barrier. This would reduce the rate of proton captures by Be⁷ in the Sun by about a factor of two, reducing the expected flux of the highest energy neutrinos by a comparable amount.

The first ionization potential (FIP)? Well, every element has one. The question is whether stellar coronae and winds are enriched in those that are relatively easy to ionize. Yes for the Sun; apparently not for some other stars (Ap98, § 7.4); and maybe for Capella (Brickhouse et al. 2000), with deficient coronal neon the only representative of high FIP elements.

Are there other stars like the Sun? Well, sort of. But Glushneva et al. (2000) are not the first to note that stars with the same spectral type generally have different colors, and conversely. Sekiguchi & Fukugita (2000) conclude that the standard solar B−V = 0.62 is either wrong or anomalous or both.

OB runaway stars, liberated by supernova explosions of their binary companions, were firmly expected before the concept of mass transfer early in the evolution of close binaries was established. Now we prefer to expect runaway binaries, with a neutron star left from the first explosions still bound to the OB secondary. Examples remain rare, and just as Israelian et al. (2000) were getting our hopes up for HD 188209, with what seemed to be a 6.4 day period, they let us down with the phrase quasi‐period and the conclusion that it is not a binary.

Models of stellar atmospheres have been around long enough that even we are not quite old enough to remember when the reversing layer approximation (Schuster 1905; Schwarzschild 1914) was the best availalble. These days they generally have to be non‐LTE (Przybilla 2000, and no way of writing this out with any number of hyphens will convey the idea that it is the equilibrium that is being negated not the local), non‐spherical (Allende Prieto et al. 2000), stratified in velocity (Luttermoser 2000), extended (Hauschildt et al. 1999), expanding (Fitzpatrick & Massa 1999), and probably other things in the dozen papers we indexed but haven't cited. And even then, you cannot always get so simple a thing as the temperature of a star consistently from the whole spectrum (Kotnik‐Karuza & Jurdana‐Sepic 2000). By the way, Karl Schwarzschild also invented LTE.

L and T dwarfs, unknown, or at least unnamed, a couple of years ago, are now so numerous that even a centipede has to take off his shoes to count them (Fan et al. 2000b; Kirkpatrick et al. 2000; Burgasser et al. 2000a), the largest yield currently coming from the 2MASS and SDSS surveys. Increasingly, it is possible to make both statistical and extremal statements. Lithium line strength peaks at L6.5. The temperature gap between type L8 and the hottest T V is less than 350 K, which is why we don't catch many transition objects with methane just phasing in (Leggett et al. 2000). Activity indicators are slightly discordant, but the latest X‐ray source is vB10, and it may appear only in flares with no quiescent material hotter than 10⁶ K (Fleming et al. 2000) No stars later than L4.5 show Hα emission; and Beuermann et al. (2000) suggest that the brown dwarf secondary of the polar EF Eri has no magnetic field, which may be characteristic of the class.

A real and significant gap is that in the numbers of companions with masses between big planets (10 M_J = 0.01 M_⊙) and small stars (0.1 M_⊙; Halbwachs et al. 2000), though there are now about 10 companions with measured masses ≲0.1 M_⊙ (Delfosse et al. 1999).

Among the L and T extrema we find the first double brown dwarf with a spectroscopic orbit, PP 15 (Basri & Martin 1999) and three probable visual binary BDs (Koerner et al. 1999; Reid & Mahoney 2000), and the coolest and faintest, Gl 570D, at T = 750 ± 50 K and L = 2.8 ± 0.3 × 10^-6 L_⊙ (Burgasser et al. 2000b).

A dozen or more papers address L and T spectra. Some of the interesting conclusions are (a) you need dust opacity to fit the observations (Pavlenko et al. 2000), (b) BDs have their very own unidentified spectra features (McLean et al. 2000), (c) there are modest differencies in the spectral subtypes assigned by different groups (Martin et al. 1999), and (d) the absorption in the Na i D doublet takes out so much yellow light that the net visible color might well be purple (Liebert et al. 2000; Burrows et al. 2000a). Draft poems beginning "I never saw a purple star" should be submitted to the more purple author in plain brown envelopes.

Lithium, despite its fragility at high temperatures, is to be found in at least a few stars at every major evolutionary phase (pre‐white‐dwarf), not to mention in 20‐some papers. Some stars still have it (Stout‐Batalha et al. 2000) and some have it again, either because they have produced it themselves (de la Reza et al. 2000) or because they have not yet fully digested a dinner of lithium‐rich brown dwarf or planet (Pilachowski et al. 2000) or both (Denissenkov & Weiss 2000).

Other surface anomalies can be fairly firmly assigned to upward mixing of nuclear reaction products. Technetium remains the classic example, making it to the top as semi‐regular variables evolve into Miras (Lebzelter & Hron 1999). Other anomalies are surely due to radiative levitation and other surface processes. The spotty surfaces of the Ap and Bp stars are classic examples (Leone et al. 2000; Khokhlova et al. 2000). Still other anomalies are still being debated between the nuclear and surface camps, although Roby et al. (1999) had to wait almost 3.5 years between submission and publication to be allowed to do so.

Many authors asked for more or earlier mixing than you naively expect for both main sequence (Gonzalez & Wallerstein 2000) and red giant (Boyarchuk et al. 2000) stars, but Takeda & Takada‐Hidai (2000) were unusual in asking for anti‐mixing so that the surface ratio of H to He goes down in their stars as N/C goes up.

And, though we have collected several dozen other papers addressing a comparable number of other issues in stellar structure and evolution, it is time to say goodnight, Dick (Good night, Dick) and admit that, not only do evolutionary tracks calculated by different groups using equally plausible (though different) physics and methods disagree by up to 20% on important issues like main sequence lifetime (Dominguez et al. 1999), but sometimes no one set of tracks will fit both members of a close binary system (Lastennet et al. 2000), which brings us naturally to binary stars.

Cowling (1941) thought that having a close companion might drive stellar pulsation (next section), though he didn't consider damping. In fact, the two characteristics seem for the most part to be randomly assorted. There are a few anti‐correlations. None of the rapidly‐oscillating Ap stars is a spectroscopic binary (Hubrig et al. 2000), and only one of the pulsating white dwarfs, GW Lib, is known to have a companion (D. Koester, private communication to Hubrig et al.). Thus one can arrange the traditional questions about binary stars into the same sort of evolutionary sequence found for single stars, multiplied by itself. In a few cases, we begin with the answers.

There are no Population III binaries (and no Population III single stars). The incidence in Population II is much more like that in Population I than was thought a generation ago, when there were supposed to be almost no Pop II binaries. There may still be a modest relative deficit at the separations characteristic of visual binaries and common proper motion pairs (Allen et al. 2000a). Occasional disruptions of wide pairs over the past 15 Gyr are at least as likely as some initial difference.

When do binaries form? Early, in the sense that the incidence among T Tauri stars is, if anything, rather larger than for main sequence pairs at the separation probed by speckle interferometry (Koehler et al. 2000, who also note that the fraction of speckle binaries varies from one star formation region to another). Apparently either some pre‐MS binaries fall apart or their orbits shrink. A pair with separation of 8900 AU contains the youngest stars in Orion (Chini et al. 2000).

How and where do binary stars form? Boss (2000a), in a report from an April 2000 meeting in Potsdam says, "fragmentation is the only game in town" (because separate formation and capture cannot account for the shared circumstellar disks often seen). Tsuribe & Inutsuka (1999) partially demure, suggesting that only the cooler clouds can fragment to binaries. Boss et al. (2000) favor the non‐isothermal clouds, which is perhaps not quite the same set. Bate (2000) reaffirms that fragmentation plus accretion favors the production of pairs with mass ratios near one among short period system, and that mass ratios less than 0.1 are quite difficult to produce. Soederhjelm (2000) provides data from the Hipparcos catalog supporting the association between large mass ratios and short periods.

When and how does companionship affect evolution of stars? From the very beginning imply Looney et al. (2000) on shared versus separate disks and envelopes in pre‐MS binaries of various separations. The intermediate case of a shared envelop happens at a = 150–3000 AU. Early interaction is sufficient to synchronize what otherwise might be separate events, since visual double T Tauris are, as a rule, either both classic type or both weak lined, meaning that the partners lose their disks together (Duchene et al. 1999)

Some classes of stars are always binaries, whether by definition or hard‐won observations, including (a) W UMa's by definition, and more work is still needed to understand how the components maintain different effective temperatures despite sharing an envelope (Wang 1999), (b) RS CVn stars, whose spots hover around preferred longitudes; Holzwarth & Schuessler (2000) blame symmetry breaking due to the companion, but we think single active stars can also have preferred spot longitudes, (c) Ba ii giants, a product of hard work, continuing with the deduction that some must have been polluted by mass transfer from their companions while they were still on the main sequence, because the white dwarfs have cooling times longer than the giant lifetimes (Böhm‐Vitense et al. 2000), (d) Beta Lyrae, with (for lack of other explanations) a circumbinary ring which must have formed out of material lost during the evolution of the system (Umana et al. 2000), and (e) the cleanest set of binaries we know, those with orbit parameters determined by Daniel M. Popper, of which HS Aqr is the last, at least in this world (Popper 2000).

All the cataclysmic variables are, of course, also binaries by definition, and frequently also by observation. New subtypes are declared from time to time, including smooth versus featured light curves (Rosenbush 1999b), TOADs that also have normal dwarf nova outbursts (Wheatley et al. 2000), a continuum between SU UMa and WZ Sge types (Baba et al. 2000, the latter are, roughly, the TOADs), and (d) a class of dwarf novae with periods smaller than the underinhabited P = 2–3 hour gap and with magnetic fields between those of the AM Her stars and those of the intermediate polars (King & Wynn 1999). "Supermediate" polars is the name that comes to mind. The prototype is EX Hya. Schmid et al. (2000) have gone the other direction, proposing a unified model for some previously separate classes, based on our orientation relative to the orbit plane, rather like "unified" models of active galaxies.

But the burning front before us is, "Do novae hibernate?" That is, after an outburst, do they eventually fade (reduce their mass transfer rates) far below the levels seen a few years to decades pre‐ and post‐explosion. A canonical nova outburst brightens the system by at least 10 magnitudes. Do we have evidence for CVs much fainter than +5? Yes, undoubtedly. Gaensicke et al. (2000a) and Reimers & Hagen (2000) provide specific examples, where mass transfer is 10^-12 M_⊙/yr or less and doesn't even keep the white dwarf warm. But we have no guarantee that these systems ever have had or will have classical nova explosions. The alternative approach is, therefore, to try to recover known old novae, the strategy that originally led to the hibernation scenario. Most recently, Robertson et al. (2000) have probably found Nova 1678 (V529 Ori) and a number of 20th century events, but not Nova 1612 and several others. A look at the numbers suggests the increasingly familiar answer, some of them do and some of them don't.

The largest class of binary papers during the year dealt with X‐ray binaries (a dozen on quasi‐periodic oscillations alone). The middle two papers (in temporal sequence, that is, all of them are, of course, of above average quality) reported (a) the third case of ≈1 Hz QPOs in a low mass XRB which persist during the low intensity state (Jonker et al. 2000 and Miller 2000), and, conversely (b) QPOs that disappear at high mass transfer rates (the high state) (Cui 2000; Campana 2000).

From among the other dozens of XRB papers, we pluck at random another probable example of mass transfer that is sustained by X‐ray heating of the donor star (Greiner et al. 2000b). Some pulsar binaries like 1957+20 have been advertised as doing this for years, however much it may sound like pulling yourself up by your own bootstraps (no, we don't remember bootstraps, having always used button hooks).

And, in case you are interested in black holes and their formation, Israelian et al. (1999, with a commentary from Cowan 1999) have provided evidence that the donor star in BHXRB GRO 1655–40 (one of the galactic superluminals or microquasars) has on its surface 6–10 times the solar abundance of C, Mg, Si, and S, but normal iron, identifiable with the ejecta of a typical Type II supernova. This rated a News and Views commentary because it implies that supernovae can leave black holes as their direct core product, there not having been enough time in the life of the system for mass transfer to drive a neutron star above the Oppenheimer‐Volkoff limit (the NS equivalent of the Chandrasekhar limit for white dwarfs) and feed it up to the present mass. Having an F–G III donor, this system also adds one to the very small group of XRBs (prototype Her X‐1) with intermediate mass donors, although the data are out of period (Shahbaz et al. 1999).

The final large class of binary star papers dealt with assorted examples of changing orbital periods and reasons these might happen. An exception is 64 Peg, whose period is probably not changing after all (Boden et al. 1999, reporting resolution of the pair with the Palomar Testbed Interferometer). Such resolution can only become more common and increase our supply of accurately measured stellar masses, e.g., Pourbaix (2000) reporting on 40 systems and Griffin & Griffin (2000) on one that would surely benefit from that sort of attention.

Of the papers reporting period changes, the middle one (again temporally) was Liu & Yang (2000), who found ΔP/P = -5 × 10^-11 for W UMa between 1949 and 1998. This may seem like an awfully small ΔP to be able to detect, but 40 years is an awful lot of periods of W UMa. The amplitude is very much what you would expect from the thermal oscillation model of how W UMa stars earn their livings (Lucy 1968).

Orbital inclinations can also change if a third star exerts suitable torques. We thought briefly that there had been two new examples during the year, but in fact Milone et al. (2000) and Torres & Stefanik (2000) were both talking about SS Lac, which started eclipsing in about 1885, had central eclipses around 1912, and hasn't blinked since 1937. It should start eclipsing again in about 1150 or 2300 years. The two previously known quitters are AY Mus and V107 Sco.

At some level, all stars vary (probably on every time scale from that of atmospheric waves to the main sequence lifetime). Out of the range of possibilities, about 30 types, dozens of individual stars, and many theoretical considerations appeared in something like 100 indexed papers. Unfairly short shrift to all except the few to which we responded, "I didn't know you did that" (as the bishop said to the actress).

• •  
  A few dozen M giants with periods less than 10 days (Koen & Laney 2000). Spots and ellipsoidal shapes can be ruled out, and very high overtone pulsation (often more than one mode per star) seems to be all that is left.
• •  
  Contrarily, there are B stars with unexpectedly long periods of 1–5 days (De Cat et al. 2000a). The authors suggest radial modes of high order. About half the stars are in binaries, which may or may not have anything to do with the price of modes (half of all stars are in binaries, after all).
• •  
  In the spectrum of α² CVn, lines coming from ions with different distributions on the surface show different radial velocity curves (Gerth et al. 1999). The oddness of this star was known to Struve & Swings (1943) and to Belopolsky (1914) before them.
• •  
  Epsilon Per is a multi‐mode, pulsating close binary system (De Cat et al. 2000b). Despite the 13 pages of the paper, the authors do not seem to find space to mention the spectral types, but since the analysis uses Si iii lines, we conclude that at least one of the stars is hot. Gies et al. (1999) kindly clarify that the primary is a B III star, and the largest‐amplitude example of giant variability. De Cat et al. (2000b) note that there is a probable third, visual, component.
• •  
  Procyon A is bright enough to be monitored by WIRE, and comparison of data with models should eventually yield very precise values of its mass, radius, and luminosity (Chaboyer et al. 1999; Barban et al. 1999).
• •  
  Some Herbig Ae stars are also Delta Scuti non‐radial pulsators (Marconi et al. 2000b). It is not quite clear why this needed the urgent publication of a Letter, since the first example was reported in 1927. But then it has also been known for many years that some snarks are boojums.
• •  
  In the O9 III+B1 III system ι Ori, the variability is actually due to modest tidal distortion, reminding us, that as Cowling (1941) said at the beginning of the previous section, one can imagine cases where, given some sort of resonance between an orbit period and a natural frequency, binary companions could drive pulsation.

Lowell (Percival) would surely be pleased that the telescopes at his observatory continue to be upgraded (Wilkinson 1999) and even more pleased at the ever‐increasing abundance of water on Mars, or at least of papers discussing it (Malin & Edgett 2000).

See Naples and then die? Well, not immediately, but radar interferometry, using the ERS 1 and 2 satellites shows that there has been subsidence in the region of subway construction (Lanari 1999).

Kolmogorov‐Sinai entropy is properly credited by Heinamaki et al. (1999), and if you get it wrong you have to wander for 40 years.

A "bad‐fish zone" occurs in a plot of profile shapes for elliptical galaxies, derived from various kinds of stellar orbits by Zhao (1999). There are also some bananas and pretzels, and your fate, should you try to swallow them all, is foretold in the multi‐versed song that begins, "Found a peanut ..."

Most of the early women astronomers worked closely with male relatives. Add Sophie Brahe (1536–1643, sister of Tycho) to the more familiar Elisabeth Kropmann Hevelius (1647–1693, a wife), Caroline Herschel (1750–1848, a sister), and Margaret Huggins (1848–1915, a wife, who could conceivably have met Caroline H. but probably did not).

The farming of fishivores can use up more pounds of fish protein than it produces (Naylor et al. 2000).

Descartes lives, or at least something rather similar to his vortex model for the formation of planets (Godon & Livio 2000).

The forest method is, of course, an offshoot of tree codes (Yahagi et al. 1999).

MOCS is a way of apportioning credit for citations to multi‐authored papers among the authors (Jones 1999). Given the high level of wit manifest in this (once again) weekly column, it is ominous that the author has left for us the obvious remark about not judging a colleague until one has walked a day in his MOCS.

TLAs: The best set of the year appeared as section heading in Blandford's (1998) concluding remarks at a conference whose proceedings appear in Astrophysics and Space Science Vol. 261. The worst are the ones that duplicate a set already in use, for instance RGB for ROSAT‐Green Bank survey rather than red giant branch (Laurent‐Muehleisen et al. 1999). The paper is about the ratio of radio to X‐ray luminosity from BL Lac sources. What is a TLA? Oh, a three‐letter acronym, of which there are only 17,576 possible. We would like to claim dibs on our own initials, but they are already in use.

Giordano Bruno is not, it seems, to be pardoned (Sodana 2000).

Asperger's syndrome is over‐represented among people in mathematics, science, and engineering (Baron‐Cohen et al. 2000). Normally, we would make you look it up, but it spoils the joke not to realize that the syndrome is a borderline version of autism, often associated with great intellectual gifts but very limited "people" skills.

