24 SURFACE GEOLOGY. 



Groveton a ridge of till, from 60 to 100 feet above the river, projects half 

 a mile westerly into the valley, or half way across it, appearing like a ter- 

 minal moraine. Horse-shoe pond, on the north-west side of this ridge, 

 occupies a portion of a deserted river-channel. These ancient river-beds 

 are frequently shown by such ponds, commonly called sloughs or moats, 

 of which Baker's pond, near Lancaster, is another example. We see the 

 river now slowly changing its position by wearing away one or the other 

 of its banks, and it has thus swept many times from side to side in exca- 

 vating its valley between the bordering terraces. 



Fiftecn-milcs Falls. From the mouth of John's river the Connecticut 

 has a rapid descent for twenty miles, amounting to 370 feet, falling from 

 830 to 460 feet above the sea. The bed of the river is a nearly continu- 

 ous slope of coarse till, showing abundant boulders, but with scarcely any 

 exposure of solid ledges. The only place where these were noticed in 

 our exploration was at the "lower pitch," or foot of these rapids, about 

 a mile above the mouth of Passumpsic river. Here there is a precipitous 

 fall of a few feet, and this is said to be the only point of abrupt descent. 

 In other parts of its course the falls of the Connecticut are produced by 

 ledges, and the channel, except at such falls, is composed of gravel, sand, 

 or silt. Nowhere else below West Stewartstown, except at the falls of 

 Northumberland and Columbia, and rarely, if at all, southward, does the 

 river flow over the glacial drift or till. 



The noticeable features of the valley in this distance are, that it is deep 

 and narrow, with sloping sides of till, and destitute of the level alluvial 

 terraces and intervals which occupy a large width everywhere else along 

 the river. Where any modified drift does occur, it is coarser than usual, 

 being generally gravel, sometimes imperfectly rounded or water-worn, and 

 its surface has commonly an irregular slope. The upper portion of these 

 rapids is especially destitute even of such alluvial deposits, the highest 

 •that occur being from 60 to 75 feet above the stream. It is frequently 

 evident that the source of these deposits is not the main river, but a trib- 

 utary, as in the case of Niles stream in Concord, Vt., and on both sides 

 at Upper Waterford. These deltas are greater in height, as well as in 

 amount, than the scanty remnants of the alluvium of the main valley. 



On the lower portion of these rapids the modified drift appears in 

 greater quantity and at a much increased height. Opposite Lower 



