Topology in QCD

and this suggests that alot of the topological charge density tends to cluster in lumps that are uncorrelated at larger distances – just like a ‘gas’ of instantons. So, for convenience, this is the language I shall use in this talk. I will start with the aspect that is no doubt on the firmest theoretical ground: the large mass of the η [1] and the value [2] of the topological susceptibility χt ≡ 〈Q2〉/V . I dwell on the quenched case because it provides one area in which the lattice calculations are in very good shape, even if there is not much new this year (probably because they are in such good shape ...). I will then move onto the possible role of instantons in driving the spontaneous breaking of chiral symmetry [3–5]. Here the exciting news is that we have finally got lattice fermions that are good enough to address this question realistically. More speculative is the influence of topology on hadrons [5]. This provides a motivation for trying to extract the properties of the ‘gas’ of instantons in the vacuum. There has been some interesting new work on the latter during the past year. Finally I turn to the role of instantons in confinement. The usual view is that there is no such role. I will discuss a recent calculation that claims the opposite. This will be the one area where fermions don’t appear at all.