CP and PN are among the newly detected molecules in IRC +10216 (Cernicharo et al. 2000), and it is not in any way the fault of the molecules or the authors if we thought initially that they should be (a) violated and (b) centered on a PNN.

L6 is a class of semi‐regular variables as well as a spectral type of brown dwarf (Kerschbaum 1999), and so much for the reason that type P was rejected (Ap98, § 10.5).

Dendrograms are a way of showing hierarchical clustering (Durret et al. 2000) and work like clad diagrams for illustrating evolutional relationships. Dendrochronology is tree‐ring dating. The root is the Greek word for tree, and if the derivatives make you think of dandruff, it is probably because they are deciduous trees.

MyCn 18, He 111, KjPn 9, and Mz 3 are all planetary nebulae with supersonic outflows (Redman et al. 2000). The first two are short for Mayall (Margaret, not Nicholas) and Cannon and Henize. As for the others ... .

Oosterhoff III globular clusters have larger average periods for the RR Lyrae stars than Oosterhoff II's (I's are the shortest), but the two known examples also have the largest metallicities, among the clusters, reversing the trend from high Fe/H in type I's to lower Fe/H in type II's (Pritzl et al. 2000).

Whether the Casimir force really takes place between actual, rather than gedanken, metal plates is still being debated, we were surprised to hear, at the two or three!?! level (Bordag et al. 2000). Sadly, it is no longer possible to ask Casimir.

Epagomenal, embolismic, and bissextile all appear in a book review (Yallop 2000) without definitions, but we suspect they have nothing to do with digestion, failure of the circulatory system, or, um, er ... .

Aging observers: Stanton (1999) finds no correlation between observers' ages and the errors they make estimating magnitudes, up anyhow to age 65. The author of Observatory, 120, 188, who is approaching that limit, however, concluded that it would not be possible for him to delay publication until the orbit of HD 146117, whose period is just 7 days longer than 4 years, moved its phase = 0° around to the time of year when it is up.

Very remote observing: It has been called to our attention that the Wyoming 2.3 meter telescope has been operated from Saratov State University.

Cenek (Vicenz) Strouhal, about whom we enquired last year (§ 11.1) was a pupil of Ernst Mach and worked on singing wires (e.g., Strouhal 1878).

NOIDS are non‐optical identification objects (PASP, 112, 304, results). Iogenic means "made by Io" (AJ, 120, 488, abstract). FREDs are part of the title of ApJ, 534, L35, and we didn't catch the definition. CaTs are interpreted in A&A, 359, 18. Radix‐4Butterfly is part of a title (PASJ, 52, 447) and leaves one wondering whether it is an astronomical object, a data processing technique, or something else.

Milankovich cycles as the cause of ice ages were proposed about 50 years earlier by Croll (1875). Given the customs of eponymity, this proves we have the right name for them.

Herbig‐Haro object numbers: If you would like to keep count, along with the numbers of McDonalds hamburgers sold, the website is http://casa.colorado.edu/hhcat/.

Maryland is the best‐funded state in the Union, measured by federal R&D dollars per resident, and, before you try to descend, locust‐like, on one of our institutions, remember that NIH is in Bethesda. The level is $1513 per resident (Science 288, 2111). New Mexico is number two, but remember LLNL and Sandia before you try to cart off the VLA!

"Subdwarf," about which we earlier enquired, was coined by Kuiper (1939). Not surprisingly, his invention quickly won out over the earlier "intermediate white dwarf."

Dionysius Exiguus (of the missing year zero) was born in what is now Romania, and James Kaler claimed to be the only native Albanian astronomy (i.e., native of Albany, NY).

And Alan Cousins (2000) holds the record for astronomer who published over the longest period of time. We would gladly have sung a suitable paean of praise if we had known him a bit better.

The dictionary meaning is "growth or increase by external additions," implying that something is already there. Thus accretion is, at most, the last part of a formation process (e.g., from disks around young stellar objects or accretion of dwarf companions by larger galaxies; §§ 9.2 and 9.3), and may be completely distinct from it, as in the case of accretion onto evolved stars or black holes. We take a look at some of these, from inside in the case of the Milky Way, from outside for the others (especially the black holes).

6.1.1. High Velocity Clouds (HVCs)

6.1.2. Cooling Flows

6.1.3. Accretion Disk Instabilities

6.1.4. Advection‐Dominated Accretion Flow

6.1.5. Young Stellar Objects

"Consider a spherical cow" is a standard joke at the expense of physicists or engineers (depending on who is telling it) and is probably not a bad approximation for calculating heat loss, evaporation, and food and water requirements. The next step is, of course, the oblate or bipolar cow. ("One end is moo the other milk" said Ogden Nash⁴), and a remarkably wide range of subjects of astronomical investigation are, to second order, bipolar, collimated, or jet‐like in their excreta. The classes we found were young or youngish stellar objects (including Herbig‐Haros and Eta Car), planetary nebulae, cataclysmic variables, supernovae, X‐ray binaries, gamma‐ray bursters, active galactic nuclei (where one of the truly ancient questions is whether typical jet are one or two sided), and undoubtedly we forgot to look for some others. The Sun is apparently not an example, with oblate corona and a wind of more complex topology (§ 2.6).

6.2.1. Young Stellar Objects

6.2.2. Planetary Nebulae

6.2.3. Cataclysmic Variables

6.2.4. Supernovae

6.2.5. X‐Ray Binaries and "Microquasars"

6.2.6. Gamma‐Ray Bursters

6.2.7. Active Galactic Nuclei

This acronym, which stands for stellar collisions and mergers, was the title (innocently adopted, the organizers said) of a three‐day workshop, which the less innocent author attended, coming away with the expectation that there would be lots of papers on the topic during the year. She was wrong, hence the relative compactness of the section.

The SX Phe stars and other blue stragglers (stars seemingly too hot, massive, and young to belong to the neighborhood) continue to be the class for which collisions and mergers in binary systems are most often invoked (Ap91, § 5). Some are binaries and some are not (Preston & Landolt 1999). Analysis of their pulsation modes remains the best mass indicator for the single ones (Zhou et al. 1999), and the name "dwarf Cepheid" does not add any information and should probably be discontinued.

Blue stragglers are commonest (or at least easiest to recognize) in globular clusters (Rood et al. 1999 on M3), and the distribution of their ages in 47 Tuc may be telling us that the cluster went through core collapse a couple of billion years ago (Sills et al. 2000). In the halo field there are many blue, metal‐poor, intermediate age, binaries with small mass functions (Preston & Sneden 2000). These must have arisen out of the mass transfer process proposed by McCrea (1964) rather than from complete mergers. In any case, multiple origins are required. On the one hand, mergers appear unlikely in the four open clusters studied by Andrievsky et al. (2000) where the stragglers rotate more slowly than the other stars. And, on the other hand, collisions of stars not previously in binary pairs is unlikely for NGC 6791 which has a small central density (Yong et al. 2000). Stellar encounters, collisions, envelope‐stripping, and mergers may or may not have anything to do with making some globulars bluest at their centers (Howell et al. 2000).

Other sorts of stars for which SCAM may be part of the story include (a) luminous blue variables (Pasquali et al. 2000), (b) helium stars with large surface gravity (Saio & Jeffery 2000), (c) the disk of Beta Pic (Kalas et al. 2000), and (d) the most massive stars, which are both rather difficult to form through a single accretion process and preferentially found in large, dense star clusters (Massi et al. 2000; Figer et al. 2000). Ballantyne et al. (2000) showcase an isolated O8.5 V star, asking whether a merger origin is possible and coming down firmly on the fence, while Norberg & Maeder (2000) definitely favor accretion. In any case, stars formed in dense environments like the galactic center, will inevitably be vulnerable to collision damage (Alexander 1999), though the collision damage waiver is almost never a good investment.

Stars near the centers of active galactic nuclei also find themselves well within the madding crowd, and stellar collisions probably contribute to AGN variability (Torricelli‐Ciamponi et al. 2000), an idea that can be traced all the way back to the first, 1963, Texas Symposium (Gold et al. 1965).

Galaxies and clusters thereof are forever tearing each other apart and putting the pieces back together again. Indeed the processes are really a continuum, since loss of identity (especially for the smaller partner) is a prerequisite to becoming incorporated in the larger whole (almost enough to make a dwarf galaxy feel religious).

Thus, "What is a galaxy?" and "Is it a group or a galaxy?" are not such silly questions as they sound to astronomers who remember when galaxies were "the basic building blocks of the universe." Even with so casual a definition as "something that feels its own self‐gravity more strongly than the effects of its neighbors" you may have to look hard to see the difference. On the one hand, for instance, HCG 18 is really a single galaxy with star‐forming clumps (Plana et al. 2000) on the grounds that the velocity field is continuous across it. HCG 31 and 54 may be similar (and the only other cases in the Hickson catalog of 100 or so compact groups). But, on the other hand, something like 20% of what are usually described as ultra‐luminous infrared galaxies are really small groups with mergers in their immediate past or future (Borne et al. 2000). Evans et al. (2000a) show some of the details for IRAS 14348−144, where, so far, only the less luminous galaxies have been robbed of their individual gas components. Jones et al. (2000c) and Romer et al. (2000) identify additional group mergers in progress from X‐ray images. If compact groups resist this process for longer than you would have expected, it may be because they actually contain many more faint galaxies than are usually included in the dynamical calculations (Tovmassian & Chavushyan 2000; Murayama et al. 2000, reporting a seventh member of the Seyfert Sextet; it plays piccolo).

Even our own beloved Local Group will someday be a giant elliptical, with specific frequency of globular clusters = 3 (Forbes et al. 2000). Not having figured out any better place to put it, we will here mention that the Milky Way has, of late, come to seem a much more equal, or even dominant, partner with M31 than it used to be (Wilkinson & Evans 1999; Loinard et al. 1999; García Cole et al. 1999; Côte et al. 2000a; Evans & Wilkinson 2000; Anguita et al. 2000; Evans et al. 2000b; Dehnen 1999). For more on the Local Group (whose zero‐velocity surface has moved out a bit to 1.18 Mpc) see the reviews by van den Bergh (1999, 2000c).

Also in small groups of galaxies you find (or rather Holmberg did) the Holmberg (1969) effect. It is the relative deficiency of satellite galaxies with projected positions close to the equatorial plane of the dominant galaxy, combined with a greater extent of the satellite systems along the polar axis. The mechanism could be either destruction or inhibition of formation (Zaritsky & Gonzalez 1999). Hartwick (2000), looking at the Milky Way and M31, concludes that the main contributor is preferential accretion of companion galaxies from along filamentary large scale structures.

The alternative fate for a group of galaxies that doesn't settle down into a single potential well is to gather with other groups to become a cluster, which itself may find other clusters to commune with, and so forth. Iwasawa et al. (2000), for instance, suggest that the X‐ray morphology and high temperature for its luminosity of the Zwicky cluster 1718.1−0108 mean that it is actually three or more groups caught in flagrante delicto. That such crimes are common follows from the evidence for guilt (substructure) displayed by both of our favorite, nearby clusters, Virgo (Lee et al. 2000a; Kikuchi et al. 2000) and Coma (Watanabe et al. 1999; Bravo‐Alfaro et al. 2000). Common even in recent gigayears (Patton et al. 2000), and even commoner in the past say Reshetnikov (2000a, 2000b) and Le Fevre et al. (2000), reporting evidence for redshift dependences of (1 + z)^{4 ± 1} and (1 + z)^{3.2 ± 0.6}, respectively.

Evidence for mergers can include both substructure in whole clusters and damage to individual galaxies. X‐ray images and temperature distributions are particularly good indicators of ongoing and past processes (Molendi et al. 1999; Kikuchi et al. 2000; Takizawa 2000; and Molendi et al. 2000 on Abell 2256, where the crim. con. has also disrupted a former cooling flow relationship). Other sources of complimentary information include weak gravitational lensing (Umetsu & Futamase 2000) and old fashioned optical photometry (Metevier et al. 2000 on the Butcher‐Oemler effect in two clusters).

Both Metevier et al. and Maurogordato et al. (2000, on Abell 521) are trying hard to push us on to the next subtopic, because they point out that individual galaxies as well as the overall cluster structure show collision damage. Before following them, we pause only to note (a) that Abell 222 and 223 will surely merge in the future (Proust et al. 2000) to form Abell 2221 /2, or maybe Abell 445, and (b) that Chiueh & Wu (2000) have provided an alibi in the form of a mechanism to assemble clusters with two cores even without any merger intervening. This may or may not be part of the explanation for close pairs of galaxies that seem to show no evidence for interactions (Zasova et al. 2000).

"Il faut souffrir pour être belle," the Farmer's daughter used to say. For galaxies, it is a bit the opposite—the most elegant, gas rich, grand design spirals are generally found only where nobody is around to bruise their arms. Indeed, the gas is the first to go, yielding first to tidal removal and then to ram pressure stripping (Lee et al. 2000a) and leaving anemic spirals (Vollmer et al. 1999 on NGC 4548 in Virgo) and trails of gas and radio‐emitting plasma behind them through the intracluster medium (Henriksen et al. 2000; Stevens et al. 1999a; Drake et al. 2000; Clemens et al. 1999). The stars, too, get pulled out and about, as you have known since Toomre & Toomre (1972) told you so. Curiously, gas and stars can leave on quite different trajectories (Hibbard et al. 2000, on three Arp galaxies). And, in due course, off also comes part or all of the extended, dark matter halo, to join a communal pool (Sensui et al. 1999), leaving the rest of the galaxy even more vulnerable (Dubinski et al. 1999).

Grenacher et al. (1999) suggest that our own galactic halo may have been flattened by interactions with the Magellanic Clouds, leading inexorably to

The relatively new, big question here is whether the halo stars are largely or even all rubble from tidally disrupted dwarf spheroidal galaxies, whose relic cores are the globular clusters, or a subset of them. Notice that this makes at least numerological sense. If a (proto) globular cluster is the 10⁵ M_⊙ core of a 10⁷ M_⊙ dSph, then the total stellar mass of the halo should be roughly 100 times the 10⁷ M_⊙ found in the 200 globular clusters. This is at least approximately right. A more detailed argument in favor is that about 10% of the low metallicity stars in the outer halo have kinematic properties suggesting that they came from a single disruption (Helmi et al. 1999). A project is now under way to acquire enough kinematic data to look for similar residual structure through most of the halo (Morrison et al. 2000).

Destruction is certainly a present phenomenon. The Large and Small Magellanic Clouds both show damage (including loss of gas in a stream) and will not be with us for many more orbit periods (Weinberg 2000; Kunkel et al. 2000). Most of our favorite dwarf spheroidals are also losing some of their marbles, er, stars (Majewski et al. 2000 on Carina, Johnston et al. 1999 on IGI, Colpi et al. 1999 on Fornax and IGI). If IGI (the dwarf spheroidal in Sagitarrius) has no Magellanic‐style gas stream (Burton & Lockman 1999), it is only because it has no gas to speak of. Fornax has managed to hang on to an elderly globular cluster of its own (Oh et al. 2000), but van den Bergh (2000b) believes that Fornax is not a representative of the class conceivably ancestral to the outer globular clusters, while IGI (which also has one of its own) is.

Do the globulars themselves vote in favor of this scenario? Well, ω Cen (our most massive) has stars of several ages as well as of several metallicities (Lee et al. 1999b), with stellar ages and abundances correlated in the same way they are in dSph's (Hughes & Wallerstein 2000; Smith et al. 2000a; Pancino et al. 2000). A prediction is that ω Cen should have its own dark matter halo (Carraro & Lia 2000), which one might find by measuring velocities of stars more than 20 pc from its center. That it, like nearly all other galactic globular clusters, has a tidal tail of stars being lost (Leon et al. 2000) is perhaps an argument against significant dark matter in the cluster. On the other hand, it does have just about the smallest of these tails. "Stars being left behind" is, by the way, not the right way to describe these fugitives, since the "tails" are expected to extend along the cluster orbits in front of them as well as behind (Combes et al. 1999). These are, we suppose, by analogy with elephants, tidal trunks (or moos, in the case of the bipolar cow).

Another long‐standing issue in galactic kinematics is the relationship among star formation processes in the disk, open clusters, moving groups, star streams or other relics of open clusters and associations, and the general field population of the disk. Open clusters anything like as old as we think the disk must be are, as you know, rare (Carraro et al. 1999 on NGC 6793 and Berkeley 17; Sarajedini et al. 1999). Many young clusters are, as first highlighted by Adriaan Blaauw and V. A. Ambartsumian more than 40 years ago (separately!), expanding (Mirzoyan et al. 1999), while the cluster within which our Sun formed is long gone (Ida et al. 2000a, on what the other stars did to Kuiper Belt Objects). In between come the star streams, like the one allied with the Tucana association (Zuckerman & Webb 2000). The stars concerned are all quite young, but they make up part of what Eggen (1992) called the Pleiades group, which includes much older stars.

Clearly lack of coevality is a fatal objection to regarding a bunch of stars as having formed in a single open cluster, but we think that Dehnen (2000) is proposing that our location, just outside the outer Lindblad resonance of the Milky Way bar, makes it possible for stars with similar location (in the rotating coordinate system) and kinematics to form over a considerable fraction of a galactic rotation period. Maturity provides no absolute protection to clusters. HR1614 and its associated moving group are something like 2 Gyr old and were gravitationally bound until a rotation period or two ago (Feltzing & Homberg 2000).

We had intended to close this section with words of wisdom about Gould's Belt, but are slightly put off by the probability that its stars (including the low mass ones now recognized, Makarov & Urban 2000, and gas—Lindblad's ring) were assembled by a local explosion, rather than a star formation event (Poeppel & Marronetti 2000). It is not clear that this explains the observation that the whole systems seems to be rotating (Comeron 1999), but then neither do the competing hypotheses of giant star formation event or impact of a high velocity cloud. The kinematic time scale of the rotation is 3–6 × 10⁷ yr, like the ages of the brightest member stars (Torra et al. 2000), and this is also at least a plausible age for a population of intrinsically weak gamma‐ray sources that are found in the same part of the sky (Gehrels et al. 2000), if they are fading pulsars (Grenier 2000).

