The mullions shown in most modern structural geology textbooks (Hobbs et al., 1976; Ramsay and Huber, 1987; Pilger and Schmidt, 1957) are in many respects similar to structures found in nearby outcrops, which started their life in structural geology as the type example of boudins (Lohest et al., 1909).

The reason for this confusing evolution in nomenclature is the inconsistent, multilingual usage of the initial morphological definition of the term boudin, which was based on a striking resemblance to the long, parallel strands in which blood-sausages are usually displayed in S-Belgium (Wilson, 1982) (Fig. 1). This nomenclature was originally strictly non-genetic, and the term boudinage was introduced for the (as yet unknown) process of formation of these structures (Lohest et al., 1909). Later usage of the term boudin became genetic, inspired by the extension during preparation of most other sausages and corresponding rock structures. The term boudin is now well established for structures formed by layer parallel extension of a competent layer in a less competent matrix, with veins or matrix emplaced in the boudin necks. In the second half of the last century some studies of the locus typicus for boudins at Bastogne concluded that the major part of the evolution of these structures was a shortening sub-parallel to bedding, instead of just extension. To solve the conflict in nomenclature the names K-boudin (K for the German word Kompression) and L-boudin (L for the German word Längung) were proposed (Bruhl, 1969). However, these terms never found general acceptance. After a "visit of English colleagues to Dedenborn" the name mullion was introduced (Pilger and Schmidt, 1957) and photographs of these structures found their way into the structural geology textbooks (Hobbs et al., 1976; Ramsay, 1967). Here, mullions are described as regular cuspate-lobate folds of an interface between two materials with a large competence contrast, the cusps poining towards the more competent material. This was interpreted (Ramsay, 1967) as being a compressional feature grown from an instability at the deforming interface.

Figure 1. The classic 'boudin' locality

(a) Morphology of the base of a psammite layer near Rouette, showing the characteristic cuspate-lobate structures of mullions. Width of picture is approximately 1 m. (b) Boudins in a shop near Bastogne, showing the characteristic morphology of sausages in this part of Belgium. Note the absence of pinch and swell structures along the sausage. Width of picture is approximately 20 cm. (Click for enlargement)

In other studies of the Bastogne boudins it was proposed that these structures were not initiated during a pronounced layer-parallel extension event and are therefore not boudins in the usual (extensional) sense, perhaps modified during later layer parallel shortening (Jongmans and Cosgrove, 1994; Rondeel and Voermans, 1975), although there is no general agreement.

In the last decade of the twentieth century it was re-emphasized (Spaeth, 1986) that all the cylindrical cuspate-lobate structures in the areas (on both sides of the border between Germany and Belgium) belong to the same class as the original boudins at Bastogne.

In this paper, to keep nomenclature consistent with the present genetic usage and to avoid further confusion, we call all these compressional structures mullions (Kenis et al.,, 2001) although this is inconsistent with the original definition and is not yet generally accepted. In future works, careful documentation of terminology will be required, because the term mullion is also not always clearly and consistently defined.