Non-radial oscillation in the solar-temperature star 51 Pegasi

We present results of high-resolution spectroscopic observations of the solar-temperature star 51 Peg taken between 1989 and 1996. Variations in the shape of the Fe I lambda 6252.53 Fe I spectral line are investigated in detail to establish their reality, nature, and likely cause. Because our spectroscopic data were gathered for purposes other than the present one, they are thinly distributed over the 7 years. This makes it difficult for us to prove beyond doubt that the variations in the shapes of the profiles are one and the same as the 4.23 day period of radial velocity variations found by Mayor & Queloz and Marcy et al. Nevertheless, we show that the probability of our data matching the periodicity of the radial velocity data the way it does by pure chance is only one in several hundred. Since the probability strongly favors the reality of the 4.23 day profile shape variations, we proceed to model them with nonradial oscillations having low order and low degree. The shifts and distortions of spectral lines induced by oscillations having l = -m = 4 fully account for both the radial velocity observations and the changes in line profiles delineated by our high-resolution spectroscopy. The planet hypothesis, proposed in the above mentioned papers, cannot account for implicit variations of the spectral line profiles. Assuming these variations are real, the planet hypothesis is no longer viable, and the need to explain the unseen and puzzling planet in an unusual orbit no longer exists. Instead, the door may have been opened to the important new area of research: low-order nonradial pulsation in solar-temperature stars. In the Appendix we refute various suggestions by which the originally proposed planet might induce the line-profile variations.