Thesis (M.S.)--University of Hawaii at Manoa, 2004. / Includes bibliographical references (leaves 159-185). Also available via World Wide Web.

The Harrington Peak Quadrangle is located within the Caribou National Forest of southeast Idaho. Within this quadrangle are outcrops of olivine trachyte of Pliocene(?) age overlying sedimentary rocks of Mississippian to Tertiary age. The region contains thrust faults and later normal faults (generally trending north-south} formed during Basin and Range extension. The Largest outcrop of olivine trachyte (approximately1 1/2 X 3 km) probably formed as the result of a fissure eruption. Two other outcrop areas show evidence of being sites of local extrusion. Whole-rock chemical analyses revealed the olivine trachyte to have moderate amounts of SiO2 and Al2O3, high MgO and CaO, and K2O in excess over Na2O (approximately 2:1). Mineralogical characteristics include microphenocrysts of Mg-rich olivine and diopsidic augite in a groundmass of Ba-rich sanidine, diopsidic augite, Fe-Ti oxides, and less commonly phlogopite and/or plagioclase. The olivine trachyte closely resembles the ciminites from the Viterbo region of Italy and has some petrological and mineralogical similarities to many other continental potassic volcanic rocks. The olivine trachyte may have formed by partial melting of a heterogenous mica peridotite mantle source enriched in incompatible elements during a previous tectonic event.

The origin of compositional gaps in volcanic deposits remains controversial. In Akaroa Volcano (9.6 to 8.6 Ma), New Zealand, a dramatic compositional gap exists between basaltic and trachytic magmas. Previously, the formation of more evolved magmas has been ascribed to crustal melting. However, the interpretation of new major and trace element analysis from minerals and bulk-rocks coupled with the mechanics of crystal-liquid separation offers an alternative explanation that alleviates the thermal restrictions required for crustal melting models. In a two-stage model, major and trace element trends can be reproduced by polybaric crystal fractionation from dry melts (less than 0.5 wt.% H2O) at the QFM buffer. In the first stage, picritic basalts are separated from an olivine-pyroxene dominant mush near the crust-mantle boundary (9 to 10 kbar). Ascending magmas stagnated at mid-crustal levels (5 to 6 kbar) and fractionated an olivine-plagioclase assemblage to produce the alkali basalt-hawaiite trend. In the second stage, trachyte melt is extracted from a crystal mush of hawaiite to mugearite composition at mid-to-upper crustal levels (3 to 5 kbar) after the melt has crystallised 50 vol.%. The fractionated assemblage of plagioclase, olivine, clinopyroxene, magnetite, and apatite is left in a cumulate residue which corresponds to the mineral assemblage of sampled ultramafic enclaves. The results of trace element modelling of Rayleigh fractionation using this extraction window is in close agreement with the concentrations measured in trachyte (= liquid) and enclaves (= cumulate residue). The compositional gap observed in the bulk-rock data of eruptive products is not recorded in the feldspar data, which show a complete solid solution from basalt and co-magmatic enclaves to trachyte. Complexly zoned plagioclases further suggest episodical magma recharge events of hotter, more mafic magmas, which lead to vigorous convection and magma mixing. In summary, these models indicate that the Daly Gap of Akaroa Volcano formed by punctuated melt extraction from a crystal mush at the brittle-ductile transition.