Gould's Belt is stored in our mental file drawer very close to the radio emission feature called the North Polar Spur (probably because they appeared in the same graduate course on galactic structure), and somehow this seemed to imply that both should also be fairly close to Earth by astronomical standards. True for Gould's Belt, but perhaps not for the spur (despite its enormous angular size). Sofue (2000) has attributed it to a starburst at the galactic center.

Serious archeologists do wonders with kitchen middens, so perhaps you can draw conclusions that we have not from the possible‐to‐probable existence of (a) star streams in the giant ellipticals of the Virgo and Coma Clusters (Pierce & Berrington 2000), (b) disrupted galaxies in the Coma and Centaurus Clusters (Calcaneo‐Roldan et al. 2000), (c) remains or even ghosts of the globular clusters originally owned by galaxies that participated in mergers near the centers of most rich clusters (Blakeslee 1999; Gebhardt & Kissler‐Patig 1999; Capuzzo‐Dolcetta & Tesseri 1999), (d) pairs of star clusters in the process of merging (De Oliveira et al. 2000), and (e) at least one runaway star cluster, moving at 50 km/sec perpendicular to the disk of the Large Magellanic Cloud (Graff & Gould 2000). It is the cluster in which the progenitor of SN 1987A seems to have originated and is apparently escaping from the scene of the crime.

Dust is remarkably ubiquitous in the universe, with even an intergalactic component occasionally invoked to explain the apparent faintness of distant supernovae (Aguirre & Haiman 2000). We logged in roughly 40 dusty sites, nine components, and seven processes. Here appear a small subset of the most and least familiar.

A very broad absorption feature near 2200 or 2175 Å was an early discovery of ultraviolet astronomy. Common in the Milky Way (though strangely not in nebulae where the exciting star can be identified; Zagury 2000), it is by no means spread over all other galaxies, and no QSO is known to display the feature (Pitman et al. 2000, pointing out that an earlier report was a false alarm). The least exotic carrier mentioned during the year was graphite with a range of grain sizes (Will & Aannestad 1999).

The most exotic constituents we picked out were magnesium protosilicates (blamed for a 2 μm feature in Cas A and elsewhere; Chan & Onaka 2000) and nanodiamonds (credited with some 3–15 μm bands; Jones & d'Hendecourt 2000). The former is a new dust candidate; the latter is, of course, a nanogirl's best friend. And there are lots of ices, H₂O outweighing all the rest put together by a factor near three (Gibb et al. 2000). The rest, in order of declining abundance toward the protostar W33A are CH₃OH, CO₂, CH₃, HCO, NH₃, CO, HCOOH, H₂CO, XCN, CH₄, HCOO⁻, HDO, C¹³O₂, and OCS, at one‐fivehundredth of H₂O.

Places where dust is found (in addition to back of our refrigerator) include:

• •  
  The B[e] X‐ray binary CI Cam (Clark et al. 2000a).
• •  
  The cooling‐flow clusters of galaxies Sersic 159‐03 (Hansen et al. 2000) and Abell 2670 (Hansen et al. 1999).
• •  
  The winds coming off hot stars, including luminous blue variables (Parthasarathy et al. 2000; Voors et al. 2000), some but not all Wolf‐Rayet stars (Williams & van der Hucht 2000; Cherchneff et al. 2000), and post‐AGN stars (Arkhipova et al. 2000).
• •  
  Novae, both before (Kawabata et al. 2000 on V4444 Sgr, 1999) and after the explosion (Taranova & Shenavrim 2000 on V1016 Cyg).
• •  
  Symbiotic stars, where the polarization caused by dust scattering varies in days to months (Brandi et al. 2000).
• •  
  Some but not all ultracompact H ii regions, if defined by their radio emission (Molinari et al. 2000b).
• •  
  Planetary nebulae (too many papers to cite, from Stasinska & Szczerba 1999 to O'Dell 2000, on the Helix nebula).
• •  
  Not‐so‐compact H ii regions (Orsatti et al. 2000),
• •  
  The same E/SO galaxies and AGNs that have gas (Tomita et al. 2000).
• •  
  The Moon, some of whose dust ought to reach Earth's upper atmosphere, though apparently none has ever been seen (Yamamoto & Nakamura 2000).
• •  
  Comets, whose dust has been variously described as just interstellar stuff glued into various‐sized grit (Evlanov et al. 2000) or as the product of post‐condensation processing (Rietmeijer et al. 1999). The break‐up of Comet LINEAR into six chunks suggests a size scale between dust and nuclei, to be called cometesimals.

Formation and destruction of grains must be the dominant processes, in the way that birth and death are for humans, that is to say, not necessarily the most interesting ones. In between, analogous to getting your first job, comes alignment (Smith et al. 2000, the first demonstration that it happens to carbon‐rich as well as oxygen‐rich grains, so that both can produce dichroic polarization). Magnetic fields are part of the deal (Straizys et al. 1999) in a tradition that goes back to Davis & Greenstein (1951). Pick your own analogy for the observation that dust is frequently also flaky or fluffy (Vaidya et al. 2000; Blum et al. 2000) and dizzy or spinning (de Oliveira‐Costa et al. 1999; Finkbeiner et al. 1999).

And if you fear that nanodiamonds might be too expensive, there is also spinel, MgAl₂O₄ to be had at the interstellar jewelry store (Posch et al. 1999).

The other 99% of the interstellar medium is gas, with all the usual elements in the usual proportions, minus "depletion," meaning whatever is condensed out on the grains (Welty et al. 1999, chosen from among many papers on the topic because it is, probably, the last on which Lyman Spitzer will appear as an author). Four fairly old questions and some partially‐new answers follow.

What is the ratio of H₂ to CO, and is it constant, so that the latter can be used as a proxy for the former? Really big in starbursts, and so no, say Bergvall et al. (2000). Apparently small in Seyfert galaxies, and so no, again, say Papadopoulos & Allen (2000), but much the same in the bulge of M31 as in the solar neighborhood (Melchior et al. 2000a). For explanation of the units in which the average value is H₂/CO = 2 × 10²⁰ H₂ cm^-2 (K km s^-1)^-1; see Ap95, § 7.2.

Just how much deuterium enrichment is possible in molecules? Spectacularly, up to N₂D⁺/N₂H⁺ = 0.35 in a couple of cloud cores, report Tine et al. (2000).

Where is most of the oxygen in molecular gas, since it is neither O₂ nor H₂O, according to Goldsmith et al. (2000) and Bergin et al. (2000), both part of the set of first results from SWAS. No answer was provided, but our candidate is sodium bithusilate.

Where is the warm absorber? Well, all right, first, what is the warm absorber? What it absorbs is X‐rays, coming from the nuclei of Seyfert galaxies, and the "warm" part is dictated by the presence of absorption features due to ions like O vi. Things said about it during the year include:

• •  
  It's an MHD wind from the black hole accretion disk (Bottorff et al. 2000).
• •  
  The UV and X‐ray absorbing gas are not mostly the same stuff, even when they are at about the same temperature (Kriss et al. 2000a, 2000b).
• •  
  The outflow velocity is about 440 ± 220 km/sec for the Fe xviii–xxi (X‐ray absorbing) stuff in NGC 3783 (Kaspi et al. 2000a). And, though the outflow velocity and line width are comparable in X‐ray and UV gas, relative strengths require two components. In other galaxies, the gas warm enough to emit and absorb O viii X‐rays is flowing out somewhat faster than the gas which absorbs the UV lines (Kaastra et al. 2000). These two slightly contradictory results both derive from Chandra data.
• •  
  You need not just two but three absorbing zones, or a continuum, to fit all the observations of MCG −6‐30‐15 (Morales et al. 2000), a somewhat better‐known AGN than you might think from its name.
• •  
  And a mini‐campaign on NGC 1068 reveals that gas near 10⁶ K both emits and absorbs over a wide range of wavelengths (Lutz et al. 2000; Nicastro et al. 2000; Alexander et al. 2000).

"Do you guys have a G dwarf problem?" remains the question we would most like to ask when intergalactic communication has been established. Our G dwarf problem is that there are fewer ones of very small metallicity in the solar neighborhood than you would expect from a simple model of gradual enrichment in a homogeneous, closed box, where the same proportion of high and low mass stars form in each generation. The local sparcity of stars with [Fe/H] - ≤4 (Arghast et al. 2000) and even <−3 (Anthony‐Twarog et al. 2000) persists. The standard solutions, as you might expect, do away with one or more of the simple assumptions (Martinelli & Matteucci 2000), and, if your first extragalactic contact is with someone in the Large Magellanic Cloud (not totally unreasonable!) the answer will apparently be, "yes" (Dirsch et al. 2000).

"Stars of Negligible Content" (Reiz 1954), as the Astrophysical Journal ⁵ running head of long ago charmingly called them, continue to elude us, and it is not, at least for smallest masses, just that they have managed to make or acquire enough CNO in the interim to affect their evolution or hide their negligibility (Weiss et al. 2000), though Preston (2000) notes that it is not clear just how many you would expect to find, even for the simple, closed box (etc.) model.

Creeping cautiously away from Z = 0, we find the age‐metallicity relationship. Two recent samples (Garnett & Kobulnicky 2000 and Rocha‐Pinto et al. 2000) show much tighter correlations (real scatter ≤0.15 dex) than earlier work (Edvardsson et al. 1993). We are told that the new results are better and that the Swedish sample may have had some odd bias, but the Swedish samplers do not appear to have been consulted.

It is not news, nor even your father's Olds, that the local average metallicity for even the youngest stars is less than solar (Rocha‐Pinto et al. 2000; Garnett & Kobulnicky 2000; Korotin et al. 1999; Andrievsky et al. 1999).

Uncertainties multiply as we look outward to other galaxies and their stellar populations, which must generally be studied from integrated spectra and colors. The core problem, which the more adjustable author continues to think of as "the curse of the adjustable parameter," is that, even for a single, isochronal population, the effects of total metallicity, variations from solar element ratios, population age, duration of starburst, and initial mass function can be very difficult to distinguish. The difficulties grow when you are examining the sum of many generations of stars, in which all of the above may have changed, and gas with more or less than its fair share of heavy elements flowed in or out as well.

Overviews of the star population problem are given by Nakata et al. (1999) and Vazdekis & Arimoto (1999). Kennicutt et al. (2000) discuss H ii regions, for which gas temperature is an additional confounding variable. We caught something like three dozen papers focused primarily on some particular indicator for some particular parameter and (rather arbitrarily) note here (a) Yi et al. (2000) on convective overshoot and metallicity as uncertainties in determining ages of galaxies near z = 1.5 (the range is 1.5–3.5 Gyr), (b) Tamura & Ohta (2000) on age versus metallicity as the dominant factor for color gradients in elliptical galaxies at moderate redshift, (c) Pilyugin & Ferrini (2000) on the combination of star formation rate versus time with loss of metals in winds for S and Irr galaxies, (d) Molla & Garcia‐Vargas (2000) on fitting indices of line strengths in elliptical galaxies with a range of a gas and metallicities, and (e) Ferreras & Silk (2000) on the degeneracy between the efficiency of star formation and gas outflows. This topic was initially indexed under "don't hold your breath," and the cited papers are generally very frank about the continuing large uncertainties in deconvolving the major contributing factors.

The possibility of variations in the interstellar ratio D/H is a part of galactic chemical evolution with cosmological implications. Last year, we emphasized evidence for real variability, so fairness requires citing this year Bluhm et al. (1999), who say that all the measured ratios are consistent with 1.5 × 10^-5 at the 1–2 σ level. At one time, the interstellar value, augmented by however much deuterium you thought had been destroyed in stars, was the best available estimate for the cosmic value. See Dolgov & Pagel (1999) and Lemoine et al. (1999) for interesting takes on the possibility that both the processed interstellar deuterium abundance and the unprocessed, intergalactic D/H may vary.

The distance to the galactic center was roughly 0 kpc to William Herschel and Heber D. Curtis (and many in between), 20 kpc to Shapley, and has cycled several times between 10 and 8.5 kpc since Oort (1927, 1928) entered the fray. A double handful of values published during the index year (many with three "significant" figures) had a mean of 8.1 kpc. Only one set of error bars took in everybody else's numbers, R₀ = 7.2 to 8.6 kpc from Beaulieu et al. (2000). Mishurov (2000) noted, not for the first time, that we are quite close to the corotation radius, and Salim & Gould (1999) predicted that, within 30 years, we will know R₀ to better than 1% from the analysis of orbits of stars around the central black hole.

For the local rotation speed, 220 km/sec remains a popular value (Nakashima et al. 2000, who also find that the circular speed declines to 175 km/sec 5 kpc outside the solar circle). But we caught somewhat smaller numbers (about 215 km/sec in Beaulieu et al.) and considerably larger ones (270 km/sec, according to Mendez et al. 1999). The local escape speed exceeds 420 km/sec (García Cole et al. 1999). And the combination of Oort constants, A–B, which should be the circular velocity divided by R₀, is 23–27 km/sec, says Mignard (2000), who looked at Hipparcos distances and proper motions for 20,000 stars.

Do other galaxies (sosies, de Vaucouleurs used to call them) resemble the Milky Way? Undoubtedly, though which and how is harder to say. NGC 4527 has the most similar rotation curve (and you have just seen how well ours is known) according to Sofue et al. (1999). And NGC 891 has a similarly complex, multi‐phase interstellar medium (Howk & Savage 2000). The Milky Way also has (and analogies to all can be found in other galaxies):

(a) a bar, which can be traced out to the solar circle (Feast & Whitelock 2000),

(b) spiral arms, four grand design ones says Sevenster (1999), two on our side and two on the other (roughly analogous to how you get six elephants into a Volkswagen), and a large contrast in gas density between the arms and interarm regions (Ungerechts et al. 2000),

(c) a bulge containing 3–4 × 10¹⁰ stars (Lopez‐Corredoira et al. 2000), which is just what it should have to go with its 2.6 × 10⁶ M_⊙ black hole (§ 11),

(d) a disk with a warp, which is either short‐lived (Drimmel et al. 2000) or evidence in favor of Modified Newtonian Dynamics (MOND; Brada & Milgrom 2000) or both, or, we trust, neither,

(e) three local stellar populations, defined from both kinematics and composition (Caloi et al. 1999).

(f) an usually small, or badly misdetermined ratio of radius at which the optical disk cuts off to scale length (Pohlen et al. 2000), and

(g) H ii regions with helium abundance as large as Y = 0.28 (Deharveng et al. 2000). This says that the sum total of chemical evolution over galactic history has produced ΔY/ΔZ close to 3 (larger than many modelers of stellar nucleosynthesis would have expected). Deharveng et al. also note that even O6.5 stars produce helium Stromgren ameboids smaller than their hydrogen Stromgren ameboids. This is, of course, the sort of factoid that casts doubts upon values of the primordial helium abundance determined from H ii regions in metal‐poor galaxies. In any case, the "observed" value of ΔY/ΔZ remains poorly known. Ribas et al. (2000a) report 2.2 ± 0.8, found by forcing both members of each of 50 double‐lined, spectroscopic, eclipsing binary systems to have the same X, Y, Z, and age, while Guenther & Demarque (2000) deduce ΔY/ΔZ≲1 from a similar‐analysis of Alpha Cen A and B.

And if there is anything else you would like to know about the Milky Way, consult the set of review articles introduced by Irion (1999).

The Local Group has a new member (Whiting et al. 1999), which can be found by looking through Cetus. (Have you ever looked through a whale? Did you see Giapetto? Jonah? Partly digested krill?) It is a dwarf spheroidal or elliptical and brings total LG membership to 36 (van den Bergh 2000c), extending to 1.18 Mpc (van den Bergh 1999), with some dynamical distortion caused by the much more massive Cen A group (van den Bergh 2000a). Andromeda IV remains excluded from the canonical LG, but it is perhaps a background galaxy to, rather than a star formation region in (Ap93, § 2.7), M31 (Ferguson et al. 2000).

The Large Magellanic Cloud has its center somewhere between 45.1 and 62.5 kpc from the center of the Milky Way (Groenewegen & Oudmaijer 2000). The first nine million LMC stars plotted on an HR or Hess diagram reveal a different history of star formation from that in the Milky Way, but with about the same start time (Alcock et al. 2000). These stars, along with the horizontal branch ones in the dwarf irregular WLM (Rejkuba et al. 2000) are at least part of the answer to the very old question of whether seemingly‐young galaxies actually have underlying old stars (yes, as a rule). A Hess diagram, like an HR diagram, is a plot of luminosity versus some indicator of effective temperature, but showing density of stars at each point rather than a separate dot for each star (Hess 1924). Who was Kienle (assuming you have just looked at the reference)? Martin Schwarzschild's thesis advisor, among other achievements. Who was Seeliger (Hugo von, 1849–1924)? He was one of the people who suggested that nova explosions happened when swarms of meteoritic material hit stars (with thanks to Cole et al. 1999a for getting us started on this path).

The spiral galaxies form a two‐parameter family (two papers), a three‐parameter family (four papers), or a four‐parameter family (only one paper, Bershady et al. 2000, contrary to what you were expecting). It is probably not true that each one of the 60+ papers on spiral galaxy structure during the year answered a different question (though they may each have given a different answer), but there were surely more than the five mentioned here.

• 1.  
  Do spirals have maximal disks, meaning the sort that puts as much mass into the central disk, bulge, and bar as the luminosity profile allows, eliminating the need for a core concentration of dark matter? Yes, it would seem, for some (Blais‐Ouellette et al. 1999) and no for others (Maller et al. 2000).
• 2.  
  Do spirals have non‐thermal halos, which was once primarily the question Does the radio emission from the Milky Way have a spheroidal component? Yes, for relativistic electrons in the Milky Way (Moskalenko & Strong 2000). Yes, but they are made of discrete loops and filaments (Irwin et al. 2000). And, not much, for NGC 5907 (Dumke et al. 2000).
• 3.  
  How much of the UV gets out? Less than 2% locally say Tumlinson et al. (1999), thus nearly all the local intergalactic background is due to active galaxies. Quite a lot from large star formation regions and starburst galaxies, say Witt & Gordon (2000) and Beckman et al. (2000). The Milky Way comes in between, with 3%–30% escaping, say Dove et al. (2000).
• 4.  
  Are there disks without bulges? Yes, UGC 7321 (Matthews et al. 1999), for example.
• 5.  
  Does NGC 5907 have an unusual halo? Yes (Zepf et al. 2000), though it probably has little to do with a population of faint stars tracing a dark matter potential (Ap94, § 5.8).

