“…Moreover, the TS distribution, being a generalisation of the canonical BG, has a non-compact support, and thus cannot be valid close to x ≈ 1, where it should have a cut due to the finiteness of the collision energy. Therefore, in order to achieve good agreement with data for large x, microcanonical effects, due to the finite available phase space, have to be taken into account, as it has also been proposed in [24]- [29]. In this paper, a possible microcanonical generalisation of the TS distribution is proposed, which gives a good fit to data on fragmentation functions measured in e + e − collisions for 0.01 x 1.…”