the ice invasion, but the Erie and Ontario escarpments and some other similar features now recognized, suggest that in Tertiary times it resembled the present condition of the upper portion of the Mississippi valley, where bold, rock escarpments border wide stream-worn depressions. Deep drift-filled channels are known to cut across the Erie and Ontario escarpments. These seem to have been formed by streams tributary to the main drainage line to the north. If this conclusion is well founded, a study of the hard-rock topography should reveal other similar channels and finally indicate a well matured drainage system. If even the broader and stronger features of the pre-glacial surface can be determined, then the modifications due to glacial abrasion will become conspicuous, and the amount that glaciers have broadened and deepened the basin be determinable. A study of the lithological character of the drift south of the present lakes should show, at least in a rough way, what portion of it was derived from the waste of rocks within the Laurentian basin. This inquiry has already been undertaken by at least two geologists, and estimates of the quantity of material removed from the basins of lakes Michigan and Erie respectively, have been made. This method may be extended so as to embrace a larger area, or some special portion of the great depression best suited for the trial may be selected. If the material removed from the basin or re-distributed within it by glacial action can be shown to be approximately equivalent in volume to the amount of rock excavated in order to form the depression, it would evidently tend to support the hypothesis of glacial erosion. If, on the contrary, the amount of débris derived from the basin should fall far short of what would be requisite to refill it, no very definite conclusion would seem to be indicated unless account could also be taken of the fine material carried away by glacial streams. As the case stands at present it appears that there is evidence of a pre-glacial valley or series of valleys as has been claimed by several geologists, and that all but the boldest features of the old topography have been obliterated or greatly modified by glacial erosion followed by glacial and other sedimentation. Additional observations should show somewhat definitely the amount of work assignable to particular portions of the history. How far the results of subaërial and of glacial erosion have been modified by other agencies, more especially by orographic movements, has also to be considered.