Elliptical galaxies featured in a comparable number of papers (but with the ratio of theory to observations running at 2:1, vs. 1:2 for spirals). Questions to which you might or might not have been wanting the answer include:

• 1.  
  Do they rotate? Some hardly at all, less than 30 km/sec at the radius where the circular velocity would be 400 km/sec in the cD NGC 1399 (Fornax A, Saglia et al. 2000). Some quite a lot, but they are at most, very faint X‐ray sources (Pellegrini 1999), to which M87, with V_c/σ_v = 300/450 at r = 32 kpc, would seem to be an exception (Cohen 2000).
• 2.  
  Are they segregated, meaning at systematically different places from spirals, especially in rich clusters? Yes, of course (as was probably known to Herschel, though not in those words), not even concentric with the S's in Abell 426, the Perseus cluster (Bruzendorf & Meusinger 1999). And the types don't all have the same distribution of luminosities either (Saracco et al. 1999, and a large handful of other papers during the year).
• 3.  
  Did they form from monolithic collapses of large, non‐spinning blobs or by repeated, hierarchical mergers of smaller units? Blobs, say Kawata (1999) and Hozumi et al. (2000). Hierarchically, say Papadopoulos et al. (2000, and many others, who unfairly, go uncited). And, of course, some of each, say Côte et al. (2000) and Gratton et al. (2000), both as it happens talking about the Milky Way, rather than ellipticals.

SO (lenticular) galaxies raise the traditional question, why? That is, why no young stars, conspicuous arms, or gas? Some combination of ram pressure, tidal stripping, and a last starburst, conclude Quilis et al. (2000) and Sadler et al. (2000). And you may wish to meditate on the profound question of why the starburst associated with the departure of the gas is always the last one (Quilis et al.) and why your missing car keys are always found in the last place you look. Yeah, we tried buying one of those devices that makes the keys chirp at you when you ask them where they are, but we misplaced the device. Still, they say you don't really have to start worrying until you can't remember what keys are for (this is not the same as being unclear on whether B744 is the department office and R602 the Xerox room, or conversely).

Dwarf galaxies must be loved by the Prime Mover, because she has made so many of them (like beetles and attorneys), indeed is apparently still making them, at least in galaxy mergers (Braine et al. 2000; Weilbacher et al. 2000; Alonso‐Herrero et al. 2000b; Johnson & Conti 2000). You might want to ask whether they have their own dark matter halos. And by now you know that we are going to tell you that some very probably do (Côte et al. 1999 on And II) and others may well not (Tamura & Hirashita 1999). Have we found all the dwarfs? No, not even those within 1000 km/sec and within the local super‐cluster (Schneider & Schombert 2000; Karachentsev et al. 2000a), especially the ones with low optical surface brightness and/or small H i mass. The missing ones are probably about as numerous as those already catalogued.

Are there dwarf galaxies in voids? Indeed there are only dwarf galaxies in voids, insist Popescu et al. (1999). Other examiners are not so severe though they agree that dwarfs are over‐represented (Vennik et al. 2000) and probably formed late in the history of the universe (Grogin & Geller 2000). But, as Grogin & Geller (1999) very properly remind us, the statistics aren't very good, because there are not very many galaxies in voids!

Galaxies live (mostly) in clusters, and, it seems, have always done so at some level. The z = 2.4 agglomeration of Lyman alpha emitters reported by Keel et al. (1999) is not a record (a z = 3.09 cluster from Steidel et al. 2000 and an out‐of‐period press release on the group around 3C 294 with z ≈ 4 may be), but it comes with the additional information that the sky coverage by z = 2.4 clusters is less than 0.4%, which is to say that there are not very many of them compared with z = 0 clusters, Abell's census of which was only one‐half complete (Holden et al. 1999). As for when clusters stopped forming, the answer is, "Not yet." Massive ones are still contracting and adding outliers from their surroundings (Clowe et al. 2000; Balogh et al. 2000).

Galaxy formation, or study thereof, remains a major industry, yielding 30‐some papers during the reference year. Kay et al. (2000) explore the consequences of varying 13 input parameters to the models, one at a time, to see which really matter. Unfortunately, the ones that do include both poorly‐known physics (like gas cooling rates) and aspects of the calculations themselves (like spatial resolution). And, as we have remarked ad nauseum, the formation of galaxies and larger scale structure remains TMIUPIMA (This is actually an acronym for "the most important unsolved problem in modern astrophysics," not the Telugu word for ingrown toenail).

Here are some aspects of the universe as a whole that did not seem to fit into § 12, including background radiation, source evolution, and re‐ionization.

8.5.1. Backgrounds

8.5.2. Long Ago and Far Away

8.5.3. Reionization and the Early History of Star Formation

Most of the things in the world—shoes, ships, sealing wax, cabbages, kings—are non‐photons. From the point of view of the astronomy, however, the inventory is pretty much limited to meteorites and Moon samples (mentioned in § 3), cosmic rays, neutrinos and other dark matter particles, and gravitational radiation. On the subject of other dark matter particles, the literature offered us only an upper limit to some product of the number density and cross‐sections of WIMPs (Abusardi et al. 2000). The only actual gravitational radiation data published during the year (Weber 2001) reported a statistically significant correlation between pulses recorded by a pair of bar antennas and events from the bursting pulsar.

8.6.1. Cosmic Rays

8.6.2. Neutrinos

The GRB year began with a compact review of the association with supernovae by Jan van Paradijs (1999) written very shortly before his death and triggered by the pair SN 1998bw and GRB 980425. Through the year, it became increasingly clear that this event(s) was/were neither a typical GRB, having a denser than average circumstellar environment (Li & Chevalier 1999) and being faint even for its duration (Norris et al. 2000) nor a typical supernova, having made rather a lot of Ni⁵⁶ for a core‐collapse event (Sollerman et al. 2000a) and displaying an assortment of spectral anomalies (Stathakis et al. 2000). If, however, you are going to use the peculiarities as evidence for this being a single event, then it is unfair to count among them that 1998bw was the first Type Ib/c (indeed any sort of Type I) supernova to be seen at photon energies exceeding 2 keV (Pian et al. 2000). The small inventory of possible additional associations was augmented by SN 1997cy ≟ GRB 970514 (Germany et al. 2000; Turatto et al. 2000), another supemova that seems to have been profligate in Ni⁵⁶ production. Some of its other nucleosynthetic products may have been swallowed by a black hole.

Most supernovae, even failed supernovae or collapsars, do not make GRBs, perhaps because they do not form black holes sufficiently rapidly or leave them spinning fast enough (Macfadyen & Woosley 1999). Black holes remain the energy source of choice (e.g., Brown et al. 2000a, who also propose the galactic superluminal source (microquasar) GRO J1655−40 as a post‐GRB). But no model goes unsung. Kaiser et al. (2000) alternatively describe microquasar radio behavior as the time‐resolved version of processes that Rees & Meszaros (1994) put into GRBs themselves (shocks along a relativistic jet). Wheeler et al. (2000) propose a magnetar for intrinsically weak GRBs like 980425. And Wang et al. (2000e) and Bombaci & Datta (2000) prefer a second collapse from neutron star to strange quark star to do the GRB part.

Ap99 (§ 7) left as an open question whether the Gamma‐Ray Bursters are all at heart the same sort of beast. The answer is now pretty clearly, "no". Majority opinion finds two classes, long duration events (a subset of which have X‐ray, optical, and/or radio tails, host galaxies, and redshifts) and short duration events, lasting at most a second, with somewhat harder spectra and, to date, no counterparts at other wavelengths, though HETE II may be changing this even as we write (or you read).

The long (etc.) ones are quite likely to come from collapse of a single, massive star core to a rapidly rotating black hole, so that the event location will still be near a star formation region; and the short ones are the product of mergers of pairs of neutron stars or neutron star plus black hole (Fryer et al. 1999; Hakkila et al. 2000). Closely associated is the conclusion that, while some of the perceived correlations among peak fluence, duration, spectrum, and total flux are the result of redshift and time dilation, there are also intrinsic correlations (Mitrofanov et al. 1999; Atteia 2000; Lloyd et al. 2000).

Within this picture, only the long‐duration events would be expected to share the redshift history of cosmic star formation (cf. Boettcher & Dermer 2000), and the "no host" problem largely vanishes. Some of the apparently bright events, located accurately with the Interplanetary Network or from afterglows are truly bright, far away, and in dwarf starforming galaxies below the detection limit (e.g., GRB 990519, Beuermann et al. 1999), while merger products have had time to move some distance away from the star‐formation regions where they were born.

The host galaxies observed so far seem to have roughly a Schechter distribution of luminosities (Schaefer 2000; Hjorth et al. 2000 on GRB 990712 in a galaxy with L = 0.2L^*). Are you surprised that optical tails can be seen from within star formation regions? Waxman & Draine (2000) were, and so sat down to calculate how much dust the gamma event itself would vaporize in the first 10 seconds. The answer is, along the axis of beamed gamma rays (the direction we are looking), a whole giant molecular cloud's worth.

Third classes or subclasses of several sorts have also been proposed, based on progenitors, environments, or burst properties. An NS/BH pair is less likely to be expelled from its host galaxy than a double neutron star (Belczynski et al. 2000) and, we suppose, will develop somewhat differently (Janka et al. 1999). Just what happens when jets slow down and dump their contents into their surroundings (the late afterglow phase) is bound to depend on whether the environment is a relatively dense circumstellar shell shed by a single massive progenitor, the average interstellar medium, or something still more tenuous. The set of real afterglows seems to include some of each (Chevalier & Li 2000; Frail et al. 2000a; Thompson & Madau 2000; Livio & Waxman 2000), the authors being divided on whether this should cast doubts on all such events coming directly from core collapses.

Alternative third classes whose distribution on the sky or in space is different from the main two (isotropic!) sets were proposed by Cline et al. (1999a, durations less than 0.1 sec and V/V_m and log N–log S Euclidean), by Belousova et al. (1999), and by Meszaros et al. (2000, intermediate duration and dim). No specific progenitors were suggested.

Statistical studies rest heavily on completeness of samples. Have we been getting enough GRBs? The raw BATSE data can be mined for events just below the threshold at which you get an automatic shout of "burst here" from the software. Kommers et al. (2000) have looked in this twilight zone, finding mostly long‐duration bursts with V/V_m = 0.177 (meaning that half the events are in the closest 17.7% of the volume surveyed, whatever it is). They conclude that more sensitive detectors would not have increased the yield much above the one event per day seen. Stern et al. (2000b) on the other hand, find about as many events below the trigger flux as above, and conclude that number versus flux received flattens but does not turn over. Neither paper seriously disagrees with the conclusion that events we have not seen are a lot like the ones we have.

A quick summary of the energy situation for GRBs is that, with a reasonable degree of beaming, any of the suggested prime movers can provide enough (Meszaros et al. 1999), while without it, none can (e.g., 10⁵⁴ ergs from GRB 990123; Briggs et al. 1999).

All estimates of the amount of energy you "save" this way are, in some sense, theoretical, since NASA has so far failed to provide a mission that can observe a single burst from more than one direction. But the radio and optical afterglows do probe an interval when relativistic ejecta are slowing down and coming to emit more or less isotropically (Rhoads 1999; Hurley et al. 2000; Huang et al. 2000; Dal Fiume et al. 2000). The early X‐ray emission is also beamed (Greiner et al. 2000a; Kuulkers et al. 2000), the later X‐ray emission perhaps isotropic and even a flickering standard candle (Kumar 2000). As usual the X‐ray cone has an opening angle roughly equal to Γ⁻¹, where Γ is the bulk relativity parameter of the outgoing stuff.

One should not expect the correction factor from apparent total energy for an isotropic emitter to the real world to be the same for all events, since it will depend both on just what the ejecta are doing (Levinson & Eichler 2000) and on just where we are relative to the axis of the emission cone. Frail et al. (2000a) offer us only a factor of 10 for GBB 970508, while Harrison et al. (1999) are more generous, proffering a factor 300 from the optical and radio declines of GBB 990510. Kumar & Piran (2000b) and Beloborodov et al. (2000) suggest that one could get an independent estimate of the beaming angle from early, rapid fluctuations of gamma‐ray flux. Seeing this will require the collecting area and time resolution of HETE II.

Why are the ejecta and the emission collimated, relativistic, and beamed? Well, apparently just about all cosmic ejecta are at least collimated (§ 6.2). The GRBs, however, get a head start, if, as is now widely advocated, the central engine is (at least for the ones with measured redshifts) a rapidly rotating, highly magnetized black hole (Macfadyen & Woosley 1999; Ruffini et al. 1999; Aloy et al. 2000; van Putten 2000; Lee et al. 2000b), even if you get there by a fairly complex path (Vietri & Stella 1999).

Well, mostly they are gamma rays, and mostly they are made by charged particles wiggling. Now all we have to figure out is (a) how do they keep from being degraded into lower forms of existence? (b) what are the charged particles, and (c) why are they wiggling so fast for such a short time? The phrases "baryon‐poor" (confusingly transmogrified into "baryon‐pure" in Phys. Rev. Lett., 85, 2669), "relativistic electrons," and "compact" have been part of the response to a, b, and c for some time. A new phrase that has risen to prominence this year is "internal shocks," meaning bits of jets that are collectively moving out with sizable bulk Γ's also catching up to each other at large relative velocities.

Keeping baryons away from the emitting jets (lest they turn the gamma ray prince into X‐ray frogs) remains important. Lee & Kluzniak (1999) and Janka et al. (1999) tackled the neutron star plus black hole case, and Rosswog et al. (2000) the binary NS case. Fuller et al. (2000) note that a ratio of neutrons to protons ≳10³ greatly reduces the baryon loading problem, at least for the first 11 minutes. Meszaros & Rees (2000) go on to associate various degrees of baryon loading with the sorts of photons that will get to you, from photospheric (least loading) to mostly synchrotron to subrelativisticallly Comptonized (most loading).

As for what you wiggle, relativistic electrons got most of the votes, though Tokuhisa & Kajino (1999) preferred meson synchrotron emission in a field of 10¹⁶–10¹⁷ G. As for how the wiggling is done, Smolsky & Usov (2000) revived what is called synchrocompton radiation (Rees 1971). This is what electrons do when they find themselves in such very low frequency electromagnetic radiation that the electric and magnetic vectors look to them like static fields (the former to accelerate, the latter to make circular orbits). With wind gammas of 100 or more driving the waves, you get first‐round photons up to 10 MeV and a second round of inverse Compton scattering up to 10¹³ eV.

The words synchrotron and Compton were mentioned in various permutations, mostly and favorably by Granot et al. (2000), Dermer et al. (2000), and Brainerd (2000), with, however, a firm negative vote from Ghisellini (2000) who concludes that the spectrum comes out wrong and proposes that Comptonization by quasi‐thermal electrons might work better. A very high therm would seem to be required.

Many of the events show millisecond pulse substructure, and this rapid variability implies compact emission regions, and perhaps multiple ones. Several variants appeared, including a shotgun (Heinz & Begelman 1999), multiple subjets (Nakamura 2000), multiple shocks and shells (Kumar & Piran 2000a), and very compact emission regions each with a small range of particle γ's (Walker et al. 2000a).

There was a handful (five papers, literally) of discussion of internal shocks in jets and what you should see as a result (GRBs of course, or this paragraph would be in a different section). In order of appearance, they were Meszaros & Rees (2000), Spada et al. (2000), Daigne & Mochkovitch (2000), Beloborodov (2000), and Ramirez Ruiz & Fenimore (2000).

Lower‐energy photons are the province of the tails or afterglows now associated with a number (but still a small fraction) of burst events. Such tails were actually predicted from a fireball model (meaning lots and lots of electron‐positron pairs and photons in a small but expanding volume, Meszaros & Rees 1997). This remains the model to beat, according to Frail et al. (2000b), who endeavored to fit the radio as well as optical and X‐ray fading of GRB 991216. The scenario called PEM (Pair ElectroMagnetic Pulse) is apparently different enough that the authors (Ruffini et al. 2000) did not find it necessary to cite much of the fireball literature.

Only the brightest events yield gamma‐ray spectra with many wavelength resolution elements. Preece et al. (2000) show a set from BATSE. Pre‐BATSE evidence for cyclotron resonances in GRB spectra remains one of the piles of dirt under the rug, because the required magnetic field of 10¹² G or so would imply emission near the surface of neutron stars, the sort of model generally accepted before it became clear just how isotropic on the sky the GRBs are (Ap97, § 11.2; Ap92, § 6.5). Barat et al. (2000) have gathered all available photons from the very bright GRB 910406 and suggest that it had an umpteen‐component spectrum, so that drawing an envelope over them all might give the impression of 10–50 keV features. This would be a happy resolution, in which everybody gets prizes, either for accurate reporting of observed spectra or for rethinking what they must mean.

And, as the Sun pulls away from shore and our boat sinks slowly in the west, you are left with another 40‐some interesting GRB papers unmentioned, but a review article (Piran 1999) that catches many important issues omitted here.

10^50–100, the potential, finite, future of life in the universe, according to Krauss & Starkman (2000). The units probably don't matter.

8 × 10¹⁰ photometric measurements made by the MACHO project (Alcock et al. 1999).

5 × 10¹⁰, the galactic millennium in years. Some interesting things will happen in the meantime (Hodge 2000).

5.25 × 10⁸ lines of TiO and H₂O included in a model atmosphere for pre‐main‐sequence stars (Allard et al. 2000).

17,000,000 infrared point sources in the first release of data from DENIS (Epchtein et al. 1999).

105,924 sources in the ROSAT All‐Sky Survey (Voges et al. 2000). Each is represented by at least six photons, and the catalog will not be issued in print form.

83,992 visual binary systems in the Washington Double Star Catalogue. All are exactly where they should be, but we misplaced the reference.

60,000 OB stars in the Milky Way (Reed 2000), for which we have reasonable data on about 3000. The implied formation rate is about 1/40 years, not enormously different from estimated rates of core‐collapse supernovae.

22,168 regular orbits included in a dynamical model of the bulge of the Milky Way (Haefner et al. 2000).

18,811 sources from the ROSAT All‐Sky Survey in the archival literature (Voges et al. 1999). This is less than 20% of the total, as per the old high energy astrophysics conference protest slogan, "free the ROSAT 100,000."

