Verhakungen, dislocations, solitons, and kinksThe paper retraces, from a personal viewpoint, the development of atomistic models of dislocations in crystals from the model of Prandtl, Dehlinger, Frenkel, and Kontorova, first conceived in 1912, to recent work on kinks in dislocations. Among the topics discussed in some detail are the emergence of the theory of solitons, the rate theory of kink-pair generation, and the interplay of experiment and theory in the quantitative investigation of kinks by mechanical relaxation and flow-stress measurements. An outcome of this interplay, the determination of the planes of the elementary slip steps of screw dislocations in refractory body-centred cubic metals, has become the 'open sesame!' of quite a number of puzzling phenomena in which the plastic deformation of bcc metals differs from that of fcc metals and alloys. These phenomena include non-uniformities in the flow-stress -temperature relationship, anomalous slip, tension -compression asymmetry in uniaxial straining tests, alloy softening, and the enigma of 'reversible' vs. 'irreversible' c-relaxation. The key for their understanding is a first-order transition transforming the cores of a 0 <111>/2 screw dislocations from a low-temperature configuration capable of slipping on {110} planes to a configuration with a {112} slip plane at elevated temperatures. Various problems needing further research are pointed out.