S2 



NEfW YORK STATE MUSEUM 



In case the sea invaded snch a valley during the ice retreat, it 

 would control to a certain extent the deposition of washed gravels 

 on the sides of the ice tongue but, unless the submergence were 

 very great as compared with the depth of the valley, local embar- 

 rassments to seaward drainage would undoubtedly occur. Such 

 embarrassments would arise where spurs entered the valley be- 

 tween side streams, or where the ice melted less rapidly,' thus 

 giving rise to levels of building above sea level. 



Application of theory to the Hudson valley. The peculiar form 

 of the Hudson valley, its rock benches or terraces inclosing a deep 

 gorge, and the Highlands through which the river passes by a nar- 

 row defile with a constricted development of these benches, must 

 have affected in a marked manner in its different sections the 

 mode of retreat of the ice margins arid consequently the distribu- 

 tion of the sediments laid down in the presence of the ice. First 

 the north and south depression through the Highland section 

 whether or not a continuous river valley as in postglacial times 

 would have guided a strong current of ice southward and during 

 the period of final melting would have given rise to a long tongue 

 of ultimately stagnant ice occupying the valley north of the High- 

 land gorge. 



It is to be presumed that the barrier opposed to ice movement 

 by the Highlands would have led in the advance, as in the retreat, 

 to a stage when the moving ice banked up against the northwest- 

 ern wall of the Highland ridges would have poured through the 

 defile at West Point as a small valley glacier spreading out on 

 the rock terraces below Peekskill or pushing south wholly confined 

 within the Hudson gorge; at least in the retreat this was the case 

 when certainly this gorge had its present general cross-section. 



Wherever during the retreat the ice front crossed the river and 

 deployed on the banks to the east and west, the streams discharg- 

 ing from the ice would bring down heavy loads of clay, sand and 

 gravel, and bank them up against the ice front in the river gorge 

 and over the neighboring rock terraces. Such deposits might 

 originally completely fill the gorge, to be subsequently partly re- 

 moved in the renewal of ordinary river drainage in the area. 



When the ice had thinned so as to form a long narrow tongue 

 filling the lower portion of the valley, covering the gorge 

 and a considerable breadth of the rock benches on either 