13,875 X‐ray stars from the cross correlation of the Tycho (Hipparcos) and ROSAT catalogues (Guillout et al. 1999).

11,310 candidate galaxies from a visual search of 560 square degrees in the galactic zone of avoidance. Only 152 of them were previously catalogued (Roman et al. 2000).

9286 stars with polarization data (Heiles 2000, who prints only sample tables).

3000 languages that have become extinct since 1900 (Krauss 1992). About 3000 remain.

2355 galaxies in Abell 496, for which Moretti et al. (1999) report luminosities.

2049 Cepheids in the Small Magellanic Cloud, with light curves from the OGLE project (Udalski et al. 1999, who show all the light curves!).

1952, the Carrington rotation number in September 1999, which was Space Weather Month (Kozyra & Webb 1999). Carrington must have started counting before 1875 (because that is when he died). In fact he and Hodgson were the first to report seeing a white‐light solar flare, in September 1859.

1781 periodic variables found so far by the Robotic Optical Transient Search Experiment (Akerlof et al. 2000). One of the less burning question is whether the acronym ROTSE is pronounced like ROTC. A more significant point is that, since their primary goal is to find non‐periodic variables (transients), these are in a sense all noise.

1302 young stellar objects with IRAS colors (Iwata et al. 1999).

1049 sources (plus one CLEAN artefact) in a survey down to 0.1 Jy at 408 MHz (Vigotti et al. 1999). We remember when 3C went all the way down to about 9 Jy (but it looked at a larger fraction of the sky).

758 ROSAT sources in the Large Magellanic Cloud (Haberl & Pietsch 1999).

719 star clusters in the Small Magellanic Cloud, with 46 added by the OGLE project (Bica & Dutra 2000) and perhaps 40 more to be found, compared to 900 predicted by Hodge (1986).

745 star clusters in the LMC, with 126 added by OGLE (Pietrzynski et al. 1999).

674 X‐ray sources in a statistically complete subsample (yeah, the ROSAT ASS again). They include 274 stars, 26 galaxies, 284 AGNs, 78 clusters of galaxies, and 12 empty fields at visible wavelength (Krautter et al. 1999).

636 Delta Scuti stars, about half newly‐recognized since 1994 (Rodriguez et al. 2000).

476 catalogued LINERs seen by FIRST (Carrillo et al. 1999). Most are bright, early‐type galaxies.

475 Herbig‐Haro objects, including 15 new ones near compact reflection nebulae (Aspin & Reipurth 2000).

463 flare stars with ultraviolet data (Gershberg et al. 1999).

380 variable‐star light curves obtained with a 13.5 cm telescope, by a single observer, looking at a small part of the sky (Pojmanski 2000).

280 Cepheids with light curves from Hipparcos (Groenewegen 1999). About 10% are W Vir (Pop II) stars, and 216 are Population I, pulsating in the fundamental mode.

271 diffuse absorption bands produced by the interstellar medium, with wavelengths between 4460 and 8800 Å (Galazutkinov et al. 2000). Most of the central wavelengths are said to be accurate to 0.1 Å, leaving us in doubt about just how diffuse they can actually be (not anyhow more than 16 Å on average!).

226 confirmed diffuse interstellar bands over a somewhat different wavelength band (Tuairisg et al. 2000), or 213 (Weselak et al. 2000). All three papers make the point that the inventories are open‐ended, with new bands being seen and previously reported ones not always confirmed. The first lists came from Hegen (1922) and Merrill (1934), who showed that the origins were truly interstellar.

212 extragalactic radio sources imaged in total and linearly polarized intensity (Reid et al. 1999).

156 nearby eclipsing binaries, whose light and radial velocity curves may be clean enough for direct distant determinations (Kruszewski & Semeniuk 1999). The method goes back at least to Joel Stebbins' work with a selenium photometer on Delta Orionis in 1915, and the eventual goal is to reach extragalactic systems.

134 candidate planetary nebulae in M33 (Magrini et al. 2000).

118 supernovae discovered in 2000, up to the witching data of 30 September (2000dn in IAU Circ. 7494, dated 29 September). The rest of the calendar year will clearly use up the 2000e's and probably the 2000f's. 1999 nearly ran off the g's (SN 1999gu in IAU Circ. 7346), and one can see dimly ahead to a time when the IAU may need to reconstitute its Supernova Working Group to decide upon a method of naming events beyond, say, 2008zz. The first SNWG was Zwicky's creation and died with him. The most useful product of the second, which declared itself supernumerary after SN 1987A made the word supernova conscious, was probably the resolution declaring that Zwicky's use of double lower case letters after the first 26 events in a year should be official. Our favorite was 2000cp, since the composition (unknown to us) was surely at least chemically pure.

106 giant molecular clouds in a more or less complete census of the LMC (Fukui et al. 1999). The paper is part of a "first results" package from NANDEN, the Japanese millimeter (the wavelength of operation, not the size) telescope at Las Campanas. The H₂ mass adds up to about 10% of the H i mass of the LMC.

70 authors on a paper reporting optical detection of a gamma‐ray burst (990328; Schaefer et al. 1999). The telescope is in Venezuela, which may or may not have anything to do with it. Anyone who is puzzled by the ordering of the authors' names has probably never met the first author.

70 sinusoidal light curves for variable stars in the Tycho Catalogue (Koen & Schumann 1999). The telescope probably passed over Venezuela.

39 primeval galaxies, meaning ones with infrared luminosity in excess of 10¹³ L_⊙ and star formation rates in excess of 10³ M_⊙/yr. Most are, however, at redshifts of only 0.3–1.0 (Rowan‐Robinson 2000).

34 modes seen in a single Delta Scuti star, U CVn (Breger et al. 1999). Most have frequencies in the range 52–112 μHz, and about half are linear combinations of the other half.

30 pulsars known to have glitched (Urama & Okeke 1999). Glitch activity is approximately logarithmic in dP/dt, the time derivative of the period.

26 broad photometric bands? Well, perhaps not, but Steele & Howells (2000) have reached Z without revealing the secret of its wavelength. I–Z is, anyhow, not a good temperature indicator for stars of type L.

23 carbon atoms in the longest chain molecule expected in IRC +10216 according to Millar et al. (2000), whose reaction network used 3851 reactions among 407 species.

20 pulsars in the globular cluster 47 Tuc, reported in the Newsletter of the Australia Telescope.

20←19 = n, the highest Rydberg line seen in the Sun (Clark et al. 2000b). It has a wave number of 29.6 cm^-1 and is limb brightened.

19 X‐ray emitting supernova remnants in the SMC (Haberl et al. 2000).

17 moons belonging to Jupiter. This is merely a progress report, with additional satellites being reported for all the giant planets on a near‐weekly schedule. For instance, IAU Circ. contains the names for the 18th, 19th, and 20th Uranian moons. None of them is much good for spooning under, and Uranus isn't real strong on June either.

13 Gamma Dor stars (Kaye et al. 1999). We knew the first (Ap97, § 8.3; Ap98, § 3.8) back when it was still called 9 Aurigae.

12 different codes from which the development of very large scale structure in the universe can be computed. They are compared by Frenk et al. (1999) and found to be in reasonable agreement.

11 CH₃OH masers (Val'tts et al. 1999).

11 classes of stars with evidence for magnetic activity (Jetsu et al. 1999).

10 or 11 X‐ray dippers (Smale & Wachter 1999). These are low mass systems in which a flared accretion disk occults its own hot center once per orbit period, of importance in confirming that these sources really are binaries.

10 supernovae seen as X‐ray sources (Schlegel 1999).

Nine sites from which TeV gamma‐ray astronomy is being done (Catanese and Weekes 1999). Actually there is a 10th, at Yangajing, Tibet (Amenomori et al. 1999), but the paper still fits here because it has a total of nine authors who are either Tibetan, surnamed Zhang, or both. We won't tell you how to recognize a Tibetan author, but it once caused the editor of another journal to ask a lead author whether he was absolutely certain about the names of his colleagues. In either case, the installations outnumber the confirmed sources: three or four supernova remnants and a comparable number of low redshift blazars.

• Eight  
  — ROSAT candidates for Be‐star X‐ray binaries in the LMC (Stevens et al. 1999).
•     
  — low mass X‐ray binary systems in which the frequencies of the quasi‐periodic oscillations trace out a Z‐shape in suitable coordinates that involve X‐ray "color" or hardness ratio (Smale & Kuulkers 2000).
• Seven  
  — the number of mathematics problems for the solutions of which the Clay Mathematics Institute has offered $1,000,000 each (Jaffe 2000).
•     
  — pulsars probably detected by EGRET as gamma‐ray sources (Kuiper et al. 2000). The most recent is J0218+4237 with a period of about 2.5 msec.
• Six  
  — AM CVn stars. These are the odd cataclysmic variables whose prototype is HZ 29 and which apparently consist of two low mass, helium white dwarfs in very tight orbits. Skillman et al. (1999) and El‐Khoury & Wickramasinghe (2000) report that AM CVn itself has both an orbit and a superhump period near 1000 sec.
•     
  — stars of spectral type WO, presumably the last evolutionary stage in the life of a massive star en route from the main sequence via other forms of Wolf‐Rayet star to a bare CO core (Afanasev et al. 2000).
•     
  — low mass X‐ray binaries that show X‐ray eclipses (Wachter et al. 2000).
• Five  
  — medal winners on the US team at the International Physics Olympiad held in Leicester UK (Hargreaves 2000). All were silver or bronze medals, only the Chinese team receiving five golds. Other high scorers were Russia, Hungary, India, and Iran (including the woman with the highest score). The US team members were all men, including two from Southern California and two from the Washington suburban area (for which the author who sporadically occupies both these sites can take absolutely no credit).
•     
  — roAp stars (meaning "rapidly oscillating") which also have Ap companions (Gelbmann et al. 2000).
•     
  — measured orbit periods for low mass X‐ray binaries in globular clusters (in't Zand et al. 2000).
•     
  — types of H i shells in the interstellar medium of the LMC (Kim et al. 1999). Shells and supershells occupy most of the ISM there.
•     
  — lower limit to the number of sorts of instabilities in stars with both magnetic field and differential rotation (Spruit 1999).
•     
  — blue‐shifted, iron‐rich absorption line systems in the spectrum of a single QSO (Dobrzycki et al. 1999). The clouds presumably all result from a single explosive event.

Among the notable fours are:

Binary pulsar J1811−1536, at P = 0.104 sec and 18.8 days, probably the fourth with a second neutron star as its companion (Lyne et al. 2000a).

The heart chambers of a small ornithiscean dinosaur, whose name we do not remember. They don't answer anyhow (Fisher et al. 2000).

The dwarf novae with deep eclipses and period longer than the gap in N(P) (Gaensicke et al. 2000b).

P4M, where the fourth P means "parabolic" (Brieu & Evrard 2000). What about the other three P's? They are all particles; the acronymial method of studying evolution of complex dynamical systems, P3M, means particle‐particle/particle‐mesh.

A bunch of quadruple star systems, including LHX 1070, all of whose members are M and L dwarfs, less than 20 pc from us (Leinert et al. 2000); HR 266, all of whose members are B9‐A1 dwarfs, in a hierarchical pattern with periods of 83 yr, 4.8 yr, and 4 days (Balenga et al. 1999); and HD 98800, in which two spectroscopic binaries make up a visual binary 38 AU apart, and only one of the pairs is surrounded by a dust disk (Koerner et al. 2000). Anything that isn't forbidden is compulsory as Gell‐Mann is supposed to have said.

There is also, perhaps, a 4‐armed spiral to be found in the distribution of OH stars in our own galaxy (Sevenster 1999), though Drimmel (2000) avers that the real distribution of mass has only two arms (even if the gas shows four). He is also looking at infrared data and the main body of the disk, suggesting a real contradiction. The average would, of course, be an m = 3 spiral, which is at least approximately what the galactic nuclear disk looks like. Vollmer & Duschl (2000) find that most of the motions therein are Keplerian and turbulent, not outflowing. We cannot, in any case, complete with NGC 1228 whose two arms split into three each outside the bar, according to Fuchs & Moellenhoff (1999).

3.5 are the post‐Newtonian equations of Rezzolla et al. (1999). We are in no position to object to other people doing things by halves, but are still struggling over the adjective. Is it the 3‐and‐one‐halfth, the three‐point fifth, the third‐and‐one‐half approximation or what?

The source J005734.78−272822.4 (Page et al. 2000) presents similar problems of pronunciation, but equally worrying is the rapidity with which reality has caught up with what was meant to be a parody (Trimble & Sciatti III 2000).

The threes, some of which clarify astrophysical classes and some of which confuse include:

The third sdOB star in a visual binary (Makarov & Fabricus 1999). The orbit periods are all of order 100 years, so it will take a while for measured masses to settle the old issue of just where these stars live in the evolutionary scheme of things.

The third optical burst from an X‐ray transient binary (Gneiding et al. 1999, not all from the same source).

Three meanings for the acronym AIC. Accretion‐induced collapse came first (Ap91, § 6.1). And then there was the Akaike'i Information Center (Ap&SS, 271, 213) and Achromatic Interfero‐Coronagraph (A&AS 145,, 141).

The third case of transient interstellar absorption (Price et al. 2000). This one is caused by H i in the Orion‐Eridanus supershell. Some others are due to other parts of Orion and to the Vela SNR, which was the first reported (Hobbs et al. 1982).

Three components in the emission lines of Seyfert 2 NGC 1069 (Crenshaw & Kraemer 2000).

Ni⁴⁸, the third doubly‐magic nucleus of nickel (Blank et al. 2000). The others are 56 (which you know about from supernovae) and 78. Only nickel can make this claim.

The third detection of Mkn 501 at TeV energies (Andreeva et al. 2000), apparently a delayed report, given the nine installations mentioned above.

The third‐largest Kuiper Belt Object, 1996TO₆₆ at about 326 km (Hainaut et al. 2000). Bigger are Pluto (oh, you knew that?) and Charon. The rotation period is 6.25 hours and other variability may be due to comet‐like activity.

Three basic X‐ray states of the microquasar GRS 1915+105 (Belloni et al. 2000a).

And the three protons of the stable if not very familiar ion H⁺₃ (Encrenaz et al. 2000; van der Tak & van Dishoeck 2000).

Two brings us to:

The second medal awarded to Visnjan Observatory for discovery of a comet by amateur astronomers.

The second binary Cepheid (meaning that both stars belong to the class). It is EV Sct (Kovtyukh & Andrievsky 1999). The first was CE Cas. Both are in open clusters, NGC 6664 and 7790, which given the rarity of Cepheids in clusters strikes us as odd.

The second m = 1 spiral in a disk around a binary Wolf‐Rayet star, WR 98a (Monnier et al. 1999).

The second detection of polarization in a non‐mased radio emission line, CO in an IRAS source (Girart et al. 1999). The first was reported nearly 20 years ago (Goldreich & Kylafis 1981).

The second compact blue dwarf galaxy (Tol 65) with a companion of low surface brightness (Papaderos et al. 1999). The first was I Zw 18 (famous for its minimal metallicity). More examples, the authors note, are needed, in order to test their suggestion that the pairs in each case formed from a single H i concentration.

The second closest star of the future will be less than 0.5 pc away within 20 Myr, according to extrapolations of Hipparcos measurements (Garcia‐Sanchez 2000). The closest will still be at 1 AU.

The second interstellar molecule to be found in double deuterated form is ND₂H in L134N (Roueff 2000). The first was D₂CO.

Two acceleration mechanisms for the gas stream in QQ Vul, a magnetic cataclysmic variable or AM Her star (Schwope et al. 2000).

The Second Tycho Catalogue (Høg et al. 2000). It includes positions, proper motions, and two‐color photometry for the 2.5 million brightest stars in the sky (and so could have been listed as 2.5 × 10⁶ above).

Second example of gravitational lensing by a small group of galaxies near z = 1 (Rusin et al. 2000). The first was reported in 1999 and in both cases the group members are all low mass galaxies. Additional examples could reveal whether the dark to luminous mass ratio is the same in these groups as in richer clusters nearer the present epoch.

The second basaltic asteroid, 1459 Magnya (Lazzaro et al. 2000). The first was 4 Vesta and they are thought to be fragments of the same parent body.

The second faintest white dwarf at M_V = +16.8 (Hambly et al. 1999). The authors suggest that it may be a halo star (i.e., MACHO candidate). There are, of course, probably very many still fainter white dwarfs but, like the trans‐Lehrerian elements, they haven't been discarvard (rhymes with Harvard).

This section of extrema is appropriately named for George Washington, with politically correct improvements, because so many of the discoveries were financed by federal grants. We caught about 25 firsts and three or four times as many other extrema and lay out for you a subset chosen for potential astrophysical significance. This lot is ordered from near to far. Mentally supply "the first" before each item.

Drawing of a sunspot dates from 1128. The artist, John of Worcester, presumably a firm believer in the immutability of the heavens, is said to have thought he was seeing opaque objects passing by (Stephenson & Willis 1999). In case you wondered, transits of Mercury are not naked eye events, and there were none of Venus that decade.

Stable compound incorporating argon (Kriachichev et al. 2000). It is HArF, made in Finland. Only helium and neon still resist chemical bonding, but similar methods may work for them (and presumably also the low temperature).

Spectroscopic orbit for a pair of brown dwarfs. Both show lithium, and the mass ratio is 0.85 (Basri & Martin 1999).

Triple white dwarf (Maxted et al. 2000). The close pair is a spectroscopic binary, with one helium and one carbon‐oxygen component, also a first; other known WD pairs have mass ratios close to one and the same core composition for both stars.

Binary white dwarf consisting of two polars, TX J1914+24 (Ramsay et al. 2000).

Balmer lines due to deuterium (Hebrard et al. 2000). Seen in Orion, they are fluoresced by Lyman alpha, which is good for detection but very bad for measuring the D/H ratio.

Planetary nebula distance measured using optical proper motions (Plus radial velocities, Reed et al. 1999). The radio proper motion technique was pioneered by Masson (1989), and the first optical proper motions measured by William Liller, who was not cited, in the late 1960's.

Planetary nebula made by both stars of a binary dying in close succession, KjPn8 (Lopez et al. 2000).

