Abstract Mafic lamprophyre dykes in northern England were emplaced soon after the final closure of the Iapetus Ocean at the end of the Caledonian orogeny. High Mg/(Mg+Fe) ratios and Cr and Ni contents are consistent with equilibrium with mantle peridotite. Incompatible trace element abundances suggest that the mantle sources were metasomatized prior to the melting events. Three components are recognized in the lamprophyre chemistry: (i) a depleted mantle source, taken to be that overlying the subducting lithosphere; (ii) a H 2 O-rich subduction zone component, related to the dehydration of the subducting oceanic crust; and (iii) a CO 2 -rich phase thought to result from the degassing of the mantle after ocean closure. This multi-component origin may be applicable to other lamprophyres of calc-alkaline affinity.
There are three distinct zones of recent volcanic activity in the Andean Cordillera: a northern zone in South Colombia and Ecuador, a central zone in South Peru and North Chile, and a southern zone in South Chile (Figure 1). Cenozoic intrusive rocks occur between these areas, and on the coastal margin of the active zones. It is therefore natural to suppose that the igneous processes that take place along the Andean plate margin have simultaneous volcanic and plutonic components and, consequently, that the two kinds of magmatic activity are interrelated. This supposition is expressed in countless diagrams and cross-sections purporting to illustrate destructive plate margin processes.
The Tertiary lava succession of eastern Iceland dips at shallow angles towards and beneath the axial Quaternary volcanic zone. A 3‐km vertical section through part of this Tertiary pile, studied during the Iceland Research Drilling Project, contains flows ranging in composition from basalt to icelandite. This succession can be divided into three distinct stratigraphic groups: upper and lower groups, 1.7 and 0.6 km thick with Zr/Y ratios of about 4.5, and a 0.7‐km‐thick middle group with a Zr/Y ratio of about 6.5. The middle group also has a significantly higher Ce/Yb ratio and higher elemental Sr abundances. The data on within‐group variation are equally compatible with high‐level fractionation of a plagioclase‐olivine‐clinopyroxene assemblage with accessory apatite and opaque minerals or with open system fractionation models. Dikes occur throughout the 3‐km section and are comparable in trace element composition to the lavas of the upper group. The dikes are part of a N‐S trending swarm originating from the Breiddalur volcanic center some 15 km to the south. These dikes were injected laterally, northward into the Reydarfjordur lava pile at the time of the eruption of the upper group lavas. The lower two major lava groups may be related to similar, but unidentified, discrete volcanic centers perhaps buried downdip. The new data on this deeper section corroborate earlier studies which indicated that on a gross scale, the eastern Iceland lava pile becomes less mafic with depth. This vertical change occurs because each major volcanic center erupted more silicic lavas close to the center and more magnesium‐rich compositions toward the periphery. The formation of the lava pile occurred by tilting toward the spreading center and overlapping repetition of such major units which resulted in the more mafic edges of the lava groups forming the topographically higher parts of the crustal section.
The igneous rocks of the British Tertiary Volcanic Province (BTVP) comprise intrusive central complexes and associated lava fields in northwest Scotland and northern Ireland. These centres are associated with linear dyke swarms which are radial around the major central complexes. The most extensive dyke swarm is related to the Mull intrusive complex and includes the Cleveland dyke, which appears to extend some 430 km from Mull through the Scottish Midland Valley (SMV) to the coast of northeast England. The dyke may have been emplaced by lateral magma migration from Mull, by vertical magma migration, or by a combination of these processes associated with the emplacement of the Mull centre and the presence of a regional stress field in northern Britain. Petrographic, mineralogical, and geochemical data for samples collected across and along the Cleveland dyke have been used to evaluate its petrogenesis and emplacement mechanism. The segment of the dyke north of, and along, the Southern Uplands Fault, the southern boundary of the SMV, is not comagmatic with that to the south, which is now defined as the Cleveland dyke sensu stricto. The Cleveland dyke is an olivine-free, plagioclase- and pyroxene-phyric basaltic andesite. Plagioclase mineralogy and bulk composition indicate that it experienced a complex magmatic history involving polybaric fractional crystallization and minor crustal contamination. Despite this complex evolution, the dyke magma is relatively homogeneous and shows chemical characteristics closely similar to tholeiitic rocks from Mull. The data substantiate lateral emplacement from this BVTP centre, rather than by vertical emplacement through heterogeneous lithosphere. Numerical modelling of dyke dynamics is consistent with emplacement of the Cleveland dyke as a single pulse of magma from the Mull centre, flowing in a manner transitional between laminar and turbulent conditions. According to this model, the dyke (volume c. 85 km3 was initiated in a large magma chamber below Mull subject to a small excess magmatic pressure. Lateral migration at relatively high velocity (1–5 ms−1) caused emplacement of the dyke in 1–5 days. Following emplacement, minor vertical ascent of magma may have contributed to the local en echelon distribution of dyke segments.
The Mona Complex comprises a 'Bedded Succession' of late Proterozoic low-grade metasedimentary schists, a group of gneisses of uncertain age, and the Coedana Granite. New Rb-Sr whole-rock isochron data for gneisses from the Holland Arms area indicate a late Precambrian metamorphic episode at 595 ± 12 Ma and a possible Cambrian igneous episode represented by an age with relatively large errors of 562 ± 31 Ma obtained from orthogneisses. A Rb-Sr whole-rock isochron age for the Coedana granite is 603 ± 34 Ma. The relatively low initial 87 Sr/ 86 Sr ratio for the late Precambrian gneisses, 0.7061 ± 3, is considered to preclude a long crustal history for these rocks and they are identified as metamorphosed equivalents of part of the Bedded Succession. The initial 87 Sr/ 86 Sr ratios of the orthogneisses and the Coedana Granite (0.7081 ± 8 and 0.7086 ± 9, respectively) are identical within analytical error and are consistent with an origin for these igneous rocks either by crustal contamination of mantle-derived melts, or by melting of young crustal material such as the older gneisses.
Abstract Millstones used in Cyprus between the Late Bronze Age and Roman periods (saddle querns, reciprocally operated hopper‐rubber mills, cylindrical rotary querns, and hour‐glass‐shaped Pompeian style mills) were frequently manufactured from igneous rocks including vesicular lavas clearly imported to the island. Potential sources include volcanics of mainland Greece, the Aegean, Anatolia, Egypt, and the Levant (Syria, Israel, and Jordan). Geochemical analysis using X‐ray fluorescence of thirty‐seven Cypriot millstones showed that sources were used as follows: Late Bronze Age saddle querns were made of Levant lavas and of local Troodos rocks, the later hopper‐rubbers were imported from the Aegean island of Nisyros, a rotary mill was of Santorini lava, and the Roman Pompeian style mills were manufactured from Levant lavas, including sources in north Syria and probably around Lake Tiberias. Millstones were probably exported from the Levant ports of Akko, Tell Abu Hawam, and Laodicea, and imported to the Cypriot towns of Salamis and Nea Paphos.