Studies of distribution and character of Pennsylvanian coals in the eastern and central United States have shown that their accumulation resulted from several environmental patterns, as follows: (1) distribution controlled by the building of a widespread delta (Illinois 5, 5a, and 6 coals in southern Illinois basin); (2) unfilled channels, either of alluvial streams (western Illinois No. 4 coal) or unfilled delta distributary channels (No. 5 coal, southeastern Illinois) – thinner coal is found away from the unfilled channels; (3) coal accumulation in an estuary prior to drowning (No. 1 coal of Illinois); (4) deposition on a narrow coastal strip comparable with Atlantic coastal marshes of today (Mulberry coal of Missouri and Kansas); (5) local accumulation in a lagoon behind an offshore barrier (Ogan, Winters, Clarion, and Scrubgrass coals of southern Ohio); (6) local accumulation in a cut-off meander of a stream, with clastic partings and no underclay (local coals in Pleasantview Sandstone of western Illinois); (7) a plain exposed following abrupt marine regression (Lexington coal of Missouri and Iowa); (8) a level depositional plain formed by burial of pre-Pennsylvanian topographic irregularities (No. 2 coal of Illinois; Croweburg coal of midcontinent). Some coals formed on delta plains or flood plains are separated from channels by natural levees, and an example has been found in the substrata of the No. 5 coal in southeastern Illinois. Where a fluviatile or deltaic channel was filled, because of greater compaction of shales a way from the channel, a coal may thin or wedge out over the channel sandstone. A study of the environments in which Pennsylvanian coals have formed contributes also to the understanding of the genesis of underclays. It is suggested that an understanding of composition and geometry of coal substrata should guide exploration for future coal reserves.

Abstract Pennsylvanian major marine cyclothems in midcontinent North America comprise the sequence: thin transgressive limestone, thin offshore gray to black phosphatic shale, and thick regressive limestone. Typically, the cyclothems are separated from one another by well-developed paleosols across an area of perhaps 500,000 km 2 from southern Kansas to Iowa and Nebraska. Southward, the offshore shales extend into the foreland basin of Oklahoma, and the regressive limestones and paleosols grade into deltaic to fluvial clastics derived from the Ouachita detrital source. Texas, Illinois, and Appalachian marine cyclothems are detrital-rich like those in Oklahoma, but appear to be separated by paleosols like those in the northern midcontinent. Because the black phosphatic offshore shales of the midcontinent record sediment-starved, condensed deposition below a thermocline in about 100 m of water, sea-level rise and fall of at least that amount is required over the entire northern midcontinent region to account for the widespread dark shale-paleosol cyclicity. Sparsity to absence of deltas between most cycles on the northern shelf rules out delta shifting as a control over major cyclothem formation there. Continuity of all major cyclothems across both the Forest City basin and adjacent Nemaha uplift rules out local differential tectonics on the northern shelf as a major control. Confinement of all reasonable estimates of cyclothem periods within the Milankovitch band of orbital parameters (20 ky to 400 ky), which controlled Pleistocene glacial fluctuation, points to glacial eustasy as the major control over midcontinent cyclicity. Moreover, only the documented late Quaternary post-glacial rates of sea-level rise significantly greater than 3 mm/yr are sufficient to exceed carbonate accumulation consistently and produce the characteristic thin transgressive limestone overlain by the widespread thin subthermocline black shale in each major cyclothem. Although tectonic subsidence helped provide space for sediment accumulation, tectonic control over cyclothem deposition would require both subsidence and uplift of the midcontinent (or an equally large region) at Milankovitch band frequencies. However, currently developed cyclic tectonic mechanisms that can achieve the required depths repeatedly in a cratonic area act at periods at the very least 5 times greater (2 my+) and at rates of sea-level rise at least 30 times too slowly (maximum of 0.1 mm/yr). Firm biostratigraphic correlation of major midcontinent Upper Pennsylvanian cyclothems with similar depositional cycles in Texas and Illinois allows a strong glacio-eustatic signal to be identified in those regions, with little evidence so far of temporally differential tectonism among them. The lithic differences between carbonate-rich midcontinent cyclothems and detrital-rich cyclothems in Texas, Illinois, and the Appalachians are attributable directly to the greater detrital influx in the latter areas, which could relate as much to appropriate climate and accessibility to detrital provenance as to tectonic activity. Preliminary correlations showing that only certain bundles of major marine transgressions extended into the Appalachian basin suggest that the absence of the others may reflect tectonic uplift there, and that with more definite correlation, a longer-term tectonic signal can be isolated in that area.