X‐ray lines of Cr and Mn from a source outside the solar system (Hwang et al. 2000a) in the supernova remnant W49B.

Luminous blue variable displaying a coherent period, 8.26 days for B416 in M33 (Shemmer et al. 2000). No cause was suggested.

Chimney that has blown out of the plane of the Milky Way on both sides of the plane (McClure‐Griffiths 2000). Our notes said "to north and south," but perpendicular to the galactic plane is north and south only in the sense that the San Diego Freeway can be said to run north and south through Orange County.

EGRET detection of a microquasar (LS 5039, this is not actually one of the superluminal ones; Parades et al. 2000).

Star of type Onfp(Ocf). No, we aren't quite sure what it is, but it must be quite bright, since it is in the SMC (Walborn et al. 2000).

Star cluster in IC 1613 (Wyder et al. 2000). Baade (1963) had looked for, but did not find, clusters in this dwarf irregular galaxy.

Superluminal Seyfert, III Zw 2 (Brunthaler et al. 2000).

These are, on the whole, ordered from far to near, and the 'est words are supplied, since it is not always self‐evident just what is extreme. The quote is from Alexander Pope, who also said of governments, "What e'er is best administer'd is best." We are not sure what fraction of his proposals were funded, but it's clear he understood astronomers, to wit, "To observations which ourselves we make, We grow more partial for th'observer's sake."

Record redshifts. z = 5.5 for the QSO RD 300 (Stern et al. 2000b), based, however, only on R‐band drop out. At M_B = -22.7 (H = 100) it would be the faintest high redshift QSO. z = 5.03 for a QSO in the Sloan Digital Sky Survey (Fan et al. 2000a), also reporting z = 4.92 as a probable record for a QSO with broad absorption lines. The most distant BL Lac object (Fan et al. 1999, again from SDSS) has a damped Lyman alpha system which puts the source at z≥4.62. And the most distant BAL QSO with detected X‐ray emission is also the optically brightest (Brandt et al. 1999).

Some other QSO and AGN items. The largest QSO nebula (meaning ionized from the nucleus) extends to 200 pc (Shopbell et al. 1999). The gas is in smooth rotation and must belong largely to a host galaxy, not a merger or cooling flow. The closest pair of QSOs (meaning to each other, the redshift is 2.24) is CTQ 839, with a projected separation of 21 or 8 h^-1 kpc, and a velocity difference of less than 100 km/sec (Morgan et al. 2000). They are not a good candidate for lensing, however, since the spectra are very different. They must have rather small host galaxies and be at risk of collision or merger, if they haven't already begun to interact and share a host.

The brightest narrow‐lined Seyfert 1 galaxy is Ark 564 (Vaughan et al. 1999, who leave us wondering who was Ark?). M81 is the closest LINER (Pellegrini et al. 2000).

The HST Key Project on the cosmic distance scale has truly reported their last galaxy and finally got I before K in the alphabetized part of the author list (Sakai et al. 1999).

Some other galaxy extrema. The most metal poor remains I Zw 18, though the search for competitors continues (Kniazev et al. 2000). The limit occurs close to [Fe/H] = -2 because star formation is very inefficient at smaller metallicities and one O V star can ionize a whole cloud of 10²–10⁴ M_⊙ (Nishi & Tashiro 2000). The most massive spiral galaxies top off at 2 × 10¹² M_⊙ according to Salucci & Burkert (2000), who say it is related to a breakdown of scaling of core density with core radius at the largest disk masses. The closest weak lensing occurs in Abell 3667 (z = 0.055; Joffre et al. 2000). One gets, of course, a measurement of the cluster mass. It is not unusual.

The reddest galaxies. These are conceivably important because they might be displaying unusually large amounts of dust, star formation that ended a really long time ago, very high metallicity, or, of course, measurement errors. The answer appears to be "all of the above," although not all authors have in mind quite the same definition of very red. Cimatti et al. (1999) present a sample where some are dusty and some are old. A SCUBA sample with I−K≥6–7 in the Hubble Deep Field includes galaxies not actually seen by Hubble, a major impediment to assigning types (Smail et al. 1999b), but others with a bit more information include both AGNs and star formers, both, of course, with lots of dust (Cooray 1999). Soifer et al.'s (1999) sample is defined by R−K>5, and the underlying galaxies are not all the same sort of beast, even after corrections for reddening. Finally, Margoniner & de Carvalho (2000) report that a previously‐advertised population of very red galaxies in rich clusters at moderate redshift belongs to the "oops" category.

Star clusters. The puniest NGC one is 6994 with three bright members at its core (Bassino et al. 2000). Actually a few NGC numbers correspond to asterisms with no physical members at all, or, as a colleague once said, "Either your phone number hasn't changed, or NGC 6948 is an interesting object." It is, in fact, one of those asterisms.

Neutron stars. The largest mass this year, 7.944 M_⊙ (Negi & Durgapal 2000) is, not surprisingly, a calculation not an observation. It requires the star to have a speed of sound equal to the speed of light in its core, a polytropic envelope (carrying the new, higher postage rate), and a radius of 33 km. The shortest rotation period for a neutron star, as opposed to a strange quark star, is also a calculation (Glendenning 2000). The most interesting bit is that the Keplerian period is not unique for rotating neutron stars because frame dragging depends on latitude.

X‐ray binaries. The longest orbit period for a low mass one is 304 days for GX 1+4 (Pereira et al. 1999), besting the previous record by a factor 10. The shortest orbit period for a LMXRB is 80 minutes (King 2000). This reflects a turn‐around in the evolution of the period analogous to the one in cataclysmic variables. The fastest quasi‐periodic oscillations are at 1200 and 1800 Hz in Cen X‐3, a high mass XRB (Jernigan et al. 2000). The longest X‐ray binary burst fueled by nuclear reactions was 86 minutes for 4U 1735−44, seen by BeppoSAX (Cornelisse et al. 2000). There were no other bursts for several days before or after, in accordance with the long‐honored models with characteristics of a relaxation oscillator (the sort of toilet you have to fill up completely before you can flush it again, even if it wasn't totally emptied the first time).

White dwarfs and cataclysmic variables. The shortest rotation period for a white dwarf is 33 sec in the CV AE Aqr (Choi et al. 1999). VY Scl is a CV and part of a triple system with P = 5.8 days, probably the shortest "long" period known in a triple (Martinez‐Pais et al. 2000). The oldest recovered nova is 1678 (Robertson et al. 2000). Non‐recoveries of others like 1612 means that they must be even fainter. The most stable stellar optical clock is the pulsating WD G117‐1515A, with P = 215 s and dP/dt<3 × 10^-15 sec/sec. Once its proper motion and such are sorted out, this should turn into a measured value of periodic change, to be compared with what is expected as WDs cool (Kepler et al. 2000). Some radio pulsars have even greater period stability.

Other close binary stars. The shortest period Algol is W Crv at P = 0.388 days. It must have started with a mass ratio close to one and never have had a rapid mass transfer phase (Rucinski & Lu 2000). The W UMa whose mass ratio comes closest to 1, at 0.970 ± 0.003 is V753 Mon (Rucinski et al. 2000). The total mass of the system is also fairly large, and it must be trying to tell us something about what is needed to maintain contact systems in contact.

Star formation and young stellar objects. The closest molecular cloud with recent star formation is MBM 12 at 65 ± 35 pc (Hearty et al. 2000a) or 60–90 pc (Hearty et al. 2000b). At least the authors agree with themselves about the name of the region (which tells you that it is a high latitude molecular cloud if you happen to have your Long Ranger code ring handy). The cluster is the second closest after that centered on TW Hya. The smallest measured velocity is 4–10 m/sec for the infall of CS gas in a dense core in a star formation region (Heithausen 1999). To put this in context, 7 m/sec is the still slightly superhuman 25 km/hr, that is, you could not keep it up for long, but a cheetah could. The fastest Herbig‐Haro object is in a region called 66D37 (even by its closest friends; Raines et al. 2000) and displays a proper motion equivalent to 850 ± 200 km/sec in infrared images.

The most massive stars. It remains puzzling that the largest numbers found directly in binary systems, e.g., 45 + 20 M_⊙ for the 03 V + 08 V pair HD 93205 (Antokhina et al. 2000) remain considerably smaller than the numbers extending above 100 M_⊙ found by matching observations of luminosity and temperature to evolutionary tracks. It is presumably somehow relevant that measured values of surface gravity give something in between, e.g., 37–64 M_⊙ for stars with evolutionary masses of 30–140 M_⊙ (Herrero et al. 2000), and the cause must be rather early mass loss, though the IRAS source discussed by Shepherd et al. (2000), which has already dumped 50–60 M_⊙ in 10⁴–10⁵ yr, seems excessive. Or perhaps not, if you look at the largest mass loss rates, up to 10^-3 M_⊙/yr for a few asymptotic giant stars (van Loon et al. 1999) and 10^-2 M_⊙/yr for a protostar in Orion (Nakano et al. 2000). Both must be rather short‐lived phases.

The fastest planetary nebula gas clocks in at 2500 km/sec (IUE spectra reported by Feibelman 2000), beating the 600 km/sec for knots in MyCn18 reported by O'Connor et al. (2000). But both are cheating. The latter is probably ejecta from a nova‐type explosion of the central binary, and the former is probably a hot stellar wind from the central pre‐white‐dwarf, not the material from an atmosphere of a red giant that we normally mean by the phrase "planetary nebula."

The longest comet tail, streaming 3.8 AU out of the nucleus of Hyakutake, was caught by Ulysses in 1996 (Jones et al. 2000b). The previous record of 2 AU was set in March 1843 by, we presume, the Great Comet of 1843. Hyakutake's tail also displayed conspicuous pick‐up ionization from the solar wind at about the same time (Gloeckler et al. 2000).

The largest number of references in a single parentheses is 103 (Li & Wilson 1999), beginning with Pringle & Rees (1972). The paper concerns the structure of accretion disks with magnetic fields.

The largest numerical error of the year probably came from Wilman & Fabian (1999) who attribute some part of the X‐ray background to Seyfert 2 galaxies absorbed by 10^-24≤N_H≤10⁺²⁵ atoms/cm². We think 10⁺²⁴ may have been intended, owing to the difficulty of detecting X‐ray absorption by very small column densities.

The highest observatory is Cerro Toco, a Chilean site from which anisotropies in the cosmic microwave background are being measured (Miller et al. 1999). It is at 5200 meters, and ALMA will go nearby.

The smallest limit on the charge of the photon may be 10^-29 e set by Sivaram (1999) from observations of GRB 990123. Their limit on the photon mass is 10^-44 g and is not a record.

The smallest observer is the one required to verify that the uncertainty principle still applies to energy, angular momentum, and so forth measured from inside the system (Aharonov & Reznik 2000).

If you dislike getting up in the dark as much as we do, you may always have been too bleary eyed to wonder whether the latest sunrise occurs close to the winter solstice or immediately after the onset of daylight savings time. This turns out to depend a little on just which day DST starts, a good deal on the equation of time, and most on your latitude. Post DST is later (by a few to 29 min) at northern latitudes south of 44°. The solstice period is later, by a few to 40‐some minutes at latitude 44° to 60°, beyond which you get tangled up with sunrises and sets that happen once a year rather than once a day. The calculation was generously done for us by Brian Marsden, whose 35 years as director of the IAU central telegram bureau also set a record.

The largest geoid, 8 × 17 meters, was found in Spain (Garcia‐Guinea 2000). If you think of geoids primarily as ornaments for coffee tables, then this one was clearly meant for the Immensosaurus family (and a collection of the "Astrophysics in ..." reviews for their coffee table book).

You have had a few years to get used to the idea that data for (many) galaxies reveal a strong linear correlation between the mass of the central black hole and the mass of stars in the spheroidal components (Ap98, § 9.1), with the mass of the black hole ≈1–2 × 10^-4 times the mass of the spheroid. The time has come to ask (1) is the relation generally true? (2) if true, how much real scatter is there around it? and (3) how many different ways has the correlation been explained?

First, is it true? Not apparently for the later spirals (Salucci et al. 2000). And this may be because their bulges are not the same sort of beast as larger ones. They show different correlations of stellar population, size, and surface brightness (Khosroshahi et al. 2000; Molla et al. 2000), and there has been a modest ground swell (ground bulge?) of opinion for making them from repeated dissipation of central bars rather than from collapse or mergers (Carollo 1999; Zhang & Wyse 2000). Should we keep the Milky Way on the graph? It is neither a very early nor a very late spiral (de Vaucouleurs et al. 1990), and it does quite unambiguously have a central black hole of 2.6–3.3 × 10⁶ M_⊙ (Genzel et al. 2000; Ghez et al. 2000).

A good indicator of reality for the correlation for the larger bulges is that it gets tighter with better data. Earlier papers focused on bulge luminosity as a mass indicator. Both Ferrarese & Merritt (2000) and Gebhardt et al. (2000) look instead at the velocity dispersion of the bulge stars (which should probe mass with less sensitivty to details of stellar population than does luminosity) and find that the scatter is reduced. Unfortunately, the two groups find two different functional forms, with M(BH)∝σ^x_v and x = 4.8 ± 0.5 and 3.75 ± 0.3 respectively.

Despite these disagreeing by more than their probable errors, both papers conclude that the answer to question (2) is that the scatter is largely observational. There are exceptions. Shields et al. (2000) point to the LINER NGC 4203 as a real ratio‐disturber, having an upper limit to the black hole mass of 6 × 10⁶ M_⊙, at most one‐fifth of what you expect.

Third, we found something like six or seven explanations for the correlation during the index year. Most are not readily classifiable into our earlier types of chicken (bulge came first), egg (black hole came first), and potato salad (co‐evolution). Whatever order they are presented in will offend some author (above and beyond the offense of being mentioned in these pages at all). The one adopted is simply the order in which the papers appeared on library shelves.

Fabian (1999), building on an idea from Silk & Rees (1998), watches both the spheroid and black hole grow until, at the observed ratio, wind ejects cold gas from the galaxy and stops the growth of both. One expects the wind energy to heat the intergalactic medium to as much as 10⁷ K; the ejecta may be what we see as broad absorption lines in QSO spectra (arguably all QSOs have such clouds, just not always along our line of sight, Ogle et al. 1999); and the scenario predicts that a new population of black‐hole powered, hard X‐ray sources should appear at z>1. (Cf. § 8.5.1.)

Franceschini et al. (1999) say that accretion onto a central black hole should track the rate of star formation, so that everything is big and bright together, and systems of large mass and luminosity evolve faster than others.

Monaco, Salucci, & Danese (2000) start with the spin of the dark matter halo as setting the efficiency of both black hole growth and star formation. Slow spin yields big bulges and holes, and the ones in the biggest halos turn on first (which more or less agrees with observations).

Kauffmann & Haehnelt (2000) propound a second class of egg salad model, in which the co‐evolution of black hole mass, AGN activity, and galaxy growth is driven by mergers. As time goes on from z ≈ 2 to the present, there are fewer mergers, less gas available, and a longer cooling time for the gas (the cold component being seen as the damped Lyman alpha absorption features in QSO spectra).

Merrifield et al. (2000) propose another merger‐driven scenario, with the key quantity being the length of time since the last major merger in hierarchical galaxy formation. This then determines the age of the dominant stellar population and the final mass of the black hole via the amount of material still in gaseous form and available to be accreted by it.

Ostriker (2000) invokes self‐interacting dark matter (the hot new cold candidate mentioned in § 12.4.2). It leads to both the black hole mass and the luminosity of the galaxy scaling with circular velocity as V^9/2_c, consistent, at least, with what Ferrarese & Merritt (2000) report a couple of paragraphs above and, of course, setting BH mass in a fixed ratio to galaxy luminosity. An initial central seed BH of about 25 M_⊙ is needed and comes presumably from a massive Population III star.

Seven being a lucky number, we feel compelled also to note the work of Liszka et al. (2000) on X‐ray variability of NGC 5538, which seems to require the central engine to be a dense cluster of stellar mass black holes. If these are indeed stellar remnants and you adopt a constant initial mass function, we suppose this will imply a fixed ratio of "engine" mass to mass in old stars, but do not know quite what else to do with it, except wait patiently until all the black holes merge into a single big one (Rees 1978), while singing, "Hi ho, hi ho, it's off to Kerr we go."

We tuck these in here both because of the probable connection with mergers of galaxies (§ 7.2) and because black holes brought together by merging galaxies must themselves eventually come together into a product that can get closer to the extreme Kerr limit on angular momentum, a/m = 1 in suitable units, than can be achieved by accretion spinup (Agol & Krolik 2000). And, in turn, a rapidly rotating black hole is probably a Good Thing for some of the processes by which they may power AGNs. It is the angular‐momentum related part of the mass‐energy of a BH that you can extract. Thus the sequence from binary black hole to rapid spin allows the theorist to tap some of the orbital energy of the merging galaxy pair for a later QSO. Indeed Zhang & Chu (2000) suggest that the only way to turn on a really bright quasar is through merger of a pair of galaxies, each with its own black hole. The rapid rotation is probably useful for jet collimation as well as energy extraction.

Do binary black holes actually turn up in present, past, or future active galaxies? The best known example is the blazar OJ 287 with its 12 year periodic light curve (Valtaoja et al. 2000; Abraham 2000; Katajainen et al. 2000), though doubts have been cast even on it (Kidger 2000). Other, less well known, candidates are AO 0235+164 which is also perhaps periodic (Chen et al. 2000a); 3C 273, where a second BH perhaps drives precession of the jet of the first (Romero et al. 2000); PKS 0429−014, another precessor (Britzen et al. 2000); and NGC 3597, particularly interesting because it also has some young globular clusters, whose presence could be another merger signature (Forbes & Hau 2000). OX 169 has, meanwhile, withdrawn itself from the candidate list. What had seemed to be a double‐peaked emission profile of the H‐alpha line was really a single H‐alpha feature plus the stronger of the two lines of [N ii] which flank it (Halpern & Eracleous 2000).

Clearly most galaxies do not have binary black holes now, though even for the Milky Way this is not as easy to rule out as you think it is going to be (Jaroszynski 2000), but Quillen et al. (2000) suggest that most E and SO galaxies have a black hole binary and merger in their past, based on their shapes and the scattering of star orbits required to match them.

