This proposal is to develop techniques for distinguishing cumulate from intercumulate components in rocks from the Beaver Bay gabbro complex, Minnesota, using backscatter SEM imaging, Normanski interference-contrast microscopy, and electron micro- probe traverse. From this study the trapped intercumulus liquid abundance (i.e., original porosity) will be estimated through summation of all intercumulate components, in a manner analogous to original porosity estimates in sandstones. A major study of major element, REE, and other trace element variations in whole-rocks and in mineral separates is also proposed in order to examine the relationships between whole-rock chemical composition and abundance and composition of the cumulate phases. We also wish to understand the relationship between the variation in the amount of trapped liquid and the composition of the liquid, and to test whether the hypothesis involving a compositionally uniform trapped liquid in the Duluth complex applies to Beaver Bay gabbro. The relatively small size, excellent exposure, complete stratigraphic sequence, and extreme iron-enrichment differentia- tion trend of the Beaver Bay gabbro make it well suited for the kind of study proposed here. The data obtained from this study will also lead to the characterization of the parent magma composition(s), a description of the crystallization sequences in the rocks, and an understanding of the processes to strong iron- enrichment crystallization trends in the complex.
