The relationship between K, the level of potash content in lavas erupted from an andesite volcano, and h, the depth to the center of the inclined seismic (Benioff) zone beneath the volcano, is extended using new data from Indonesia, the Lesser Antilles, and New Zealand. Coefficients of correlation between K and h of +0.86 (K55) and +0.80 (K60) are obtained by straight‐line regression. This is compared with a coefficient between K55 and distance d of the volcano from the trench of only +0.60. The coefficients of correlation of K against hmin, which is the depth to the upper surface of the Benioff zone, are not significantly different from those between K and h.

Circum-Pacific andesites, with associated basalts and dacites, are erupted from linear island arcs and marginal continental ranges whose ocean-side borders coincide approximately with the continent-side boundaries of belts of shallow seismicity that parallel adjacent trenches. The lines of active volcanoes stand above elongate subcrustal regions delineated by the intersection of inclined Benioff seismic zones and the subhorizontal Gutenberg low-velocity zone. Close correlation between content of potash in erupted lavas and vertical depth to the Benioff zone suggests that andesitic volcanism has its origins in the mantle where magmas are generated by events associated with earthquakes of intermediate focal depth.

No abstract

Volcanism, heat flow, gravity, morphology, and tectonism in the North Island of New Zealand are examined in the light of seismic evidence for inhomogeneity in the upper mantle. The high‐frequency attenuating/transmitting boundary underlies the junction between volcanic and nonvolcanic provinces, and the now active andesite volcanoes lie above this boundary. Heat flow is higher than normal over the attenuating zone and lower than normal over the transmitting zone. Gravity anomalies are positive over the attenuating zone and negative over the transmitting zone; a density model is proposed in which the density variations are envelopes of seismicity in the upper mantle. The Hikurangi trench curves in a similar manner to the attenuating/transmitting boundary and shoals in a similar manner to the seismicity. Similar data are examined for northern Japan, and there is good agreement between the two areas. The special importance of New Zealand in testing the Oliver‐Isacks model of an active margin is demonstrated.

No abstract