RIVER TKRRACES IN NEW ENGLAND 485 



of swinging and the terrace scarps come nearer together. The second theory postu- 

 lates as man}'- uplifts as there are terraces, and further requires a systematically 

 diminishing time interval between the uplifts, thus allowing the river to open a 

 wide valley floor in the early stage of terracing, but only a narrow floor in the 

 late stage. 



Miller's theory is, in efi'ect, that the lateral swinging of the terracing stream is 

 occasionally hindered by the' discovery of a rocky ledge or a mass of till in one 

 bank or the other, and that inasmuch as such ledges and till masses must (other 

 things being equal) be more frequently encountered in the later and deeper stages 

 of erosion than in the earlier and shallower stages, it necessarily follows that the 

 scarps of the low-level terraces must be held nearer together than those of the 

 high-level terraces. Where ledges are few the early-made terraces are apt to be 

 undercut and destroyed by the river in forming the later-made terraces, as in the 

 foreground of the accompanying block diagram. Where ledges are numerous, a 

 whole flight of terrace steps may not only be formed but preserved, as in the 

 background of the diagram. A small ledge, exposed for only 5 or 10 feet in the 

 bank of a stream, may restrain the swinging of the stream and thus defend the 

 surmounting terrace for some distance, both up and down the valley. The cusps 

 by which terrace fronts frequently advance into the valley are often determined 

 by ledges. A long sloping ledge may determine a group of cusps, systematically 

 placed one over the other. 



A recent examination of the terraces on Saxtons river, near Bellows Falls, Ver- 

 mont, and on Westfield river, near Springfield, Massachusetts, shows that the ter- 

 race fronts there are frequently controlled by ledges and that Miller's theory is 

 closely applicable in explaining them. A fuller account of these and other ter- 

 races will be presented in a forthcoming number of the Bulletin of the Museum of 

 Comparative Zoology at Harvard College. 



Remarks were made by N. S. Shaler. 



The following three papers were read by title without discussion : 



WISCONSIN SHORE OF LAKE SUPERIOR 

 BY G. L. COLLIE 



The paper is printed as pages 197-216 of this volume. 



LANDSLIDES OF ECHO AND OF VERMILION CLIFFS 

 BY RICHARD E. DODGE 



I Abstract'] 



The northern portion of the Echo cliffs and the Vermilion clifis along the 

 margin of the Paria plateau exhibits a wonderful series of landslides. The slides 

 occur in the basal layers of the Trias, where they are feebly supported by the un- 

 derlying Peruiian, and in some cases have slipped over the edges of the exposed 

 Permian layers, almost shrouding them. The upper clifl", made by the stronger 

 top layers of the Trias, presents a very even front where bordered by the slides, 

 and as yet has contributed but little waste to the surface of the slides. 



LXIX— Bull, Geol. Soc. Am., Vol. 12, 1900 • 



