EXTRASOLAR Jupiter-like Discovered The observation, by high-precision spectroscopy, of a Jupiter-mass planet in a nearly circular orbit around 51 Pegasi in the left wing of the Pegasus constellation not only represents the first step on the road to identifying the first extrasolar planetary system associated with a solar-type star but also raises intriguing questions concerning the way large planets form. The orbital motion of 51 Pegasi corrected for the long-term variation of the velocity of The detection of planets around main- currently limited to detecting velocity the centre-of-mass owing to the possible sequence stars other than the Sun repre­ changes of about 15 m/s. Since the reflex presence of a second companion of low sents a major astrophysical and instru­ motion of the Sun due to Jupiter is 13 m/s, mass. The points, plotted as a function of mental challenge [EN 26 (1995) 8]. A planet all searches based on spectroscopy thus the phase of orbital rotation, correspond to orbiting around a star would be expected to aim to detect companion objects which are experimental estimates of the radial velo­ reveal itself by its own emission or by a at least as large as Jupiter. city as determined using cross-correlation perturbation in the star’s characteristics. The radial velocity was computed using a of spectroscopic data. The solid line is the Effects include wobbling of the star, reflec­ cross-correlation method that concentrated theoretical curve fitted for a circular orbit tion of the star’s light on the planet, variation the Doppler information of about 5000 with a period of 4.2293 ± 0.0011 days. It in the star’s angular position (the star has a stellar absorption lines. The location of a shows that the data are remarkably stable reflex motion around the common centre-of- Gaussian cross-correlation function fitted to and sinusoidal. mass of the star/planet system which leads the spectral line intensities gave the radial to a variation of the star’s angular position), velocity. A high accuracy was achieved gravitational perturbation giving a periodic owing to the scrambling effect of the optical orbital plane. If one makes the reasonable modulation of the star’s radial velocity, fibre input and by continuously calibrating assumption that the rotational axis of 51 occultation and gravitational amplification of the instrument during exposures. Pegasi is aligned with the orbital plane, light from the star, timing of pulsar emission, The calculated orbital motion of 51 Pegasi three sets of independent measurements intrinsic radio emission, etc. is plotted in the figure giving the radial give a projected velocity of 2.2 ± 1.0 km/s. By monitoring the time of arrival of signals velocity as a function of the phase of orbital Chromospheric activity caused by the re­ from a pulsar, a complete planetary system rotation. The data have been corrected for a emission in the core of calcium lines is consisting of two terrestial-mass planets long-period perturbation that probably arises related to the speed of stellar rotation via the orbiting with periods of months has been owing to the presence of a second compa­ magnetic field created by the dynamo effect. found, but at a place where nobody ex­ nion with a small mass. Intensive monitoring In the case of 51 Pegasi, the activity indi­ pected it [1]. Planets orbiting around solar- of 51 Pegasi is in progress in order to con­ cates a rotation period of 30 days so the type stars rather than pulsars offer more firm this long-period orbit. The solid line in equatorial velocity is 2.2 ± 0.8 km/s. Com­ exciting perspectives (at least to the general the figure represents the computed motion bining these two velocities gives an lower public), so it is understandable that the assuming a circular orbit with a rotation pe­ limit of sin /= 0.4 for the inclination angle i of announcement on 6 October of the detec­ riod of 4.2293 ± 0.0011 days for 51 Peg B. the orbital plane relative to the line-of-sight. tion, for the first time, of a Jupiter-mass At the present time, the eccentricity of the Consequently the companion responsible companion orbiting around a solar-type star orbit is estimated to between 0 and about for the 4.23-day orbit has an upper mass attracted considerable attention. 0.15. limit of 1.2 Jupiter-mass. If one assumes Using data from the new high-precision The remarkably sinusoidal and stable misalignment as large as 10°, the compa­ ELOIDE fibre-fed echelle spectrograph at velocity curve, which favours the planet nion has still to weigh less than 2 Jupiter- the Observatoire de Haute-Provence, interpreation, has been confirmed indepen­ mass. and of the dently by a joint US team from the Lick The 30-day rotation period of 51 Pegasi is Observatory have reported [2] that Observatory and the Harvard-Smithsonian clearly not synchronized with the 4.23-day the presence of a very low mass (0.5 Center for Astrophysics. orbital period of its low-mass companion Jupiter-mass) companion, called 51 Peg B, An orbital period of only 4.23 days is not despite 51 Pegasi’s very short period. This in an orbit with a semi-major axis of 0.05 too surprising for a binary star system. rules out the possible presence of a low- Astronomical Units (AU), seems to be the However, it is rather puzzling if one relates it mass