Geological data permit interpretation of the Northern Belt of the Southern Uplands as an Ordovician forearc basin in a northwest-facing arc. This collided with Laurentia in the latest Ordovician, reversed, and collided with Cadomia in the early Devonian.
Since the earliest plate tectonic model for the British Caledonides (Dewey 1969) and the interpretation of the Southern Uplands as ocean floor and trench sediments imbricated at a southwest-facing arc (Mitchell 1974), the concept of northwestward subduction beneath the Northern Belt has been more or less undisputed (Mitchell & McKerrow 1975; Leggett et al. 1979; Leggett 1987; McKerrow 1988; McKerrow & Soper 1989). However, the more recent geological, as opposed to palaeontological, data (McKerrow 1987, and references therein) are more compatible with interpretation of the Northern Belt as a forearc basin formed above southeastward-subducting lithosphere.
In the proposed model (Fig. l), by the Llanvirn or Llandeilo, subduction of an oceanic basin, Iapetus 1, had begun beneath an are system to the southeast (Fig. 1A). There was no major accretionary wedge and southeast of the forearc ridge, in the Northern Belt forearc basin, diachronous deposition of Moffat Shale was followed by greywacke sedimentation and local eruption of basaltic lavas. Until the Caradoc the associated ‘Cockburnland’ volcanic arc lay immediately to the south of the Northern Belt (Morris 1987; Stone et al. 1987; Hutton & Murphy 1987) and possibly included the Lake District on the northern margin of Cadomia. The northwest flank of the volcanic arc supplied andesitic detritus to the forearc basin (Morris