[1] Palagonitized sideromelane from submarine volcaniclastic, seafloor volcanic, marine phreatomagmatic, lacustine phreatomagmatic, and subglacial volcanic settings was investigated using in situ microanalysis to test if palagonite composition and texture are related to depositional environment. Palagonitization extent varies linearly and inversely with original sample porosity, suggesting that porosity is a controlling factor of palagonitization. Water absorbance of reflected infrared light varies linearly with water content derived from electron microprobe totals. Palagonite water content has a linear, inverse relationship to palagonitization extent. REEs are immobile during palagonitization, so they can be used to construct isocon diagrams for estimating major-element concentration changes. Major-element and overall mass changes during palagonitization vary widely (particularly for FeO and TiO 2 ) and indicate that palagonitization cannot be an isovolumetric process. These parameters depend strongly on original sideromelane composition, thus requiring composition to be taken into account when performing global oceanic cation flux calculations. Subalkaline sideromelane dissolves much more rapidly than alkaline sideromelane during palagonitization. Two styles of palagonitization, burial-diagenesis (relatively long-duration, low water/rock; passive fluid circulation) and hydrothermal (relatively short-duration, high water/rock; hydrothermal fluid circulation), are recognized. Observed palagonite REE concentration gradients indicate that sideromelane dissolution must continue in the zone behind the advancing palagonitization front. MgO was found to be highly mobile during palagonitization. Observed palagonite MgO gradients are not developed during sideromelane dissolution, but instead record initiation of syn-and/or post-palagonitization conversion of the gel-palagonite layer to a phyllosillicate layer, consistent with evolution of sideromelane alteration layers toward equilibrium with the solution.

Internal and eruptive volcano processes involve complex interactions of multi‐phase fluids with the solid Earth and the atmosphere, and produce diverse geochemical, visible, thermal, elastic, and anelastic effects. Multidisciplinary experimental agendas are increasingly being employed to meet the challenge of understanding active volcanoes and their hazards [e.g.,Ripepe et al., 2002; Wallace et al., 2003]. Mount Erebus is a large (3794 m) stratovolcano that forms the centerpiece of Ross Island, Antarctica, the site of the principal US. (McMurdo) and New Zealand (Scott) Antarctic bases. With an elevation of 3794 m and a volume of ∼1670 km3, Erebus offers exceptional opportunities for extended study of volcano processes because of its persistent, low‐level, strombolian activity (Volcano Explosivity Index 0–1) and exposed summit magma reservoir (manifested as a long‐lived phonolitic lava lake). Key scientific questions include linking conduit processes to near‐field deformations [e.g., Aster et al., 2003] ,explosion physics [ e.g., Johnson et al., 2003], magmatic differentiation and residence [e.g., Kyle et al., 1992], and effects on Antarctic atmospheric and ice geochemistry [e.g., Zreda‐Gostynska et al., 1997]. The close proximity of Erebus (35 km) to McMurdo, and its characteristic dry, windy, cold, and high‐elevation Antarctic environment, make the volcano a convenient test bed for the general development of volcano surveillance and other instrumentation under extreme conditions.

Abstract-Boron has been shown to be a useful trace element in clay-mineralization reactions, raising the possibility that B studies may provide a means to investigate environmental controls on palagonitization. The objective of the present study was to address calibration, matrix effects, and B exchangeability issues such that meaningful secondary ion mass spectrometry (SIMS) microanalysis of B in thin sections of palagonite will be feasible. Silver Hill illite (IMt-1) was found to be a suitable calibration reference material, based on compositional similarity, relatively high B content, and ease of mounting on thinsection samples for SIMS microanalysis. Matrix effects of borated sideromelane and illite were compared and found to be similar, confirming previous studies which showed no matrix effects for B among minerals. Boron substitutes for Si in tetrahedral sites and also can be adsorbed in exchangeable sites of 2:1 clay minerals. Similarly, B can be found in tetrahedral and exchangeable sites within palagonite, which consists of both layered and amorphous volumes. In order to measure tetrahedral B content and isotopic ratio in the palagonite, exchangeable B was removed by soaking sample thin sections in a 1 M NH 4 Cl solution until exchangeable cation concentrations were constant. Treated samples showed decreases in B content and isotopic ratio with exchange. Extraction of exchangeable B permits the direct measurement of tetrahedral B content and isotopic ratio. The exchange technique devised and tested here should have broad applicability to thin-section microanalysis of B in clay and clay-like materials where cation exchange can be used for surface-analytical techniques. The present study represents an initial attempt to address samplepreparation, calibration, and potential matrix-effects problems for analyses by SIMS. Further refinements may improve the accuracy of the measurements, but the results presented here indicate that meaningful measurements are possible.