stellar companion since spectroscopic most convincing explanation for the ob­ to the relatively long orbital period of Jupiter. binaries of similar periods have been found served velocity variations of the solar-type 51 Pegasi is a 5.5th magnitude star located to be synchronised in all cases. Using ob­ star 51 Pegasi. Other candidate stars in a 45 light years away. It is similar to the Sun, served scaling laws, the lack of synchroni­ selected sample have also shown signifi­ having much the same temperature and zation can in principle be used to estimate cant velocity variations over an 18-month being only slightly older. Its only remarkable an upper limit for the mass of 51 Peg B. period of observation, but they require feature is a slightly higher abundance of additional measurements. heavy elements, and even this can be ac­ The Only Convincing Explanation The Geneva team has monitored since counted for by the uncertainty in measuring April 1994 the radial velocity of a total of elemental concentrations. If one assumes a Since the observed velocity variations are 142 G and K dwarf stars, this sample having random distribution of the planet-star orbital small and of short amplitude, it is necessary been selected for its apparent constant plane, there is a 1% probability that the to rule out possible alternative explanations radial velocity from a much larger sample of planet is four-times the mass of Jupiter and for the spectroscopic measurements. Rota­ stars that had been monitored for 15 years a 2.5 x 10-5 probability that it is heavier than tion of a hot spot on the star’s surface can at a lower sensitivity (200 m/s). The ELOIDE the minimum mass for hydrogen burning be dismissed on the basis of the lack of spectrograph, optimised for Doppler-shift (0.8 solar masses). So one is probably chromospheric activity and the long rotation studies, was used to measure the variation dealing with a fairly low-mass companion. period (a solar-type star rotating with a of the stars’ positions along the line-of-sight. This conclusion is supported by examin­ period of 4.23 days would have a much The sensitivity of the method means that it is ing the rotational velocity projected onto the larger chromospheric activity). Europhys. News 26 (1995) 123 Pulsation of 51 Pegasi would be accom­ panied by luminosity and colour variations as well as by phase-related absorption line The Heliosphere and its Neighbourhhod asymmetries. However, all-sky surveys car­ ried out by the HIPPARCOS satellite indi­ The general characteristics of the cate that stars with spectral features similar heliosphere (W.I. Axford, 1992). to those of 51 Pegasi are among the most stable. Moreover, no mechanisms have The Sun’s immediate neighbourhood — the been identified which are able to excite heliosphere — is dominated by the solar pulsation modes as long as four days in magnetic field and the solar wind. How it is solar-type stars. For instance, only very low embedded in the local interstellar medium amplitude (« 1 m/s) modes and periods of (LISM), and how these two fundamentally minutes to one hour are detected for the different media interact with each other are Sun. Radial velocity variations of a few days of fundamental importance. A dozen space and less have been observed for a few giant missions are presently making observations stars larger than 20 solar radii in size. How­ both inside the heliosphere by in situ moving at a somewhat higher speed, ever, 51 Pegasi is certainly not this large. measurements (e.g., Ulysses, exploring its caused by the neighbouring G-cloud (locat­ Nor does it show short-period simultaneous three-dimensional structure, and the ed towards the galactic centre), into which pulsations that are characteristic of giant Voyager and Pioneer spacecraft in the we will move sometime during the next stars. outermost regions) and outside the LISM in millennia. Photometric measurements of 51 Pegasi all wavelength bands (e.g., optical observa­ The composition of the gas in our local and of two companion stars carried out tions from the Hubble Space Telescope, and cloud can, in principle, be determined from earlier this year do not completely rule out ultra-violet and X-ray measurements from in situ observations of interstellar pick-up the possibility of a very low pulsation. EUVE and Rosat, respectively). ions in the solar wind. However, it became Although stronger constraints stemming The main task of the workshop The apparent at the workshop that the inter­ from HIPPARCOS data are expected short­ Heliosphere in the Local Interstellar Medium pretation of data needs a better under­ ly, it is noted in the meantime that pulsations (Bern; 6-9 November 1995) with over 40 standing of ionization rates and of ion trans­ are known to affect the symmetry of absor­ participants from 9 countries on 3 continents port in the interplanetary medium. None­ ption lines. Such a feature has been sought — the first event to be held at the new theless, the pick-up measurements confirm for without success using the cross-correla­ International Space Science Institute (see the view that these ions form the seed tion technique. So an interpretation of the insert) — was to bring together representa­ population of the anomalous cosmic-ray observed velocity variations as being due tives of these