A bstractThe Keweenawan intrusive rocks of the Lake Superior region may be conveniently divided into alkaline intrusions, layered tholeiitic intrusions, and quartz and olivine tholeiite dike and sill swarms. Within each group is a characteristic internal stratigraphic succession of rock types. The alkaline rocks are characterized by successions of fractionated, undersaturated to saturated gabbroic through alkalic differentiates. On the basis of data from the Duluth Complex, the layered tholeiitic intrusions have been subdivided into an early group of cumulates consisting of peridotite, anorthositic rocks and granophyre, and a later group of troctolitic through ferrogabbroic cumulates with minor granophyre. On the basis of crosscutting relations and paleomagnetic data, most quartz tholeiite dikes and sills are older (magnetically reversed) than olivine tholeiites (magnetically normal). However, some magnetically normal dikes and sills are quartz tholeiites. Age data and crosscutting relations do not provide a means for placing the alkaline rocks in a stratigraphic succession with the tholeiitic rocks; however, comparison with modern rift environments suggests the following succession of magma types: alkaline (COi-rich to C02-poor), quartz tholeiite (alkalic, H20-saturated), olivine tholeiite (low alkalies and H2O). The distribution of the intrusive products of these magmas is consistent with a transition in the Keweenawan tectonic regime from northwest-southeast wrench faulting to rifting along northwest-southeast transform faults.
IntroductionThe data base on the Keweenawan intrusive igneous rocks consists of the results of field mapping, petrographic descriptions, petrologic and geochemical analyses, and interpretive studies. The extent of the data base decreases in the order listed. In view of this and because field mapping is an iterative process, it is fair to say that knowledge of these rocks has not progressed substantially beyond the preliminary characterization of major exposed rock units. Most of the interpretive aspects of geologic mapping remain to be made or have not been verified. For example, definition of individual intrusive units, correlation of stratigraphic sequences within presumed intrusive units, and pre-and postemplacement structural control on stratigraphic relations of the intrusive igneous rocks are vague and uncertain. If this pessimistic view of the quality and extent of the available data is not overdrawn, it poses the question of what has hampered growth of a substantial data base on a group of interesting rocks that have been studied for over a century. At least two of the possible answers to this question are pertinent to this review. First, the low relief of the region,