Discussion

Abstract

Dr G. P. B lack referred to the author’s interest in the structure of the ring-complex. There had been an unfortunate tendency for petrologists to class all masses of igneous rock that had even the merest semblance of an arcuate outcrop as ring-dykes without considering the need to apply further criteria. The literature of ring-complexes showed that the rock masses described as ring-dykes fell readily into two contrasted types. Characteristics generally shown by the first type included narrow, annular, or highly arcuate outcrops, contacts everywhere with older rocks, and steeply inclined or vertical junctions whose attitude could be determined with confidence. Such masses seemed invariably to be acidic in composition. The Loch Ba felsite of Mull—the ‘holotype’ of all ring-dykes—belonged to this type, as did many examples described from Nigeria. Members of the second type formed a less well-defined group but all showed a majority of a number of features, including crescentic or irregular outcrops often of considerable breadth, contacts with younger intrusions on one or more margins so that the original form of outcrop could not be determined, and junctions that were so poorly denned or so poorly exposed that their attitude could not be ascertained. The members of this type were generally of basic composition. The best assemblage of examples came from the Ardnamurchan complex where almost all of the thirty-two ‘ring-dykes’ listed in the Geological Survey memoir (Richey & others 1930, pp. 201–2) fell into this type. Even the Great Eucrite, despite its annular outcrop,