84 NKW YORK ST ATI: MUSEUM 



tions ill the sodiniciii.ii'y history will be greatly increased. The 

 glacial days laid down in the outer belt of deposition of one 

 frontal stage may be eroded by tha overriding action of the ice 

 of the next and then slieeled over, ](artly or wholly, by deposits of 

 till or boulders as w(^ll as by sheets of coarse gravel and sand. 



Another etfect })roducing local terraces will arise during the 

 melting of ice from a gorge like that of the Hudson 

 with dissected walls quite independently of sea level so 

 long as the rock terraces rise somewhat above sea level. 

 As soon as the ice is limited to the main gorge, the 

 enibayments in the wall, receiving drainage from the ice and 

 such lateral streams as may pour into them from the open 

 country, will form temporary lakes and be filled and sheeted 

 over with sands or gravels at levels determined by the effective- 

 ness of the ice barrier and the duration of the process of filling, 

 as well as by the elevation of the floor of the area of 

 deposition. 



JSncccssive stages in the cross-section of a melting glacier in a 

 valley like iliat of the Hudson river. The glacier which covered 

 eastern New York, it may be said, was pushed on to the area by 

 the pressure of its own accumulation in the Laurentide district. 

 Eliminating the eftect of forward motion in the ice and supposing 

 the glacier to have been stagnant over the region between the 

 Highland canyon of the river and the Catskill mountains, it would 

 follow that for some time during the declination in the thickness of 

 the ice sheet the relations to the valley would be those indicated, in 

 figure 5, in which the ice sheet not only filled the valley but cov- 

 ered the divides on either side. 



Viw a long time after, when the ice had dwindled down to a 

 lon^^nc lining the bottom of the valley, its cross-section would have 

 been llial shown by liP> in figure 5 and this general cross-section 

 would have been retained till a final stage wa,s reached, when the 

 ice filled llie gorge only leaving the top of the rock terraces free 

 l'<M- lalci'Ml drainage. 



In lliis finnl slage jlie cross-section would be that shown in DD, 

 ligure n, in which Hie bi*oad rock terraces might become the seat 

 of lakes and laleral stream deiK>sits. Upstream and behind con- 

 striclions in (lie valley wliei-e Ihe terraces l)ecame wedged out as 

 in the TTJLilil.inds, by nnconsnmed spurs from the valley sides, 