Yes, we are pretty sure it takes one from column A to get one from column B, even if that is not part of your a priori definition of an active nucleus.

The most important thing in a black hole's life (like a jockey's) is its mass. This, according to Magorrian & Tremaine (1999) cannot be less than 10⁶ M_⊙ in any galaxy here and now, because that many stars will have been swallowed in a Hubble time, no matter what the initial BH and star populations were like. The limit will not be easy to test. If the observations tell you, for instance, that a small galaxy in the Local Group does not have a black hole of more than 10⁶ M_⊙, then the most probable value is zero, while actually seeing evidence for smaller ones presents observational challenges.

If you, the black hole, support a serious, accretion‐powered active galaxy for 10⁸ yr, you will inevitably swallow at least 10⁸ M_⊙ of stars, over‐ripe tomatoes, or whatever is available. Thus, points out Mathur (2000), small masses necessarily go with the youth of AGNs, whether they are QSOs at z ≈ 2 or Seyferts now. The paper is unusual in containing no equations.

What we measure most often is not the BH mass per se but the luminosity for which it is responsible. How closely are they related? Many astronomers have reported correlations over the years, but not always the same ones. Early prejudice that most AGNs would sit at their Eddington luminosities, proportional to mass, has largely gone out of fashion. This time around, we find L∝M² (Sergeev et al. 1999) and M∝L^0.8 (Wandel et al. 1999). Unfortunately, the two papers make use of rather similar, reverberation time arguments, in which the time delay by which changes in H‐beta flux lag the continuum implies a length and so a mass scale.

Why should the relationship between quasars (etc.) and star formation even be an issue? In so far as QSOs have galaxies around them, and all evidence says that they do (Ap95, § 10), stars must have formed and must be contributing to the luminosity we see. Hughes et al. (2000a) confirm that, when you move the slit of your spectrograph off the center of a quasar or radio galaxy, the photons that get through are mostly starlight, not scattered AGN waves.

Nevertheless, very considerable discussion persist on, first, whether specific sources and classes of sources are sending us mostly photons that began around black holes or mostly photons that began in photospheres and, second, on whether these sources or classes of sources can be placed in some evolutionary order, probably involving mergers. Two barriers get in the way of deciding which sorts of photons are arriving. First is that they may have been reprocessed by dust (etc.) blurring or erasing spectral and polarization signatures of their origins. The second is insufficient angular resolution (especially in the submillimeter band, but sometimes also infrared, X‐ray, and even optical for more distant sources). Thus we cannot always separate nuclear from cytoplasmic contributions, let alone decide whether "nuclear" means the inner few AU (near the Schwarzschild radius) or the inner few parsecs.

Notice that disagreement about the dominant energy source for particular classes of sources (LINERs, ultra‐luminous infrared galaxies, IRAS sources, etc.) need not mean that no evolutionary relationships exist. Perhaps the adopted categories are just not the right ones for the purpose.⁶ A few individual sources (especially the bright sort that tend to become unrepresentative prototypes) with lots of light (etc.) of both kinds are also not a real objection to an average, monotonic evolution sequence. The discovery that most really bright, distant astronomical objects have comparable contributions from AGNs and from massive stars would be a strong argument against models in which one systematically evolves to the other. Naturally, several different evolutionary sequences, and none, have been advocated.

Under the circumstances, it is not surprising that of the 230‐plus papers we indexed under AGNs for the year, more than 50 dealt with some aspect of the black hole/starburst connection. The second largest category dealt with "unification"—the primacy of orientation for relationships among subtypes. It included 48 papers, 28 voting yes (including an extensive review by Veron‐Cetty &Veron 2000), 13 voting no, and 7 voting for both or neither.

The place to begin looking for which engine is under the hood is the proceedings of a conference on the subject, for which Joseph & Sanders (1999) provided a closing debate. Since then, surveys of parts of the sky by the Chandra X‐ray satellite and the Submillimeter Common‐User (not meant as the insult it sounds like) Bolometric Array (SCUBA) have added significantly to the inventories. A minimalist summary is that SCUBA finds starbursts and Chandra finds AGNs (Fabian et al. 2000), though a subset of SCUBA‐selected galaxies looks optically as if there are both weak active nuclei and mergers or interactions (Ivison et al. 2000). Indeed the AGN contribution to distant SCUBA sources could be large enough to affect estimates of star formation rates as a function of redshift (McMahon et al. 1999).

The next step beyond this minimalist summary is to look at individual objects and subclasses and what has been said about them by observers. You might choose (a) to watch for cases where the division between black hole and star‐powered luminosity is just what you were expecting from the source name or classes, (b) to watch for cases where it is the opposite, or (c) to skip down 1500 words or so to the theory paragraphs.

LINERs (to start with the faintest) consist either of some sources of each sort (Ji et al. 2000; Barth & Shields 2000) or of sources that do both but both sporadically (Sakamoto et al. 2000). Pogge et al. (2000) collected a large number of HST images of LINERs specifically to resolve the AGN/starburst dichotomy and conclude by admitting that they have worked very hard without answering the original question. Alonso‐Herrero et al. (2000a) may well have worked just as hard to conclude that all of their 10 LINERs have both H ii regions and SNRs from a starburst and also an AGN, but do not seem to regard this as a failure.

3C 18, the brightest FIR source in the 3C catalog (>10¹³ L_⊙), shows no evidence that anything except the AGN is heating the dust (Willott et al. 2000). Indeed a bit more dust would have left it an unidentified 3C source.

Blazar PKD 2155−304 emits optically thin synchrotron all the way from the far infrared recorded by ISO to X‐rays (Bertone et al. 2000), which counts as pure AGN.

Per A=NGC 1275=3C 84 has the fraction of its nuclear infrared luminosity that is due to the AGN at least 10 times that due to stars (Krabbe et al. 2000). We claim therefore that two of the three names are misleading!

In Centaurus A, both near and far infrared are due to stars (Alexander et al. 1999; Unger et al. 2000), the extended synchrotron radio source is fueled by the AGN, and the X‐rays have some of each. Chandra caught it with an off‐center source flaring up brighter than the nucleus (Steinle et al. 2000), and whether the transient is some sort of X‐ray binary or, conceivably, a young SN/SNR, it is star‐powered.

In contrast to Cen A, most of a sample of 79 radio galaxies show at most modest evidence for extra star formation above the norm for their giant elliptical and SO types (Govoni et al. 2000). Smail et al. (1999a) note, however, that many radio galaxies in clusters at z∼0.4 have (contrary to earlier impressions) enough dust to hide current star formation from optical eyes. This is, however, what infrared is for.

NGC 4945 emits near and mid‐infrared that is half to entirely produced by stars, while the X‐rays are all (Comptonized) AGN emission (Spoon et al. 2000; Marconi et al. 2000a).

NGC 6240 is an IRAS galaxy, meaning that you think of stars first, but the X‐ray luminosity measured by BeppoSAX puts it in the AGN box as well (Vignati et al. 1999).

Conversely, Mkn 273 (z = 0.04 ULIRG), which looks like a Seyfert 2, is really driven by a nuclear starburst (Colina et al. 1999). The central region has a number of radio sources that are young SNRs and such, though the brightest central one could be an AGN (Carilli & Taylor 2000).

The host of 3C 48 is undergoing a merger and is close to the peak rate of triggered star formation (Canalizo & Stockton 2000). The less felicitously named UN J1025–0040 finds itself in similar circumstances of being vigorously powered by both sorts of engines (Canalizo et al. 2000).

In principle, there is a distinction between IRAS galaxies (extending up "only" to 10¹² L_⊙, an example of the Scott effect) and ULIRGs, even brighter and so rare that none was close enough to be caught by IRAS. Murphy et al. (1999) found no evidence for an obscured AGN in the vast majority of the IRAS galaxies they examined (broad H‐alpha emission would have counted as evidence). Heisler & De Robertis (1999) presented another IRAS sample that includes starbursts, obscured active nuclei, some unobscured Seyferts, and a few with lots of stars and an AGN. IRAS 12397+3820 is a "both please" Guainazzi et al. (2000).

The very most luminous galaxies also yield a range of deductions from the data sets. (1) ULIRGs with warm 20–60 μm colors are mid‐merger and have both starbursts and fueled AGNs (Surace et al. 2000). (2) Even the 8–25 μm mid‐infrared is largely unresolved (at Keck) and comes from an AGN (Soifer et al. 2000), (3) Optical luminosities of more than and less than 10¹¹ L_⊙ are associated respectively with frequent (87%) and infrequent (27%) presence of compact radio cores (Kewley et al. 2000), but if you take this cut as a definition of an AGN, then many of them still have more starlight than monster light, and (4) two examiners of hard X‐ray sources settle on "obscured nucleus" (Pappa et al. 2000) and "mix of photons from stars and heavily absorbed AGN" (Risaliti et al. 2000).

OK. In the preceding dozen sources and classes are to be found (1) some that are all AGN, (2) some that are all stars, (3) both, with IR and X‐ray apportioned as you would expect, (4) both, but even a good deal of the X‐ray emission is stellar, and (5) both, but even some of the far, mid, or near infrared is black‐hole powered. Under the circumstances, feel free to go fly a kite or engage in some other activity you enjoy while we try to sort out the possibilities from about 10 theory groups and 12 scenarios. In fact we ourselves were a little tempted not to return from a break necessitated by having to explain to a Kindly Old Gentleman on the telephone why our astronomy department could not sell him a star name for the 12th birthday of his great grandson, who is interested in astronomy.⁷

Choice A is that there is no close connection between galaxies in which we have seen starbursts and galaxies that are AGN hosts, on the grounds that the underlying morphology is different (Serjeant et al. 2000b). In this case, we can all fly kites until tea time.

Choice B is that fueling of a black hole and of nuclear star formation is a symbiotic (Williams et al. 1999) or competitive (Oliva et al. 1999) arrangement, with the same gas supply for both at the same time, presumably provided from a merger of gas‐rich galaxy parts, though a monolithic collapse might also work. This entitles you to play Frisbee until the Sun is over the yardarm, and since it really does sound most like the data, be sure your yardarm is in the right direction.

Choice C, that a merger leads to a starburst leads to an AGN is several years old and appears again in Wilman et al. (1999) along with some serious botany of four IRAS galaxies. Roughly concurring are Binette et al. (2000), who describe a radio galaxy with A Past (starburst) and Zheng et al. (1999) whose IRAS galaxies span a sequence from mergers of disk galaxies, to stars, to AGNs, to placid giant elliptical galaxies in their HST photometry. Boisson et al. (2000) suggest some details, based on population synthesis for a central starburst, which ages to a Seyfert 2, on to a clone of NGC 1275 (a radio galaxy, but with gas and all), to a Seyfert 1, to a normal galaxy. Other choices of initiating event would perhaps yield intermediate states resembling the more violent AGNs. Tennis anyone?

Choice D is the converse, that is mergers that do not advance from starburst to AGN, even for very bright galaxies (Rigopoulou et al. 1999) and may even go the other direction, according to Roche & Eales (2000) who find that Fanaroff‐Riley I radio galaxies (a sort of AGN) become ULIRGs. The evidence is that the latter have, on average, closer companions and assumes that mergers are the initiating events.

Finally, you cannot doubt that, if quasars are the consequences of mergers, and mergers were commoner in the past, then quasars should have been commoner in the past. This is undoubtedly true, by (1 + z)⁶ or thereabouts (Percival & Miller 1999). And, having committed the fallacy of affirming the consequent, we can but echo a much more distinguished author who said, "Seyfert galaxies are industrial accidents." We would gladly have committed the fallacy of the undistributed middle, but find that our middles have become a good deal more distributed over the 10 years of this series of reviews. Post hoc, ergo propter hoc.

The 19 values of Hubble's constant published during the year (giving only one vote to a package of five papers from the HST Key Project Team; Mould et al. 2000) had a median of 64 km/sec/Mpc, hit on the nose by Jha et al. (1999), who used 43 supernovae and a calibration that includes metal‐dependence of the period‐luminosity relation for Cepheid variables. The median continued the gentle rise with time noted in Ap99 (§ 12.1, where it was 62). Only Phillips et al. (1999) gave a three‐figure answer (63.3), we and they both hesitating to describe these as "significant" figures. They too are supernova buffs and used color curves corrected for extinction in the host galaxies. The widest error bars were admitted by Lahav et al. (2000b) and covered both 50 and 70 km/sec/Mpc, but then it is really a paper on Bayesian methods, which are excellent for changing one's mind gradually, but less good on the road to Damascus. The data are measurements of fluctuations of the cosmic microwave background radiation, and a cosmological constant that contributed 0.7 toward a critical density of 1.0 was assumed.

Cepheid variables and Type Ia supernovae were, inevitably, the most discussed of the 20 or so distance indicators we read about. At least two Cepheid problems remain. The first is that absolute luminosity at a given period undoubtedly varies with metallicity, via the mass, radius, and temperature corresponding to a given period. There are real differences (e.g., Paczynski & Pindor 2000); the models do not entirely reproduce available data (Antonello et al. 2000); and (where have you heard this before?) "more data are needed" (Caputo et al. 2000b). The model‐based corrections slightly increase the H found by the methods of the Key Project, because (Caputo et al. 2000c) most galaxies that host Type Ia supernovae are more metal‐rich than the Large Magellanic Cloud. Use of infrared as well as optical colors can improve the chances of being able to standardize SNIa's in different sorts of environments (Krisciunas et al. 2000).

The second residual error source is Cepheids whose images are contaminated by real or optical companions. If no correction is made for this, you will give the star a brighter apparent magnitude and a smaller amplitude than it deserves, and think that it is closer to you than it really is. If RW Car, with a close binary companion nearly as bright as itself (Fernie 2000) were typical, this would be a disastrous source of systematic error. On average, it is perhaps 0.1 m in measured magnitudes (Gibson et al. 2000; Saha et al. 2000) or 10% in H (Mochejska et al. 2000). This last comes from a project called DIRECT but, according to the authors, uses indirect methods!

Type Ia supernovae are also afflicted by two problems of rather similar type, one concerning the physics of the events and one observational in nature. First is potential dependence on the stars involved that might vary systematically with redshift (and which also is a worry if you are trying to get Ω and Λ from SN observations). By way of reminder, Type Ia's are nuclear explosions in degenerate carbon‐oxygen cores or white dwarfs. Thus their peak luminosity can be affected by total metallicity of the initial star (via the C/O ratio in the degenerate core), core mass, whether the progenitor is a single or binary star, and whether the burning front moves super‐ or subsonically (detonation vs. deflagration). That is, SNa Ia are not standard candles. Some of the properties, including burn speed, apparently also affect the correlation of peak luminosity with rate of fading that is normally used to standardize them (Sorokina et al. 2000). All these would matter relatively little if nearby and distant events included the same mix of progenitors and explosion processes. We found during the year one vote for "same" (Aldering et al. 2000) and one for "different" (Riess et al. 1999a, 1999b). In the latter data set, the more distant events are fainter on average. This (as remarked upon by Sidney van den Bergh at a 1996 conference) would be mostly unlikely as an observational selection effect. The dominant physics is that progenitors of smaller initial metallicity end up with cores having smaller C/O ratio and so less energy to release in the flash up to iron.

The second supernova snake‐in‐the‐grass is the possibility of gray, uniformly distributed dust dimming apparent magnitudes of distant events without reddening them (Croft et al. 2000), so as to mimic the relation between redshift and apparent magnitude in a universe with q<0 or positive cosmological constant, Λ. The discussion by Aguirre & Haiman (2000) comes quite close to a recantation from one of the stronger proponents of the dust alternative. A definitive test is, however, at hand. Intergalactic dust sufficient to spoil our Ia candles will scatter the images of X‐ray emitting QSOs enough for detection by the current generation of satellites. If we see the scattering the dust is there. If we don't it isn't.

Given some ultimate X‐ray mission, one could even measure directly the distances to Type II (X‐ray emitting) supernovae, using the time delay of radiation scattered by intervening dust. Predehl et al. (2000) have just made this work for Cyg X‐3 (9 kpc from us), though it failed for Sco X‐1 (Bradshaw et al. 1997). And the author who remembers 1965 best will always thing of this as the Slysh mechanism, because it was first proposed by Overbeck (1965).

Only 18 distance indicators remain to be addressed. Keep in mind, however, that when you can apply two or more, they often do not agree. For instance, Aparicio et al. (2000) find smaller distances to some dwarf galaxies from red giants than from Cepheids. They think that the RG distances are right, or at least that the Cepheid ones are wrong. We will, therefore, not attempt to adjudicate among the indicators, but just list them, in approximate increasing order of the distances at which they are useful, with one reference each, in case you need an entry point to learn more about them.

(1) Parallaxes and main sequence fitting for open star clusters (Gatewood et al. 2000), (2) double‐lined, spectroscopic eclipsing binaries (Ribas et al. 2000), (3) a period‐luminosity relation for Delta Scuti stars (Peterson & Christensen‐Dalsgaard 1999), (4) clump stars, which have in spades the problem of age and composition dependence (Sarajedini 1999), and put the LMC claustrophobically close to us at 44.5 kpc (Udalski 2000), compared to (5) RR Lyrae stars (Caputo et al. 2000a), though both disagree with (6) field horizontal branch stars (Carretta et al. 2000) and (7) stars at the tip of the red giant branch (which put the LMC at 51 kpc, Cioni et al. 2000), (8) planetary nebulae (Cazetta & Maciel 2000), (9) long period variables, which also have a "parent population" problem (Barthes et al. 1999), and (10) closely related red supergiants (Jurcevic et al. 2000), (11) globular clusters, with continuing uncertainty from the bimodal distribution of luminosities in many galaxies (Lee & Kim 2000) and from how you tie them to local subdwarfs (Salaris et al. 2000), (12) novae, another multi‐population species, since the accreting white dwarf is likely to be of larger mass and different composition among the youngest stars (Shara et al. 1999), (13) fluctuations of the surface brightness of galaxies (Jerjen et al. 2000, who looked at dwarf ellipticals), (14) the I−K colors of spirals, which the authors de Grijs & Peletier (1999) say is not as good as (15) the Tully‐Fisher relation between the maximum rotation speed in a spiral and its physical diameter (Koda et al. 2000, on the physical cause of the correlation—galaxy mass—and the scatter—galaxy angular momentum), (16) the Faber‐Jackson relation between velocity dispersion and physical size in elliptical galaxies, whose scatter arises from younger galaxies having formed when the cosmic density was smaller (Forbes & Ponman 1999), and (17) the Sunyaev‐Zeldovich effect, in which the electrons in X‐ray emitting clusters of galaxies upscatter photons passing through from the CMB, with X‐ray luminosity and CMB decrement depending on different powers of distance. Blasi et al. (2000b) point out that a non‐thermal tail of high energy electrons (responsible perhaps for hard X‐ray excesses in the spectra of some clusters) can mess up the whole thing royally. And, in case you think you have been deprived of the 18th method, notice that there are actually two in number (1).

