the mass an impervious and uniform hard-pan. This has generally been denominated "unmodified drift." It is that which escaped the assorting action of the water issuing from the glacier. In other places, this unmodified drift would be superficially assorted, showing the effect of running water after its deposition or in the act of deposition. Probably very much of that portion of the drift that lay in the course of the broad Mississippi, yet south of the limit of the glacier, would be superficially worked over, losing much of its clay. We actually see vast tracts on the Upper Mississippi, and even in the latitude of St. Paul, in which the surface consists, to a considerable depth, principally of stratified, sand and gravel. He would also find parallel ridges of drift-materials, consisting largely of the coarser portions, and showing stratification where water passed over or through it in being deposited. Some such ridges would still retain the most or all of the original clayey portions. This would be the case where the drainage was not powerful. Such ridges mark the places at which the retreat of the ice was temporarily stopped by a period of greater cold, the slow advancing of the ice under the propulsive forces already named serving to heap up a greater amount of detritus all along the ice-margin. These ridges are known as moraines, and they occur in all parts of the drift-latitudes. They are developed on a very grand scale in Northwestern Ohio.
There is still one important point in this discussion that must not be omitted. It is plain to see that, in some parts of the Northwest, the advance of the continental glacier would be up gentle slopes, instead of descending an incline. These slopes, of course, present obstructions to the movement of the ice in those directions. It is true, also, that the continental glacier would tend to level the country and obliterate such northward slopes. But, in the later part of the ice period, the valleys would be the last relinquished, and would be deeper dug by isolated branches or spurs from the main ice-sheet, which would conform in their direction to the contour of the valleys they might occupy. All glaciers, however, whether continental or local, would avoid an ascent if there were any other passage. Now, when a glacier, propelled by a force exerted far to the north, meets with a gentle slope toward the north, the water which issues from its foot Will be confined in a lake about the foot of the ice, and will rise to the height of the lowest outlet. Into this lake may flow streams of considerable size, bringing their sediment from the south, east, or west, according to the topography. Here we should have, then, a constant accession of drift from two sources, the chief of which would be, of course, the glacier itself. As this drift is brought under the influence of standing water, its fine parts are floated away by currents and waves, to be spread over the bottom of the lake in horizontal laminations, the principal portion, and notably the bowlders, sinking at once to the bottom unassorted. Thus, by the continued slow withdrawal of the field of
