100 ON EROSIONS OF THE EARTH'S SURFACE. 



apply to the passage cut through the trap at this place, as were made in relation 

 to that at Sunderland. 



The two last examples, being the one in sandstone and the other in trap, would 

 more logically be described in another part of this paper, but it seemed most 

 natural in passing down the Connecticut to notice the gorges consecutively. 



6. Gulf between Mddletoicn and the mouth of Connecticut river. — As you pass 

 through this gulf in a steamboat, and see how, in many places, the high and rocky 

 banks crowd down upon the river, and even jut into it, you cannot resist the con- 

 viction that the stream itself, or one of a similar character on a former continent, 

 must have had much to do with its erosion. You cannot believe either that it is 

 a gorge produced solely by original folding of the strata, or by oceanic action. It 

 is too long, say about 20 miles, and probably I might add, too crooked, to admit a 

 sufficient force of waves and tides to accomplish the work. However it is not one 

 of the most decided and certain examples of fiuviatile erosion. 



7. Ravine through which Agawam river flows, extending from Mount Tekoa, where 

 the river debouches into the valley of Connecticut river, nearly to the summit level of 

 the Western railroad, along the main branch of the river, about hventy-five miles. — If 

 we were to follow up any other branch of this river, we should find similar 

 ravines. The main one under consideration crosses the strata often at right 

 angles, and there is no evidence of their dislocation on either side ; hence its 

 erosion may reasonably be imputed to the river, or the ocean. It is deeper in 

 many places than the Ghor, on Deerfield river, and there are at least two cataracts 

 along its course of considerable height, where the work of erosion is going on. In 

 most places, however, it is wider than the Ghor, admitting of farms and villages. 

 There is scarcely any part of it that presents walls of rock so obviously eroded as 

 at Tekoa, where the river emerges into the alluvial plain of Westfield. 



From the fact that an enormous vein of granite is seen in the bed of Agawam 

 river in several places, as at Salmon Falls, I have suggested in my Final Keport 

 on the geology of Massachusetts, p. 691, that it might once have extended through 

 a great portion of this ravine ; and if so, that it gives the reason why the river 

 chose this track : because such a vein would be more easily worn away than the 

 mica slate. I still think that in this way we may account for a part of the ero- 

 sion : but I have not found the evidence that the vein occurs through any con- 

 siderable portion of the river's course. 



8. Ancient bed of Agawam river in Russell. — This is a well marked example, 

 lying immediately north of the railroad station in Russell. Standing at that spot, 

 and looking north, you have before you a rocky hill, several hundred feet high, on 

 the right or east side of which the river and the railroad now run. But on the 

 left side, the common road passes through a valley about as wide as the river, and 

 filled to a considerable height with terrace materials, gravel and coarse sand, at the 

 north end, but finer towards the south. Near the north end the road attains an 

 elevation above the present river a few rods further north of 74 feet. This is the 

 present height, or nearly so, of the old bed of the river above the existing stream. 

 But upon both sides of the old bed, the steep hills are fringed with the remnants 

 of a former terrace, rising 208 feet above the river; and this doubtless filled the 



