1995A&A...297..345C

The problem of the age difference between the classical second-parameter clusters M 3 and M 13 is revisited, using the "{DELTA}V" and the horizontal-branch (HB) morphology methods of age derivation. Our new application of the {DELTA}V method makes use of the colour-magnitude diagram recently obtained by Guarnieri et al. (1993A&AS..102..397G) for M 13. Besides, the ambiguity in the determination of the HB level for M 13, due to a scarcely populated instability strip, is minorated through computation of detailed synthetic HB models. The method suggests that the two clusters do not have a large age difference, with an upper limit of {DELTA}t(M 13-M 3)≃3Gyr. This result, which agrees with the independent analysis of VandenBerg et al. (1990AJ....100..445V), is shown to be in conflict with the "age-only" interpretation of the second-parameter phenomenon, given that larger age differences appear necessary to match the observed HB morphologies, if age is the second parameter in the halo. Evidence is presented that the super oxygen-poor stars found by Kraft et al. (1992AJ....104..645K, 1993AJ....106.1490K) and others to be present in the red giant branch of M 13 (blue HB morphology), but lacking in M 3 (intermediate HB morphology), may be related to an increased mass loss rate on the red giant branch, which could be an additional parameter affecting the HB morphology.