The bedding-plane anisotropy and structural configuration of the Mesoproterozoic Belt-Purcell Supergroup guided a narrow magmatic salient >350 km eastward from the Salmon River suture of Idaho to the foreland basin of central Montana, along a deep graben within the southern part of the Belt-Purcell Basin. The magma assimilated anatectic melt from the lower Belt-Purcell Supergroup in the western half of the graben, where the Lemhi subbasin had intersected and deepened the graben by several kilometers. The magma stepped across the stratigraphic section as it intruded eastward along the graben, spread laterally as it climbed into the overlying Paleozoic and Mesozoic strata, and eventually erupted into the foreland basin. This paper develops a model in which the magma formed a thick, east-tapering wedge beneath the Belt-Purcell carapace. The wedge elevated and tilted its lid, which failed along the trend of the graben to a terminus in the Crazy Mountains basin of the Helena structural salient, much like a tectonic-scale landslide. The carapace failed in two main phases between ca. 100 and 75 Ma. It slid ~100 km during the first failure phase, and ~40 km during the second, when the Boulder batholith and its volcanic cover filled a large pull-apart structure within the carapace. Slaty cleavage, tectonic slides that omit strata, and a nested series of hairpin-shaped allochthons characterize the failed carapace. Shear zones and nappes bound the carapace; the sinistral Lewis and Clark line bounds it on the north, and the dextral southwest Montana transverse zone bounds it on the south. The Lewis thrust fault and associated structures of the Rocky Mountain fold-and-thrust belt overprinted and displaced the magmatic salient and its carapace from ca. 74 to 59 Ma. The magma crystallized, cooled, and generated hydrothermal ore deposits from Late Cretaceous to middle Eocene time. Eocene extension overprinted the system from 53 to 39 Ma and exhumed its infrastructure in core complexes. Those exposures, together with regional structural tilt, enable reconstruction of a deep cross section of the magmatic wedge and its carapace.

South America, Central America, the southeastern United States, Arctic Canada, Europe, Asia, and Africa all have been suggested as possible or probable biogeographic sources for taxa that appeared in the Western Interior of North America during the late Paleocene and early Eocene. Recent compilations of the geographic and temporal distributions of Paleocene and Eocene mammals and new data, derived primarily from recent collections from early Tiffanian (late Paleocene) quarries in the Crazy Mountains Basin of south-central Montana, permit tests of these hypotheses, particularly those involving a southern New World origin. Significant first appearances of mammalian higher taxa in the Western Interior occur in the earliest Tiffanian, late Tiffanian, earliest Oarkforkian, and earliest Wasatchian. Those that appear in the earliest Tiffanian probably were derived from late Torrejonian forms in the same region. It appears, therefore, that there was not a pronounced geographic shift in North American mammalian faunas across the Torrejonian-Tiffanian boundary as suggested in some southern New World origin hypotheses. It has been suggested that Palaeanodonta, Dinocerata, and Notoungulata (represented by Arctostylopidae), which appear in the late Tiffanian in the Western Interior, originated in South America, but the evidence is inconclusive and highly controversial. New higher taxa that appear in the Western Interior at the beginning of the Clarkforkian, particularly Rodentia and Tillodontia, probably originated in Asia and dispersed across Beringia. Most of the suprageneric taxa that first appear at the beginning of the Wasatchian in the Western Interior (Perissodactyla, Artiodactyla, Adapidae, Omomyidae, and Hyaenodontidae) also probably appeared in Asia and Europe at essentially the same time; there is no evidence for heterochrony. Recent paleontological discoveries and paleogeographic evidence suggest that the ultimate origins of some or all of these taxa lay in either Africa or the Indian subcontinent. The latter biogeographic source has not been seriously considered previously.