218 GORGES. 



Hampshire (American Journal of Science, Vol. IX, N. Series), that prohably trap dikes once occupied 

 the clefts, we shall equally need some agency to remove the materials. 



2. Willoughby Lake. So obvious is it that this lake six miles long and one mile broad, and walled in on 

 two sides by rocky ridges 1800 feet high now occupies the place once occupied by the mountain, that no 

 one, however little acquainted with geology, doubts that the two mountains were once united, and that 

 they have either been separated or worn away. It is the most common explanation to say that the gulf 

 resulted from some "violent convulsion of nature," which is the usual resort for any unusual phenomena in 

 nature. But I could discover no evidence of any such disruption of these mountains. But the fact that 

 here is the summit level between the waters of the Passumpsic on the east, and Lake Memphremagog on 

 the northwest, makes it most probable that like Dixville Notch and some other cases to be described this 

 is a valley of erosion of the peculiar kind called Purgatory. Our theory of the mode in which they were 

 formed will be stated after a few more examples are given. 



A fact in respect to Modified Drift at the two extremities of the Willoughby gorge deserves notice. 

 Irregular hills of gravel and sand there rise, some of them 100 feet high. Yet the borders of the lake 

 show nothing of this kind above the waters, and, as these are said to be very deep, we doubt if the modified 

 drift is there. Did it fail to lodge in the gorge and accumulate at the extremities, because the waters of 

 the ancient ocean rushed through the opening with violence, or too quietly to carry the detritus with them ? 



3. The gorge between Lowell and Eden. The Missisco Eiver takes its rise in the south part of Lowell 

 and runs nearly north through Lowell and Troy, and from thence curves around the north end of Jay Peak 

 and runs westerly to Champlain. Passing from Lowell to Eden, we follow a principal branch of this river 

 to its source, which is the summit level between the Missisco running north, and a branch of the La Moille, 

 running south. Here is a gorge of considerable depth, excavated in solid rock, having the same general 

 aspect as Willoughby and others, where water and ice were the probable eroding agents. There is a pond 

 a little south of the summit at Eden. At its north end we find modified drift, which grows coarser as we 

 approach the gorge. This fact seems to indicate that the current which brought in this detritus came 

 from the south. But this point needs further examination. 



South of Eden, the valley through which the current from the Troy valley might have run was perhaps 

 that through which the La Moille now flows. Or it might have passed across from Hydepark through 

 Stowe, &c., to the White Eiver at Waterbury. 



4. Gorge near Eagle Ledge, near where the corners of the four towns of Worcester, Calais, Woodbury and 

 Elmore unite. The rock is here excavated to a considerable depth, and on the north a stream runs into the 

 La Moille, and on the south the principal branch of the Winooski takes its rise, so that these two rivers 

 were probably once connected, or rather the two bodies of water that once occupied the depressions which 

 we now call the valleys of La Moille and Winooski ; for the gorge near Eagle Ledge (we are not sure in 

 which of the four towns it lies) is 800 feet above the present bed of the La Moille. 



5. The Williamstown Gulf. If we pass up the Stevens branch of the Winooski to Williamstown, we 

 shall find a pond in modified drift at the summit level, whence the second branch of White River rises and 

 proceeds southerly. At this point is no excavation in rock, but a mile south is the gulf, cut not less than 

 100 feet deep through the ledges. Through this gulf the Winooski most probably once passed southerly, 

 and was a tributary of White River. But we regret that our observations on this route have been so 

 defective. 



6. The Roxbury Summit Level. This is the highest point on the Vermont Central Railroad, and is 

 something like a hundred feet higher than the summit level just described in Williamstown. There are no 

 striking gorges near the summit, yet a practiced eye perceives certain marks which indicate the valley to 

 have been one of erosion. In descending towards the Winooski, some pot-holes occur a few miles south of 

 Montpelier, indicating the former existence of a cataract. 



7. Summit level between the Battenkill and Otter Creek in East Dorset. If any one takes the Western 

 Vermont Railroad from Rutland and goes southerly, he will find himself soon in a deep narrow valley, 

 having the Green Mountains on his left and a part of the Taconic range on his right. At East Dorset 



