Conal twisting seems to result from dissociation between the cardiac jelly and the deep myocardial interface like the lining in the sleeve of a jacket. The conal ridges are the natural markers of the jelly and endocardium. They are responsible for septation of the conus and enable the twisting to be observed and measured. The myocardium is marked artificially by cauterisation. This marks the armature of the wall and leaves behind a zone of reduced resistance in the form of a hernia or false diverticulum. The conal ridges and marked myocardium dissociate. In the mid segment the myocardium does not play any role in twisting. In the proximal and distal segments it is only partially involved. This dissociation is observed even in the structure of the conal wall; the jelly, which is dense near the endocardium, is loose near the myocardium and adheres to the deep surface by dispersed fibres. Perfusion under pressure of the investigated specimens induces a detachment between the jelly and the myocardium and there only. This fragility only lasts during the twisting period. It is not found at the end of cardiac embryogenesis. This zone would allow not only a sliding--due to its fragility--but also a controlled sliding--by its fibres. In addition to twisting there is also conal migration. This takes place in the same direction as proximal twisting and determines myocardial rotation, which is less marked however, than that of the corresponding ridges. Experimentation may exaggerate this dissociation by preventing migration. It may also reduce or even suppress it by the formation of adhesions between the two layers. Although this mechanism is not univocal, a myocardial-jelly adhesion could stop distal twisting for example and explain malposition or transposition of the great vessels. This dissociation is no unique to superior vertebrates as it has also been found in the first living animals in whom conal twisting occurs, the dipneustes.