We report precise radial velocity measurements of the K giant i Dra (HD 137759, HR 5744, HIP 75458), carried out at Lick Observatory, which reveal the presence of a substellar companion orbiting the primary star. A Keplerian fit to the data yields an orbital period of about 536 days and an eccentricity of 0.70. Assum ing a mass of 1.05 M 8 for i Dra, the mass function implies a minimum companion mass m 2 sin i of 8.9 M J , making it a planet candidate. The corresponding semimajor axis is 1.3 AU. The nondetection of the orbital motion by Hipparcos allows us to place an upper limit of 45 M J on the companion mass, establishing the substellar nature of the object. We estimate that transits in this system could occur already for inclinations as low as 81=5, as a result of the large diameter of the giant star. The companion to i Dra is the first brown dwarf or planet found to orbit a giant rather than a main-sequence star.

Context. Future astrometric missions such as SIM PlanetQuest need very stable reference stars. K giants have large luminosities, which place them at large distances and thus the jitter of their photocenters by companions is relatively small. Therefore K giants would be best suited as references. To confirm this observationally a radial velocity survey is performed to quantify the level of intrinsic variability in K giants. Aims. From this radial velocity survey we present 34 K giants with an observed standard deviation of the radial velocity of less than 20 m/s. These stars are considered "stable" and can be used as radial velocity standards. Methods. The radial velocity survey contains 179 K giants. All K giants have a declination between −30• and +65• and visual magnitude of 3−6 mag. The Coudé Auxiliary Telescope (CAT) at UCO/Lick Observatory is used to obtain radial velocities with an accuracy of 5−8 m/s. The number of epochs for the 34 stable stars ranges from 11 to 28 with a total timespan of the observations between 1800 and a little over 2200 days.

Context. The Rossiter-McLaughlin (RM) effect, a rotational effect in eclipsing systems, provides unique insight into the relative orientation of stellar spin axes and orbital axes of eclipsing binary systems. Aims. Our aim is to develop a robust method to analyze the RM effect in an eclipsing system with two nearly equally bright components. This gives access to the orientation of the stellar rotation axes and may shed light on questions of binary formation and evolution. For example, a misalignment between the spin axes and the angular momentum of the system could bring the observed and theoretical apsidal motion into better agreement for some systems, including V1143 Cyg. Methods. High-resolution spectra have been obtained both out of eclipse and during the primary and secondary eclipses in the V1143 Cyg system, using the 0.6 m Coudé Auxiliary Telescope (CAT) and the high-resolution Hamilton Echelle Spectrograph at the Lick Observatory. The Rossiter-McLaughlin effect is analyzed in two ways: (1) by measuring the shift of the line center of gravity during different phases of the eclipses and (2) by analysis of the line shape change of the rotational broadening function during eclipses. Results. We measured the projection of the stellar rotation axes using the rotation effect for both main-sequence stars in an eclipsing binary system. The projected axes of both stars are aligned with the orbital spin within the observational uncertainties, with the angle of the primary rotation axis β p = 0.3 ± 1.5• , and the angle of the secondary rotation axis β s = −1.2 ± 1.6• , thereby showing that the remaining difference between the theoretical and observed apsidal motion for this system is not due to a misalignment of the stellar rotation axes. Both methods utilized in this paper work very well, even at times when the broadening profiles of the two stars overlap.

Context. Since 1999, we have been conducting a radial velocity survey of 179 K giants using the Coudé Auxiliary Telescope at UCO/Lick observatory. At present ∼20−100 measurements have been collected per star with a precision of 5 to 8 m s −1 . Of the stars monitored, 145 (80%) show radial velocity (RV) variations at a level >20 m s −1 , of which 43 exhibit significant periodicities. Aims. Our aim is to investigate possible mechanism(s) that cause these observed RV variations. We intend to test whether these variations are intrinsic in nature, or possibly induced by companions, or both. In addition, we aim to characterise the parameters of these companions. Methods. A relation between log g and the amplitude of the RV variations is investigated for all stars in the sample. Furthermore, the hypothesis that all periodic RV variations are caused by companions is investigated by comparing their inferred orbital statistics with the statistics of companions around main sequence F, G, and K dwarfs. Results. A strong relation is found between the amplitude of the RV variations and log g in K giant stars, as suggested earlier by Hatzes & Cochran (1998). However, most of the stars exhibiting periodic variations are located above this relation. These RV variations can be split in a periodic component which is not correlated with log g and a random residual part which does correlate with log g. Compared to main-sequence dwarf stars, K giants frequently exhibit periodic RV variations. Interpreting these RV variations as being caused by companions, the orbital parameters are different from the companions orbiting dwarfs. Conclusions. Intrinsic mechanisms play an important role in producing RV variations in K giants stars, as suggested by their dependence on log g. However, it appears that periodic RV variations are additional to these intrinsic variations, consistent with them being caused by companions. If indeed the majority of the periodic RV variations in K giants is interpreted as due to substellar companions, then massive planets are significantly more common around K giants than around F, G, K main-sequence stars.

It has long been speculated that the observed periodic radial velocity pattern for the K giant Pollux might be explained in terms of an orbiting planetary companion. We have collected 80 high-resolution spectra for Pollux at Lick Observatory yielding precise radial velocities with a mean error of 3.8 m/s, providing the most comprehensive and precise data set available for this star. Our data confirm the periodicity previously seen in the radial velocities. We derive a period of 589.7±3.5 days and, assuming a primary mass of 1.86 M ⊙ , a minimum companion mass of 2.9±0.3M Jup , consistent with earlier determinations. No evidence for any periodicities is visible in our analysis of the shapes of the spectral lines via the bisector method, so that we conclude that evidence is accumulating and compelling for a planet around Pollux. However, some last doubt remains about this interpretation, because non-radial pulsations which might be present in giant stars could in principle also explain the observed radial velocities, while the accompanying bisector variations might be too small to be detectable with current data.