An ultramafic complex at Union Bay, in southeastern Alaska, is an exceptionally well-exposed example of a series of similar complexes forming a belt parallel to the Coast batholith.
A body of gabbro, approximately circular in plan and about 6 miles in diameter, intrudes folded sedimentary rocks of probably Triassic and Cretaceous age. A moderately low grade of regional metamorphism in the sedimentary rocks is increased to almandite-zone grade adjacent to the gabbroic contact. Intrusive into the gabbro is a remarkable ultramafic complex which comprises a vertical pipe approximately 1 mile in diameter, to which is attached a lopolithic offshoot approximately 5 miles long and 3 miles wide. The ultramafic units range through hornblende pyroxenite, pyroxenite, olivine pyroxenite, peridotite, and dunite, and both the pipe and the lopolith show a well-developed concentric zoning with dunite in the center and pyroxenite or hornblende pyroxenite on the periphery. Magnetite is a primary constituent of the pyroxenite unit.
The most plausible explanation of the structure and petrology of the ultramafic complex is a hypothesis of successive injections of an ultramafic magma changing in composition with time and intruded in reverse of the order of crystallization of the component minerals. It has not been possible to explain the gabbro (which is olivine-free) by the same mechanism, yet the space and time relationships here, as elsewhere in the world, are probably not accidental. It is proposed, therefore, that a body of gabbro was emplaced, and into this was injected a plug and associated lopolith first of diopside-magnetite magma with minor olivine, then diopside-olivine magma, and finally olivine magma, each new magma intruding the central, incompletely solidified portion of the earlier magma. Minor mixing and melting explain the composition and distribution of the intermediate units of the complex. Lack of contact metamorphism of grade that might be expected to accompany intrusions of such high temperature is explained first by the fact that the ultramafic magmas were almost everywhere in contact with gabbro, not sedimentary rocks, and second by the fact that, being anhydrous, the ultramafic magmas absorbed, rather than expelled, volatiles.
The only likely source for magmas of these compositions would be melting portions of the mantle, but no proposed mechanism for supplying the energy for this process is suggested herein. With an available source of energy, however, it can be postulated that magmas would be generated in the succession proposed. The gabbroic magma probably would originate as a result of melting at a shallower level.