428 ANNALS NEW YORK ACADEMY OF SCIENCES 



we have positive proof in the submerged canyon of the Hudson of an eleva- 

 tion of approximately three thousand feet, or, according to Spencer, of nine 

 thousand feet or more. From this elevation date the trenching of the Ter- 

 tiary peneplain and its connecting base-leveled valleys and the main features 

 of the modern coastwise topography, including the cuesta of Long Island 

 and the inner lowland of Long Island Sound. 



It appears most probable, as first suggested by Upham, that the Pleisto- 

 cene ice-sheet originated in this latitude by accumulation, with the sub- 

 sequent development by movement and ablation of a bold, aggressive, 

 moraine-building front. The now drowned inner lowland of Long Island 

 Sound is undoubtedly still floored by Cretaceous clays and sands. Across 

 this floor, except at the narrow east and west ends, as shown by Merrill, 

 the ground moraine was not dragged; and the erratics from the Connecti- 

 cut shore must have been transported englacially, as also suggested by 

 Merrill. The building of the moraines is due to the deformation by the 

 thrust, and in part also by the vertical pressure, or dead weight, of the ice 

 of the plastic Cretaceous clays and sands and the overburden of Tertiary 

 gravel, and the incorporation in the latter, by the joint agency of the defor- 

 mation and glacial streams, of the erratic detritus set free by the ablation of 

 the ice. 



The transverse valleys and deep bays of the north shore of Long Island 

 are probably in part pre-glacial, — original features of the cuesta and 

 inner lowland. But in part, also, they must be attributed to the erosive 

 action of the advancing ice, and to the occupation of pre-determined de- 

 pressions by lobes of stagnant ice during the glacial retreat, while the bor- 

 dering areas were being overspread by washed or modified drift, chiefly 

 sand and gravel. In this connection it is interesting to note the close 

 agreement in trend of these valleys with the glacial movement. 



During the advance, as well as during the retreat, of the ice-sheet, con- 

 ditions favored the formation of glacial lakes; and the outflowing glacial 

 streams were, doubtless, building both delta and outwash plains of sand 

 and gravel (earlier Manhasset gravels), derived chiefly from the deformed 

 beds of Pliocene Yellow Gravel and Cretaceous sand. These plains were, 

 in turn, deformed by the continued advance of the ice and buried beneath 

 the moraines. Thus deposits essentially contemporaneous with the moraines 

 have come to be regarded as belonging to a distinctly earlier stage of the 

 Pleistocene; and, apparently, sufficient account has not been taken of the 

 disturbing and complicating agency of the ice acting in conjunction w^ith the 

 glacial waters, — fluvial and lacustral. 



The recession of the ice margin, first from the outer, and later from the 

 inner, moraine inaugurated anew general glacial-lake conditions along the 



