58 SURFACE GEOLOGY. 



stream sometimes wears away one of its banks to such a depth, that the channel 

 gradually changes towards that side, while the back water produced on the other 

 side causes a deposit, which is increased by freshets, and although its upper surface 

 becomes nearly level, it yet forms a terrace which properly deserves the name of a 

 glacis terrace. After this process of lateral change has gone on for some years, it 

 not unfrequently happens that the river suddenly deserts its old bed, in consequence 

 of having found a new channel. Successive floods fill up the deserted bed, some- 

 times so as to make a level-topped terrace : but in other cases, it is only partially 

 filled, and exhibits, at least for centuries, evidence of the former presence of the 

 stream. Such are the old river beds in Deerfield meadows, shown on Plate IV. 

 In the short one directly west of the village, the whole process has been gone 

 through since my boyish days, and I have watched its progress with interest from 

 year to year. 



22. It is I apprehend, by modifications of this process, that that variety of 

 glacis terrace exhibited on section No. 31 was produced. Sometimes they may 

 also have resulted from the accumulation of sand and loam on one shore, by the 

 lateral influence of a strong current. I am not prepared to say exactly how that 

 variety of glacis terrace, found in the Alps and other mountainous districts, con- 

 sisting of rather rapid slopes of the whole alluvial formation of a valley towards 

 the stream, was produced. It may, however, have resulted from the sliding down 

 of detrital matter towards the stream from the steep adjoining hill-sides, during 

 the semi-fluid condition of the surface in the spring, or after powerful rains. 



23. On the supposition above made, that during the drainage of a valley like 

 the Connecticut, it assumed the condition of a chain of small lakes, we can see 

 how it is, that around the gorges or straits between them, the terraces should be 

 higher than in the wider parts of the valleys. For the contraction of the stream 

 at the gorges, would check the current there, and thus cause more of the sus- 

 pended matter to be deposited. Very probably it might so fill up the gorges, that, 

 as the continent rose, it would require a great length of time to wear them down 

 to their present depth. 



24. We see then that the various forms of river terraces, whether called delta, 

 lateral, gorge, or glacis terraces, may be formed by the simple drainage of the 

 country, as the surface emerges from the ocean. Nor need we, as has generally 

 been thought necessary, suppose that there were pauses in the vertical movement. 

 That such pauses may have occurred I admit, and that in this way some terraces 

 and beaches may have been produced; but to form the river terraces we need not 

 call in their aid. 



25. I now proceed a step further, and will state certain facts, which prove that 

 river terraces in general could not have been produced by pauses in the vertical 

 movement of the land. 



1. If thus produced, they ought to be the same in number and height in the 

 different basins of the same river, and on different rivers not very remote from one 

 another. For, even though we might admit some small difference in their height 

 if thus produced, their number must correspond, since the water would sink 

 equally in the different basins. But a reference to the sections attached to this 