two approaches, as well as component, whose most prominent feature to the orbital motion of a very low mass theorists and modellers of the heliosphere, is the virtual absence of carbon. In line with companion planet seems to be the most in an enviroment which would stimulate the this view, the recently discovered, and quite convincing. exchange of views and data. The aim was to unexpected, pick-up carbon ions could be reach agreement on the principal parame­ shown to stem from a local source inside the Exciting Perspectives ters (such as the size of the heliosphere and heliosphere (their three-dimensional distri­ the density and temperature of the LISM) bution indicates that evaporating interstellar It is clear that the mass and orbital ec­ and to identify points of disagreement and grains are the source). centricity of the 51 Peg B companion are ways to overcome them. The distance to the termination shock Rt similar to those for heavy planets. But this Excellent agreement was reached regard­ where the solar wind passes from the does not mean that the companion forms in ing the direction and speed at which the supersonic to the subsonic regime, and to the same way as Jupiter. Most importantly, heliosphere moves through the local cloud, the heliopause Rh where the solar wind present models do not predict the formation as obtained from in situ observations of meets the interstellar gas, has been ad­ of Jupiter-like planets with separations as interstellar neutral gas on Ulysses and from dressed by several different methods. Per­ small as 0.05 All (well within Mercury’s orbit Hubble observations of interstellar ab­ haps the clearest indication of comes in our Solar System). It is unlikely that the sorption lines in the spectra of bright stars. from two major radio events observed by accumulation of ice grains to give a Jupiter- However, is now clear that accurate values Voyager, placing it at a distance of 110-160 like planet followed by orbital decay due to can only be obtained by monitoring helium astronomical units (AU). Theoretical modell­ dynamic effects would result in a orbit as because other constituents of the LISM (in ing indicates Rt = 2/3 ffh, putting the ter­ small as 0.05 AU. Secondly, all of the pla­ particular, hydrogen and oxygen) are affect­ mination shock well within reach of the nets in the Solar System which are heavier ed either by deceleration and pile-up up­ Voyager spacecraft, currently at over 60 AU than 10-6 solar mass have circular orbits stream of the heliopause or by deflection in and moving outward at nearly 4 AU per since they grew from a protoplanetary front of this boundary. year. A poll among the participants resulted gaseous disc. The low eccentricity of 51 Observations of interstellar hydrogen ab­ in a value of Rt = 87 ± 5 AU, which means Peg B does not represent evidence for a sorption lines and of the backscatter of solar that Voyager ought to reach the boundary of similar effect since the separation is very Lyman-α radiation from the LISM reveal that our heliosphere within the next decade. The small. Instead, dynamical evolution of the our local cloud, which is about 1 parsec (3 x result may, however, be biased to a low system rather than the formation conditions 1013 km) in size, may not be the home of value owing to the participants’ hope that may be responsible for the low eccentricity. the heliosphere much longer. Absorption they will be alive to witness this historic One possibility is that 51 Peg B resulted spectra clearly show a second component from the radiative stripping of a nearby brown dwarf of low mass. if this is the case, it would be mostly made up of heavy International Space Science Institute elements. The International Space Science Institute the new institute will be able to “bring out an Notwithstanding these more general (ISSI) was formally Inaugurated on Novem­ integrating view” to studies that encompass issues, it is the possible presence of a ber 10, several months after opening its many aspects. This approach clearly struck second long-period companion to 51 Pegasi doors to collaborators. Modelled on advan­ home to delegations to the Inter-Agency that is of the most immediate interest. If it ced studies institutes, the ISSI aims to Consultative Group (IACG) from the Euro­ also turns to be in the Jupiter range with a enable space scientists involved in today’s pean Space Agency (ESA), NASA, Japan’s nearly circular orbit, the Geneva team will many space missions to pool data and Institute of Space and Astronautical Studies have discovered the first example of an understanding in order to interpret scientific (ISAS), and the Russian Academy of Scien­ extrasolar planetary system associated with results in a broad, Interdisciplinary context ces when the IACG endorsed the ISSI’s a solar-type star. while not overlooking the results of ground- scope in 1994. In the 1980’s, these national [1] Wolszczan A., Frial, D.A., 355 (1992) and laboratory-based research. J. Geiss, agencies mandated the IACG to coordinate 145; Wolszczan A., Science 264 (1994) 538. the ISSI Executive Director and a prime missions to Halley’s and prog­ [2] Mayor M., Queloz, D., Nature 378 (1995) 355. instigator behind the initiative, thinks that rammes in solar-terrestrial physics because 124 Europhys. News 26 (1995)