Abstract

We revisit the "second-parameter" pair of globular clusters NGC 288/362 on the basis of theoretical models for red giant branch (RGB) and horizontal branch (HB) stars. The results of the most extensive set of RGB/HB simulations computed so far for these clusters are presented for two different metallicities. Using several different analytical mass-loss formulae for RGB stars, we derive relative "HB morphology ages." We compare them with the relative main-sequence turnoff ages derived by application of the "bridge test" in Paper I, where it was found that NGC 288 is 2 ± 1 Gyr older than NGC 362. We find that adoption of a higher metallicity ([Fe/H] ? -1.2), as favored by the Carretta & Gratton metallicity scale, makes age a much more plausible second-parameter candidate for this pair than is the case when a lower metallicity ([Fe/H] ? -1.5), closer to the Zinn & West scale, is adopted. However, while the different HB morphology of these two clusters can be reproduced by canonical HB models with [Fe/H] ? -1.2 and an age difference of 2 Gyr, this explanation is not without difficulty. In particular, we confirm previous suggestions that canonical models are unable to reproduce the detailed HB morphology of NGC 288 at its red end, for as yet unknown reasons. Finally, we show that the mass dispersion on the HB of NGC 362 is substantially larger than for NGC 288 and suggest that there is a correlation between the mass dispersion on the HB phase and the central density of globular clusters. This is presumably related to the way environmental effects affect RGB mass loss - another major second-parameter candidate. We argue that, if confirmed, this central density-HB mass dispersion correlation will have to be taken into account in order to conclusively determine whether age may be considered the (sole) second parameter of HB morphology for this (and other) second-parameter pair(s).