“…The condition (19) implies that, e.g., in optically thin crystals (d < 1, ) a* < 1 (Fig. 2) and the ratio /,/la should indeed be very small for the stable kinks to exist [l, 71, while in thick crystals a* 2 1 and stable kinks are likely to be present even at large lD/l,, although it is not very clear what the term "kink" means in this latter case [9]. The exactly solvable step function model enabled us to investigate the behaviour of kinks in a bounded crystal and to pin-point mechanisms which determine their stability.…”