We propose that the abundance anomalies of proton-capture elements in globular clusters, such as the C-N, Na-O, Mg-Al and Na-F anti-correlations, were produced by super-massive stars with M ∼ 10 4 M ⊙ . Such stars could form in the runaway collisions of massive stars that sank to the cluster centre as a result of dynamical friction, or via the direct monolithic collapse of the low-metallicity gas cloud from which the cluster formed. To explain the observed abundance anomalies, we assume that the supermassive stars had lost significant parts of their initial masses when only a small mass fraction of hydrogen, ∆X ∼ 0.15, was transformed into helium. We speculate that the required mass loss might be caused by the super-Eddington radiation continuumdriven stellar wind or by the diffusive mode of the Jeans instability.

We critically examine the constraints on internal angular momentum transport which can be inferred from the spin down of open cluster stars. The rotation distribution inferred from rotation velocities and periods are consistent for larger and more recent samples, but smaller samples of rotation periods appear biased relative to v sin i studies. We therefore focus on whether the rotation period distributions observed in star forming regions can be evolved into the observed ones in the Pleiades, NGC 2516, M 34, M 35, M 37, and M 50 with plausible assumptions about star-disk coupling and angular momentum loss from magnetized solar-like winds. Solid body models are consistent with the data for low mass fully convective stars but highly inconsistent for higher mass stars where the surface convection zone can decouple for angular momentum purposes from the radiative interior. The Tayler-Spruit magnetic angular momentum transport mechanism, commonly employed in models of high mass stars, predicts solid-body rotation on extremely short timescales and is therefore unlikely to operate in solar-type pre-MS and MS stars at the predicted rate. Models with core-envelope decoupling can explain the spin down of 1.0 and 0.8 solar mass slow rotators with characteristic coupling timescales of 55 ± 25 Myr and 175 ± 25 Myr respectively. The upper envelope of the rotation distribution is more strongly coupled than the lower envelope of the rotation distribution, in accord with theoretical predictions that the angular momentum transport timescale should be shorter for more rapidly rotating stars. Constraints imposed by the solar rotation curve are also discussed. We argue that neither hydrodynamic mechanisms nor our revised and less efficient prescription for the Tayler-Spruit dynamo can reproduce both spin down and the internal solar rotation profile by themselves. It is likely that a successful model of angular momentum evolution will involve more than one mechanism. Further observational studies, especially of clusters younger than 100 Myr, will provide important additional constraints on the internal rotation of stars and could firmly rule out or confirm the operation of major classes of theoretical mechanisms.

No abstract

We show that mixing by weak turbulence induced by internal gravity waves generated by turbulent gas motions near the base of the convective envelope presents a good alternative to convective overshooting and rotation‐driven instabilities as a mechanism for partial mixing in thermally pulsing asymptotic giant branch (AGB) stars. Such mixing ingests protons into the He/C‐rich zone which leads to formation of the 13C pocket, a necessary ingredient of the current theory of the s‐process under radiative conditions in low‐mass AGB stars. We demonstrate that with appropriately chosen, physically reasonable parameters partial mixing by internal gravity waves may produce a much wider 13C pocket than convective overshooting limited to an extent compatible with observations.