228 R. S. TARR — MORAINES OF SENECA AND CAYUGA LAKE VALLEYS 



margin, all testify to this origin. The moraines do not appear to be in 

 any essential respect different in origin from those now accumulating as 

 marginal deposits along the edge of the Greenland ice sheet. 



Marginal and outflow Channels 



Closely associated with the moraines are channels which were evi- 

 dently occupied by streams marginal to the ice lobes. Besides being 

 too large for streams now occupying them, and being very often in posi- 

 tions where drainage could not exist without some barrier now gone, the 

 channels are peculiar in a variety of ways. Sometimes there is only one 

 bank, the other having been the ice. Often the channels both begin and 

 end abruptly, their continuation doubtless having been on the ice, and 

 at times they terminate in gravelly deposits built of debris which streams 

 brought through these valleys. They vary in length from a few score 

 yards to 2 miles, and in depth from 4 or 5 feet to over 50 feet. Usually 

 they are in drift, but some are cut partly or wholly in the bed rock. 

 Small streams in some cases meander over the swampy bottoms of the 

 broad, flat channels. 



There are several well defined outflow channels across divides, usually 

 across terminal moraine deposits behind which ice-dammed lakes were 

 held. Of these the largest are the two that carried southward the waters 

 ponded back in the Cayuga and Seneca valleys when the ice was with- 

 drawing from these valleys. The Seneca valley overflow, known as the 

 Horseheads outflow, which carried the waters of glacial lake Newberry, 

 is the longest and best defined. 



Outwash Gravels* 



Wherever the ice stood on a grade sloping away from it, or wherever 

 by its deposits it built up such a grade, the outwash gravels have raised 

 and leveled the valley bottoms by building outwash gravel plains. The 

 best and most perfect case is in the Horseheads-Elmira valley, where 

 the major part of the gravel supply came from the main Seneca valley 

 lobe ; but a similar though less extensive outwash plain was built south 

 of the Cayuga Lake valley. Smaller outwash gravel plains were formed 

 in other valleys, and their position is indicated by the dotted area on 

 the map (plate 36). In each case the outwash gravels commence at the 

 outer base of terminal moraines either as terraces or kamey gravel hillocks; 

 but in a very short distance they become remarkably level plains, with 

 occasional kettles and crossed by channels of the outwash streams. 



* These, together with the eskers, kames, and other Quaternary deposits, are more fully de- 

 scribed in the Watkins Glen Folio, 



