Hubbard — High Level- Terraces in Southeastern Ohio. Ill 



above are the best terraces seen. Others were noted, one or 

 two in a place, at several points, but in every case except one 

 they turned out to be rock terraces. In the one case, the 

 terrace was due to landslide. It is not common for landslid- 

 ing to give a form even resembling the neat terraces, but in at 

 least three examples such was the case. There may be shore- 

 line terraces in this valley, but the author failed during a care- 

 ful search to find them. All the highland terraces seen are 

 above the outwash, and are in the region of continuous Penn- 

 sylvanian rocks whose chief characteristic is their variableness 

 as evidenced by their relative resistance to weathering. 



In closing, I desire to call attention to a few facts which it 

 may be well to bear in mind when further reviewing the prob- 

 lem of the high-level, slender terraces. 



Terraces may often be due to wave work. As fine terraces as 

 have ever been described have undoubtedly been developed 

 by this physiographic process along shore lines.* It has been 

 stated by other observers, and I concur, that the terraces of 

 the Ohio valley are commonly on the steeper slopes and along 

 the smaller valleys. Wave-made terraces would be slow in 

 forming on steep rock slopes, because there would be little 

 rock mantle, and any material derived by wave action would, 

 by the same agent, be easily scattered down the steep, sub- 

 merged slopes, without forming terraces. The walls of smaller 

 (narrower) valleys would be among the least propitious places 

 for wave terraces to develop because the waves would here be 

 weakest. On the other hand, gentler slopes, soil covered and 

 lying along the larger valleys, would be very proper places for 

 wave-cut terraces ; but under such conditions terraces rarely 

 occur. Further, their preservation in such places would be 

 easier because erosion is less active on gentle slopes. 



Granting that terraces, due to wave action, have been formed 

 along the walls of partly submerged valleys, other related 

 features should also have been formed. Deltas, at bay heads, 

 or along the sides of the larger bodies of water where streams 

 entered, should be common. Not one has ever been reported. 

 Probably some streams were too short to construct deltas, but 

 many were not. Lake clays and sands should have been 

 deposited, especially near where the smaller streams entered 

 the drowned valleys. They also have never been reported. 

 In Central New Yorkf in the finger lake region deltas occur 

 in profusion and associated lake clays occur to depths of 10-15 

 feet down .slope from many delta series. But only in rare 

 instances are cliffs or wave-cut terraces found. They have 



*U. S. G. S., Monographs I and XL 



f Watson, T. L., N. Y. State Mus., 51st Ann. Rpt., vol. i, pp. r65-117, 1899 ; 

 and. several other authors. 



