Discussion
T. J. Kemmis: I wonder if you would care to comment on how your mechanism fits the stratigraphic record left by the Laurentide ice sheet in North America? Your contentions, both of a single erosional process by continental ice sheets and of incorporation only of regolith, seem incompatible with the record left by continental ice sheets in mid-western North America. For example, an unpublished map of Hoiberg shows that the Sangamon interglacial soil in Illinois is present beneath later Wisconsin-age glacial deposits for a distance of 80 km back from the Wisconsin front, indicating that much of the existing pre-Wisconsin material was not incorporated at over a distance of 80 km up-glacier from the Wisconsin margin. Several detailed stratigraphic studies of tills in the glaciated lowlands of North America (e.g. Reference Willman and FryeWillman and Frye, 1970; Reference Karrow and LeggetKarrow, [c1976]) have shown the stratigraphic succession to consist of a sequence of texturally and mineralogically distinct calcareous tills. It is hard to envisage how these tills, which are found in the sub-surface to successively overlie one another over wide-spread areas (on the order of hundreds of square kilometres), could result solely from the incorporation or regolith.
I. M. Whillans: The main point is that there are some important problems with till provenance. I suggest regolith freeze-on as an important mechanism. Certainly some areas were covered but not eroded.
The stratigraphic descriptions For Illinois define members that cover areas of the order of 50 km by 100 km (103–104 km2). Perhaps each member is to be associated with a single glacial advance or perhaps each bed represents an advance (a bed is a subdivision of a member). It would be very interesting to determine if the compositional variations within each member can be related to a similar pattern in the bedrock and other pre-existing deposits.
