WATER-LEVELS OF BELFAST-FILLMORE LAKE. 437 



down-cutting or removal of the damming drift is difficult to estimate. 

 The waters seem to have leveled the drift without cutting to rock, the 

 channel being spacious and near to grade. 



Water-levels. — This stage of the Genesee waters covered a longer stretch 

 of the valley than any other and probably endured for a longer time 

 than any other of the lakes previous to the Warren stage. From Belfast 

 to Portageville the summit planes belong to this stage. The waters, how- 

 ever, buried the upper valley some distance above (south of) Wells ville, 

 but from the southern limit of the waters to a point above Belfast these 

 planes are overtopped by those of the previous Wellsville lake. They 

 are, however, generally so much stronger than the earlier Wellsville ter- 

 races that they can probably be recognized. In this basin we have for 

 data the reliable altitudes of the Western New York and Pennsylvania 

 railroad, based upon the canal levels. This is the main section of the 

 old river valley. . It is in places two or three miles wide, but the ancient 

 borders are often obscured by the heavy deposits of drift, which also fre- 

 quently form massive hills in the midst of the valley, morainic, kame- 

 like and esker-like. These drift hills have not usually retained strong 

 terracing by the static waters, perhaps on account of their unenduring 

 character and the great destruction they have been subjected to in the 

 middle of the valley. The higher leveling and shore benching cannot 

 be seen well from the middle of the valley or from the railroad, but only 

 from more commanding situations. The lower valle}^ is very rich in 

 stream plains and lower terraces. 



Although the high water-planes have been seen at many points in this 

 section of the valley, only a few have been measured or closely estimated, 

 but a number sufficient to serve as examples. 



Below Cuba village the bordering plain is 40 feet over the channel, or 

 1,536 feet. The Erie station is upon this terrace, which is perhaps two 

 miles south of the present water-parting and may indicate some down- 

 cutting and backward erosion by the outlet stream. Terraces correspond- - 

 ing in altitude to this are seen northward by the Cuba reservoir, at Black 

 creek, and indeed almost continually as far north as Rockville. How- 

 ever, these terraces are all in the Black Creek valley, which was the basin 

 of a local lake (see page 448), and it might properly be claimed that they 

 do not prove the level of a larger lake in the river valley. The phenomena 

 are, however, far greater and more imposing than those produced by a 

 local lake. Corresponding summit terraces are found all the way to 

 Portageville, and leave no room for doubt of the larger lake and outlet. 



At Belfast and Oram el the terraces are conspicuous, but the altitudes 

 have not been even estimated. Upon the east side of the valley north of 

 Belmont two lines of water erosion show clearly, seen from the wagon- 



