ANCIENT WATER LEVELS OF CHAMPLAIN- HUDSON VALLEYS 83 



side, the streams from the neighboring open country would build 

 their deltas against the ice margin in the form of terraces involv- 

 ing buried shreds of the ice margin, and having when the ice 

 melted away kettles or depressions marking the sites of these 

 buried or partially inhumed ice masses, and a relatively steep 

 but perhaps hummocky or kamelike terrace front overlooking the 

 river gorge and at varying distances from and elevations above 

 the gorge. The water thus impounded against the ice margin 

 would flow along the ice edge or finding its way through crevasses 

 and water tunnels in the ice escape with the glacial drainage 

 without producing marginal stream phenomena. 



Finally when the ice had melted off from the rock terraces for 

 a time a long narrow tongue would still occupy the gorge itself 

 forcing some of the drainage over the rock benches and covering 

 them with sheets of clay, sand and gravel. This coating of 

 glacial materials might here and there mantle the ice in the gorge 

 where that had been lowered by melting so that its surface lay 

 below the level of the rock terraces. In any event when the ice 

 finally melted out of the gorge the rock benches would be 

 coated with terrace drift to their edges, the deposit here 

 and there descending into the gorge as if it had once entirely 

 filled it though this may never have been the case. Unless the 

 evidence of ice contacts be found, it would be an extremely 

 difficult task to determine with certainty the original extent and 

 limits of such deposits and to discriminate them from remnants 

 left from a reexcavation of a gorge which has once been filled 

 by glacial sediments. 



Moreover, such lateral glacial deposits will depend for their 

 elevation on the hight of the rock terraces on which they have 

 been spread out. From point to point they should merge into 

 the frontal deposits of the successive stages of the retreat of the 

 ice front. Along such frontal lines the materials would be 

 coarsest, gradually passing to finer and finer materials toward 

 the south or away from the ice if the drainage was in that direc- 

 tion. Here and there lateral tributary streams would pour in 

 their contribution of detritus and produce local variations of 

 texture. or thickness of the sediments. 



If the ice retreat by successive oscillations in which the reces- 

 sional movements overbalance the forward ones, the complica- 



