Three Wisconsin age glacial lobes, the Lake Michigan, Saginaw, and Erie, extended into northern Illinois, northern Indiana, and northern Ohio. The maximum advance of these three lobes is considered to be the Minooka moraine of the Lake Michigan lobe, the Iroquois-Packerton moraines of the Saginaw lobe, and the Union City moraine of the Erie lobe. The writer considers that these moraines mark the Gary drift border in Illinois, Indiana, and Ohio. This conclusion is at variance with the interpretations of Wayne and Thornbury in Indiana, Martin in Michigan, and Goldthwait in Ohio.
The retreatal history of these three lobes involves a correlation of moraines and other evidence heretofore considered separately for each lobe. The position of the ice front during the release of meltwater, which produced the Kankakee Torrent in Indiana and Illinois, is significant in the history of all three lobes. The tracing of drainage channels which carried the Kankakee meltwater in southern Michigan and northern Indiana provides the evidence for locating the ice margin of the three lobes during the first phase of their retreatal history. Field evidence shows that both the Lake Michigan and Erie lobes readvanced across terrain formerly occupied by Saginaw lobe ice after the maximum discharge of the Kankakee Torrent. The ice position at this time is marked by the Tekonsha moraines of the Lake Michigan and Saginaw lobes and a drift border on the northwestern side of the Erie lobe in Michigan which has been traced for a distance of only 12 miles. The relationship of the border between drift of the Erie lobe and drift of the Saginaw lobe to the several moraines of the Erie lobe in Michigan, Indiana, and Ohio is not known.
In order to clarify existing late Wisconsin stage terminology the writer proposes that the moraines of the Lake Michigan lobe from, and including, the Minooka north to, but not inlcuding, the Port Huron, be used as a type area for the Gary glaciation.
Radiocarbon dates are considered to be a means of establishing an absolute chronology of events that is based on independent field evidence and not as a primary tool for deciphering Pleistocene chronology.