190 THEORY. 



2. On this supposition, also, the terraces ought to agree in height, at least upon opposite 

 sides of a valley. Circumstances, indeed, might erase all traces of their action in particu- 

 lar localities, but such great irregularity as exists in this respect, cannot thus be explained. 

 We refer to every section that is measured to show this irregularity. 



3. The difference in the number and height of the terraces on the principal stream and 

 its tributaries at their debouchure, "affords decisive proof that terraces were not the result 

 of paroxysmal elevations. Here we find two sets of terraces formed in the same bank 

 of detritus ; one set, usually the smallest in number, on the main river; and the other set, 

 formed by the erosion of the tributary through the first. For examples we refer to several 

 of the sections. Upon Fig. 49, showing the number of terraces upon the Ashuelot River 

 in Hinsdale, N. H., and upon the Connecticut at Vernon, we have five terraces on the 

 former and three or four on the latter. Fig. 54 shows six terraces at the mouth of Sax- 

 ton's River, while in Westminster (Fig. 53), a little to the south, there are only four. 

 In the north part of Vernon (Fig. 50), there are only four on the Connecticut; but on 

 West River in Brattleboro, perhaps two miles north, we find nine, and on Whetstone 

 Brook ten (Fig. 52, and Plate VII, Fig. 2.) And the latter rise no higher than the 

 former. 



Plate VII, Fig. 1, shows ten terraces upon Williams River, and five upon Connecticut 

 River at its mouth, the tenth and fifth being of the same height. And as both of these 

 sets are found in the same bank of sand and gravel, it is certain that if one set was pro- 

 duced by paroxysms, the other could not be; and if one set could be produced without 

 paroxysms, the other might be also. 



11. The same facts prove that these terraces were not produced by the bursting of bar- 

 riers : for the results would have appeared in terraces of the same number and height 

 upon the opposite sides of the stream, and the same facts that disprove the paroxysmal 

 theory, disprove this theory also. 



12. The lake terraces that occur in Vermont have the same origin as the lateral terraces 

 of rivers. They may be explained either by the second or third methods, according as a 

 current passes through the lake or not. And we have every reason to suppose that lake 

 terraces all over the country are formed in the same way. We have described three pond 

 terraces produced by the sudden drainage of Lime Pond in Williamstown. This case 

 illustrates the formation of terraces by drainage. 



The method in which moraine terraces have been produced has perhaps been sufficiently 

 considered. Ice was called in with water. It was supposed that icebergs became strand- 

 ed at the base and on the sides of hills, and that deposits Avere made around and upon 

 them, and they would have been level-topped if the ice had remained ; but in consequence 

 of the melting of the ice they are now extremely irregular. No more satisfactory explan- 

 ation of these phenomena has ever been given, and therefore we adopt this view, certainly 

 until a better one shall appear. 



13. The facts and reasonings that have been presented, exhibit to us one simple, grand, 

 and uninterrupted series of operations, by which all the changes in the superficial deposits 

 since the drift, have been produced. We see the continent slowly emerging from the 

 ocean ; rivers commencing their wearing action on the islands ; waves and oceanic currents 

 meeting the detritus of rivers and comminuting, sorting, and arranging the same, in the 