We voted against fractals (hierarchical structure persisting on scales larger than about 150 h^-1 Mpc) last year and see no reason to ask for a hand recount this year (Bharadwaj et al. 1999; Hatton 1999). Indeed the observations can almost be said to have reached a steady state on several points, with ξ(r) proportional to r^−1.7–r^−1.8, intrinsically bright galaxies more clustered than faint ones, ellipticals more clustered than spirals (Seaborne 1999 on Stromlo APM vs. IRAS galaxies), and much of the structure on sheets and filaments (Lahav et al. 2000a; Schmalzing et al. 1999; Bharadwaj et al. 2000; Guzzo et al. 2000; Fynbo et al. 2000, each with a different sample, but rather concordant results).

What you see is redshift dependent in the expected sense—that is, at z≳4 you pick out only the galaxies in rare, massive haloes which are going to be more clustered than commoner halo types by the time you reach z = 0, and this shows up as a large bias parameter, b = 5 at z = 4 (Arnoutz et al. 1999; Magliocchetti et al. 2000). In between, there is actually a good deal of what we would just the other day have called high redshift clustering, but might now call intermediate redshift clustering (Steidel et al. 2000; Williger et al. 2000; Sanchez & Gonzalez‐Serrano 1999; Kurk et al. 2000).

That radio data alone will not reveal this large scale structure apparently has held from the time of the 3C (Third Cambridge) survey in the 1960's down to the present, according to Artyukh (2000) and Venturi et al. (2000), who are unable to find even the Shapley concentration. They suggest that merging of clusters of galaxies might switch off existing radio sources, so that they do not trace the largest scales.

What remains to be sorted out? First are continuing discrepancies among several redshift surveys on how much power persists out beyond 80 h^-1 Mpc (Hoyle et al. 1999) and ditto for how much deviation there is from pure Hubble flow on these larger scales (Giovanelli et al. 1999). Next, is there really a Great Attractor? Yes, say Woudt et al. (1999), but it is mostly the Norma or ACO 3627 cluster.

Third, should we believe various, previously‐advertised, quasi‐periodicities in the redshift distributions for galaxies or clusters? No, say some observers (Drinkwater et al. 2000). Yes, say several theorists, perhaps because their theories can explain it (Kaminker et al. 2000; Krilova & Chizhov 2000), and if the period is 130 h^-1 Mpc at all redshifts, then you can use it as a standard meter stick to measure the cosmological constant (Roukema & Mamon 2000), which, of course, turns out to be about 0.7.

Fourth, where does the dipole converge? This is the local version of "how big is the largest structure?" and answers range from 60 h^-1 Mpc (that is 6000 km/sec, da Costa et al. 2000) to 200 h^-1 Mpc (Rowan‐Robinson et al. 2000) and more than 18,000 km/sec, over which peculiar velocities do not trace the presence of the obvious observed clusters and voids, and the bulk flows do not decrease with distance Karachentsev et al. 2000b). Tomita (2000a, 2000b) finds that we live in an under‐dense region of about this size. New massive data sets like SDSS should clarify most of these issues in a few years, so we feel no compulsion to brush off the hanging chads and vote this year.

On the theoretical front, simulation, and perhaps dissimulation continues. A real highlight was the comparison of 12 codes applied to the same problem (predicting what typical clusters should look like, given standard cold dark matter and non‐radiative gas) Frenk et al. (1999) report that most results agree to 50%–10%, apart from the predicted X‐ray luminosities, which are very sensitive to small fluctuations in the gas density. Incidentally, "al." includes all 12 coders.

The 24th General Assembly of the International Astronomical Union in August 2000 had in common with the 23rd GA in Kyoto in 1997 a symposium devoted in large measure to the cosmological parameters. And at least the author who attended both was surprised not by new values being reported but by the near‐constancy of the bandwagon numbers over the intervening three years. That is, most of the speakers and poster presenters either supported or thought it necessary to oppose a set of numbers:

• •  
  Ω (matter) = 0.30–0.35, of which
• •  
  a few % = baryons,
• •  
  ≲1% = neutrinos (but probably not zero), and the rest is cold dark matter,
• •  
  k = 0 (flat space),
• •  
  Ω (Λ) = 0.65–0.7 (cosmological constant, quintessence, or dark energy),
• •  
  age = 13–14 Gyr (for H = 65),
• •  
  n = 1 (Harrison‐Zeldovich spectrum of primordial fluctuations),
• •  
  σ₈ ≈ 1 (normalization of that spectrum),
• •  
  q<0 (that is expansion currently accelerating).

This seems to be 10 numbers, but they are not all independent. For instance, if you claim to know Ω_m and Λ but not to know k (the spatial curvature), then you are fighting Einstein and perhaps even Euclid. It is, however, possible to carry out measurements of these three things by different methods and arrive at results that are not mutually consistent.

It would be dishonest to claim that the literature of the index year was equally coherent. But, of the roughly 100 indexed papers, only 10 or 15 disagreed significantly (and observationally) with the consensus. Most unfairly, we note only the following: (1) consistent, fairly standard sets of numbers along these lines published by Mauskofp et al. (2000), Melchiorri et al. (2000, including new CMB data), Novosyadlyj et al. (2000, emphasizing large scale structure), Henry (2000, emphasizing X‐ray clusters), and Bridle et al. (1999 discussing the normalization σ₈); (2) a new, independent estimator of matter density from weak gravitational lensing by structure larger than clusters (Wittman et al. 2000), which also favors an open or Λ‐dominated universe, but be warned that van Waerbeke et al. (2000) believe that they have seen lensing by cosmic shear, which is not part of the consensus model, (3) an improvement of the age estimate, 14 ± 3 Gyr, from the ratio of thorium to europium in the globular cluster M15 (Sneden et al. 2000); (4) evidence for both a cosmological constant and some neutrinos from counts of quasars (Lin & Chu 1998); (5) non‐confirmation for non‐Gaussianness of the CMB (Contaldi et al. 2000); yes the word sounds awful, but non‐Gaussianhood or non‐Gaussianity aren't much better; and (5) a few candidates for "where are most of the baryons?" including low surface brightness galaxies (O'Neil & Bothun 2000), ordinary X‐ray emitting clusters (Wu & Xue 2000), hot supercluster gas (Boughn 1999), and gas at 10⁵–10⁷ K that emits various sorts of EUV continuum (Lieu et al. 1999; Tripp et al. 2000) and soft X‐rays (Scharf et al. 2000). All of these loci are concentrated around galaxies, clusters, sheets, and filaments, not spread in a uniform intergalactic medium.

Our minds are, pretty much, made up (though see § 12.5) on this: 90% or more of the stuff that contributes to gravitational potentials in the universe does not emit or absorb its fair share of light (that is, dark matter exists), and we haven't a clue what it is. One each from columns (A) baryonic, (B) hot (neutrino‐like), (C) cold (neutralino‐like), and (D) pressure (lambda‐like) must, you might suppose, surely include at least portions of the truth, though some of the authors cited below would perhaps not endorse even this most general sort of description. The ordering of the candidates is, roughly, this A to D, followed by (E) "other," and we apologize to anyone whose favorite we have misunderstood so badly as to include it in the wrong category.

Just in case you might have forgotten, dark matter can be both warm (Sellwood 2000, unfortunately no better than hot or cold for making large scale structure) and fuzzy (Hu et al. 2000). The latter consists of 10⁻²² eV scalar particles, cold because they have come from a Bose‐Einstein condensate rather than from a heat bath (like axions in this respect) and good for avoiding excess structure formation on small linear scales.

12.4.1. The Baryons

12.4.2. Hot or Cold

12.4.3. Topological Defects

12.4.4. None of the Above

12.4.5. Double Dark

If you are casting only one vote, save it for the standard general relativistic, expanding, hot big bang with one to a few kinds of fairly conventional dark matter (§§ 12.1–12.4.3). Here, however, are some of the other candidates for Universe of the 21st Century. The first set includes ideas that are "in addition to" the standard model and which may help to explain, clarify, or justify it. Next come the "instead of" models, beginning with old friends like MOND and quasi‐steady‐state. And we trickle off into mists of incomprehension. Lines between the sets and between model universes and DM candidates are not sharp, for instance in the case of the transiently accelerating universe of Barrow et al. (2000).

12.5.1. Branes, Brains, and Extra Dimensions

12.5.2. "I Never Forget a Face ..."

12.5.3. "... But in Your Case, I'm Willing to Make an Exception"

A generous senior colleague offered recently to nominate one of us for this. The actual name turned out to be slightly different. But we do think that there should be a prize given to the astronomer who did the most during the year to encourage a general pulling up of socks in the field. The colleagues who pointed out to us the following mistakes in Ap99 are all candidates.

Sect. 10.7. HDE 31685 should have been HDE 316285.

The first mentions of Hg²⁰⁴ in Ap stars (§ 6.4) and stellar rings (§ 8.6) were probably Bidelman (1956) and Hodge (1986). Well, we didn't say they weren't, but the Bidelman paper was improperly omitted from a much earlier review (Trimble 1975).

Sect. 6.7. The RV Tauri stars in globular clusters are too RV Tauri stars from very early times (Arp & Wallerstein 1956). In addition, C. H. Payne Gaposchkin and H. S. Hogg so categorized them in well‐known catalogues, and, as the chap who called it to our attention pointed out, "If General Motors calls it a Chevrolet, it's a Chevrolet."

Sect. 10.8. The blueshift should, of course, have been described as Δλ/λ = -0.1 to −0.3, not pluses, which are redshifts!

Sect. 10.8. The oldest archived plates greatly predate 1879. Harvard has one of the Pleiades from 9/10 October 1857. It shows three stars.

Sect. 5.4. A colleague attempted to clarify for us the phrasing of onset of declines in R CrB stars relative to pulsation phrase, citing Duerberk et al. (2000), who apparently got it wrong. The real story is that minimum brightness within a main fading event is also minimum brightness in the pulsation cycle, but, the explanation continued, "Unfortunately I am also 'allows' me similar mistakes. And every time I am afraid to devise a bicycle." Well, some of ours have square wheels too.

Sect. 9.2. Several readers (including an undergraduate) recognized the terminology from jacks and offered to beat either author.

Sect. 9.3. The "issue of an archival journal with a cD attached to its cover." We could have sworn that the manuscript said CD, but the commentator's reaction was that the MACHO people should be searching closer to home. Indeed a whole cD galaxy would make a spectacular "microlens" and hasten the moment at which the paper versions of major journals will grow to close the universe.

A recently‐postdoctoraled astronomer explained his possibly premature publication last year on a hot topic by saying he had not wished to be a candidate for this prize. Challis, in case you might have forgotten, actually observed and charted Neptune from Cambridge on the 12th of August, in a bit of sky he had already examined on the 30th of July, but had not bothered to compare his notebook pages until after Galle and d'Arrest had announced their discovery. It was back in 1846, so you can be excused for not remembering the details, but many of the items which follow seem to have their origins in a rush to print (or, occasionally, astro/ph).

"Goods are totally free from bark and apparently free from live plant pest." So said the document that came with a volume we recently ordered. It was, unfortunately, the best review that book received.

"Enclosed is a brief survey to help ensure that our records pertaining to legacy gifts of the type set forth in the attached survey." We didn't fill out the survey (which was part of a fund‐raising letter from the SETI Institute).

"Since the N/H ratio plays an important role in the formation scenarios of Jupiter, we wanted to confirm or infirm the Galileo result." (BAAS, 32, 1014).

"See Author (1992) for a pedestrian derivation of ... " (ApJ, 529, 433, footnote). We are pretty sure "Author" did not see this before it was published.

"Many groups ... specifically wave ... the proprietary period in their proposals." (STScI Newsletter, 16, No. 4, p. 2). As in, wave bye‐bye.

"Typical poor Abel clusters" (AJ, 119, 611). They can't even afford that second 1.

"Origin of the X‐Y Relation" is part of the title of ApJ 534, L89. The abstract explains, "There is no X‐Y relation."

"Baade (1993) cited in PASP, 111, 1150. They were tough in those days. But a number of papers seem to have had trouble keeping track of who their own authors were, for instance: A, B, C, and D are the first‐page authors of ApJ, 525, 10, but the running head is A, B, and D, with C in the acknowledgements. "This work was performed as part of the author's PhD thesis" say the acknowledgements of ApJ, 524, 372, but the paper has two authors. A, B, and C, are the first‐page authors of ApJ 528, 436, but the running head mentions only A and B. The paper A&A, 351, 358, by authors X and Y, mentions a private communication from Y. Can you whisper in your own ear?

"The distance has been revised from 65 ± 5 pc to 65 ± 35 pc" (A&A, 353, 1044, abstract) is perhaps just great (and meritorious!) honesty, but the following incorporate numerical errors of factors of two or larger, or geometry in which pi is different from the familiar value.

"19,636 km (one and a half times the earth's circumference)" (Science, 287, 53).

"Thankful to G. J. M. and Y. H. C. for allowing us to use his ... mosaics" (AJ, 118, 1408, acknowledgements). It happens also that Y. H. C. is a "she."

"We find 7 X's, including 4 Y's and 4 Z's" (PASP, 111, 1398, abstract). Well, maybe one of the Y's is also a Z, but the context makes it unlikely.

"...thermal transport of the eleven orders of magnitude of the Rayleigh number (10⁶≤R_a≤10⁷)" (Nature, 404, 837, abstract).

"Application of the results to the intracluster medium is being reduced to 5–7 orders of magnitude less than the mean free path due to Coulomb collisions" (ApJ, 533, 84, abstract and text). Can you reduce an application? Maybe. Can you reduce it to be less than a mean free path? Really hard.

Picture this! AJ, 199, 2338 shows a Hertzsprung‐Russell diagram with hot stars to the right. Deliberate clearly, but it takes a bit of getting used to. "The added grey‐scale solid symbols" of ApJ, 536, 274 are actually red. But to make up for it, MNRAS, 310, 983 Fig. 1 attempts to color‐code using red and magenta, while ApJ, 523, 555 and 557 have rediscovered quarter‐tones, the sepia variant on the sort of picture that used to be called a half‐tone.

"Le mot bust." In most of these, we think we know what the author meant, indeed some of them were probably deliberate. Feel free to add your own commentaries of the sort provided in Observatory magazine's "Here and There" section. "Future space missions ... will be able to shade new light on our knowledge" (A&AS, 145, 323, abstract). "... kinematics of a large deal of the local ISM" (A&A, 358, 299, abstract). "... and then...adding proscribed amounts of heat per particle ..." (ApJ, 532, 17). "The start solid lines represent..." (A&A, 349, 832, fig. caption). "Hereto we use" (New Astron., 4, 167). "The overall flow properties of derived by ..." (MNRAS, 319, 954, conclusions). "SOB" (PASP, 112, 662, text; it mean superoutbursts, but smile when you call me that). "If the absorptions is responsible for the ISM..." (ApJ, 534, L184). "A reddening map to control the patchiness of the dust" (ApJ, 527, 167, abstract. If only we could!). Section title "Appendix A: Appendix" (A&A, 356, 1001; luckily there is no Appendix B). "But there is a prize to pay: the initial advantage of genetic algorithms usually not to end in side‐extrema is reduced" (A&A, 357, 1180, footnote). "Endly, we have compared ..." (A&A, 352, 382, conclusion). "... suddenly shifts inward vvvv from about a few hundred ..." (A&A, 354, L67). And, undoubtedly written with heartfelt intent, "... thank C. G. for insightful harassment" (ApJS, 125, 361).

We end with a few items that could be catalogued under, "With friends like these..."

"Human skeleton dating from 11,500 to 11,500 million years ago" (Science, 287, 874, a letter to the editor). "Contributions which will appear in the PASP through the year 200 to mark the upcoming millennium" (PASP, 112, 869, a very old footnote). And, "A few years ago, astronomers determined that distant supernovae were receding from the center of the universe much faster than before" (Science, 289, 1109). But, if you are left feeling unappreciated, just remember with Robert May (writing in Nature, for 4 May 2000) that, "There are no societies to express public support for nematodes."

Author Aschwanden made use of the Astrophysics Data System (ADS) and his work was partially supported by NASA contract NAS8‐00119. Author Trimble made use of the libraries of the University of California and the University of Maryland and, increasingly (and regretfully, as library collections become less and less accessible), of private journal subscriptions. Her page charges were partially supported by fees received for writing and editing for Sky and Telescope and the American Physical Society. A special thank you to the Boise Cascade Office Products company whose "recycled steno notebook" provided space for this year's indices and some of the inspiration for § 13.

Colleagues to whom we are grateful for providing input, output, throughput, and up‐with‐it‐put for Ap2000 include Lawrence H. Aller, William P. Bidelman, Tereasa Brainerd, Kris Davidson, Luke Dones, Yuri Efremov, Martin Elvis, Michael Feast, Luigi Foschini, Martin Gaskell, Roger Griffin, Carl Hansen, Chris Impey, Rosina Iping, James Kaler, Korado Korlevich, Kevin Krisciunas, Kam‐ching Leung, Brian Marsden, Ray Martin, Adrian Melott, Leon Mestel, Eugene Milone, Michael Molnar, Leos Ondra, Bohdan Paczynski, Alexander Potekhin, Saul Rappaport, Alexander Rosenbush, Larry Rudnick, Sebastian Sanchez, Steve Shore, Zhen‐ru Wang, George Wallerstein, Doug Welch, and Lodewijk Woltjer. Prof. Christian Klixbull Jørgensen (of the chemistry department of the University of Geneva), who had provided input for a number of years, died during the index year.

Once again, co‐editor Anne P. (what does it stand for?) Cowley helped with the references at a level deserving of co‐authorship or sainthood.