Artie P. Hatzes, Volker Perdelwitz, Marie Karjalainen, Jana Köhler, Michael Hartmann

Published radial velocity (RV) measurements of the K giant star 42 Dra reveal variations consistent with a 3.9 M_Jup mass companion in a 479-d orbit. This exoplanet can be confirmed if these variations are long-lived and coherent. Continued monitoring may also reveal other companions. We have acquired additional RV measurements of 42 Dra spanning fifteen years. Periodogram analyses were used to investigate the stability of the planet RV signal. We also investigated variations in the spectral line shapes using the bisector velocity span as well as infrared photometry from the COBE mission. The new RV measurements do not follow the published planet orbit. An orbital solution using the 2004 - 2011 data yields a period and eccentricity consistent with the published values, but the RV amplitude has decreased by a factor of four from the earlier measurements. Including some additional RV measurements taken between 2014 and 2018 reveal the presence of a second period at 530 d. The beating of this period with the one at 479-d may account for the observed amplitude variations. The planet hypothesis is conclusively ruled out by COBE/DIRBE 1.25 micron photometry that shows variations with the planet orbital period as well as an additional 170 d period. The amplitude variations in the RV as well the COBE/DIRBE photometry firmly establish that there is no giant planet around 42 Dra. The presence of multi-periodic variations suggests that these may be stellar oscillations, most likely oscillatory convection modes. These oscillations may account for some of the long period RV variations attributed to planets around K giant stars. This may skew the statistics of planet occurrence around intermediate mass stars. Long-term monitoring with excellent sampling is required to exclude amplitude variations in the long-periods found in radial velocity of K giant stars.