Spectroscopic and photometric evidences have led to a complete revision of our understanding of globular clusters with the discovery of multiple stellar populations which differ chemically. Whereas some stars have a chemical composition similar to fields stars, others show large star-to-star variations in light elements (Li to Al) while their composition in iron and heavy elements stay constant. This peculiar chemical pattern can be explained by self-pollution of the intracluster gas occurring in the early evolution of clusters. Here the possible impact from a first generation of fast rotating stars to the early evolution of globular clusters is presented. The high rotation velocity will allow the stars to rotate at the break-up velocity and release matter enrich in H-burning which in turn will produce new stars with a chemical composition in agreement with observations. The massive stars have also an important role to clear the cluster from the remaining gas left after the star formation episodes. If the gas expulsion is fast enough, the strong change in the potential well will lead to the loss of stars occupying the outer part of the cluster. As second generation stars are preferentially born in the cluster centre, the ratio of second to first generation stars will increase over time to match the present ratio determined by observations. Considerations on the properties of low-mass stars still present in globular clusters will also be presented.

Recent studies have claimed the existence of very massive stars (VMS) up to 300 M⊙ in the local Universe. As this finding may represent a paradigm shift for the canonical stellar upper-mass limit of 150 M⊙, it is timely to discuss the status of the data, as well as the far-reaching implications of such objects. We held a Joint Discussion at the General Assembly in Beijing to discuss (i) the determination of the current masses of the most massive stars, (ii) the formation of VMS, (iii) their mass loss, and (iv) their evolution and final fate. The prime aim was to reach broad consensus between observers and theorists on how to identify and quantify the dominant physical processes.

The first results of new evolutionary calculations of massive stars are presented. The calculations cover ZAMS masses from 20—200 M⊙. Yields and abundances throughout the stars are presented.