Title:A Robust Test of Evolution near the Tip of the Red Giant Branch and Missing Giants in NGC 2808

Abstract: We describe a new method for robustly testing theoretical predictions of red giant evolution near the tip of the giant branch. When theoretical cumulative luminosity functions are shifted to align the tip in I-band and normalized at a luminosity level slightly brighter than the red giant bump, virtually all dependence on age and composition (heavy elements and helium abundance) is eliminated. While significant comparisons with observations require large samples of giant stars, such samples are available for some of the most massive Milky Way globular clusters. We present comparisons with the clusters NGC 2808 and M5, and find that NGC 2808 has a deficiency of bright giants (with a probability of less than about 3% that a more extreme distribution of giant stars would have happened by chance). We discuss the possibilities that underestimated neutrino losses or strong mass loss could be responsible for the deficit of giants. While we cannot rule out the neutrino hypothesis, it cannot explain the apparent agreement between the M5 observations and models. On the other hand, strong mass loss provides a potential link between the giant star observations and NGC 2808's unusually blue horizontal branch. If the mass loss hypothesis is true, there is likely a significant population of He white dwarfs that could be uncovered with slightly deeper UV observations of the cluster.