By analogy with a type II superconductor placed in a magnetic field (see section C.I.1), de Gennes, in 1972, put forward the idea that a smectic composed of chiral molecules can twist by developing screw dislocations [1]. In this model (see appendix 1 to chapter C.I), the twist of the director field is localized in the core of each dislocation and in its immediate vicinity, over a distance λ termed penetration length: in this region, the molecules are tilted with respect to the layer normal, as in a smectic C.