Abstract
Within a largely concealed, caldera-related volcaniclastic succession of the Ordovician Borrowdale Volcanic Group in the western Lake District, two thick (100–350 m) ignimbrites within the Fleming Hall Formation exhibit a number of features that in combination make them unusual deposits. They are both homogeneous with comparatively low-SiO2 (63%) bulk composition, contain only a moderate crystal content, are generally poor in lithic clasts, show uniformly very dense welding (yielding parataxitic to massive vitrophyric texture) throughout and lack associated fall-out or surge deposits. Ignimbrites of comparable bulk composition in this geological setting are usually part of zoned sheets and/or frequently very crystal-rich. Large-scale, unzoned densely welded ignimbrites are usually rhyodacitic to rhyolitic. By contrast, ignimbrites of intermediate composition that display dense welding are relatively small deposits that form by agglutination of hot, plastic spatter.

It is postulated that the Fleming Hall ignimbrites were derived from low column height, low explosivity eruptions that conserved heat and minimized entrainment of accidental lithic clasts and the formation of fine ash. The very dense welding and lack of bubble-wall shard vitroclastic textures indicate that pyroclasts were hot and relatively dry, probably occurring as mildly vesicular (scoriaceous) fragments which welded or fused together during aggradational deposition rather than by post-depositional compactional loading. There is little variation in the degree of matrix or melt crystallization throughout the two ignimbrites, despite the fact that high temperatures must have been maintained for many years following deposition. Both display virtually ubiquitous development of micropoikilitic glass devitrification texture, which suggests that the viscosity of the supercooled dacitic melt was sufficiently high, probably due to initial degassing, to inhibit significant melt crystallization after deposition.

The eruption of the Fleming Hall magmas was probably initiated by the rise or injection of hotter, more basic, magma, and not by overpressurization due to volatile exsolution resulting from cooling and crystallization. Foundering of the chamber roof caused forcible and rapid eruption of the magma, probably along a series of volcanotectonic faults rather than a central vent, and probably flooded the resultant caldera depression. It is predicted that this type of eruption will not have produced a widely dispersed deposit, the bulk of which may have been largely contained within its own caldera.