CENOZOIC TIME — QUATERNARY. 949 



or other detritus, so that the actual depth of excavation may much exceed 

 that obtained by soundings. 



From such facts it is reasonable to estimate the elevation of portions of 

 British ISTorth America along the Canadian watershed, or the great Ice- 

 plateau, to have been at least 3000 feet above the present level. This 

 subject has been recently well discussed by Upham, with this estimate as 

 his conclusion. The author, in 1871, suggested 5000 feet, and this may not 

 be too high for some portions of the Canadian region of highest ice. With 

 3000 feet for the Canada watershed south of Hudson Bay, this bay must 

 have been largely dry land. Along the coast of Maine the elevation indi- 

 cated is less than a thousand feet. 



South of Maine, on the New England coast, other fiord-like indentations of the coast 

 exist in Narragansett Bay, R.I., and the gorge of the Thames, from New London to 

 Norwich, Conn. Besides these, there are pot-holes in the gneiss of islands off the 

 Connecticut coast ; and those of the Thimble Islands, in the bay of Stony Creek, show that 

 this bay was formerly crossed and probably excavated by a freshwater stream. The great 

 depth of the bays on the north side of Long Island, 50 to 65 feet notwithstanding the 

 later drift deposits over the region, is further proof of elevation. The amount for southern 

 New England and Long Island could not have been less than 150 feet (D., 1870). With 

 this elevation. Long Island Sound in the Ice period would have been, instead of an arm of 

 the sea, the channel of a river tributary to the larger Connecticut River ; and Long 

 Island with New York on the west side and the south coast of New England on the east 

 would have been continuous dry land. (See map, page 18.) The soundings of the Sound. 

 and of the waters south of Long Island are shown on this map, and also more fully in A7n. 

 Jour. Sc, xl., 1890, with explanations in the same volume. 



If the fiords of the coast are proof of elevation, the absence of them farther south 

 should be probable evidence of little elevation or none. The submarine Hudson River 

 channel (map, page 18) indicates a former emerged condition of the sea bottom, requiring 

 an elevation of the region and the adjoining coast of 2800', judging from the deepest part ; 

 and it has been inferred by Lindenkohl and Upham that this elevation took place in the 

 Glacial period. But the facts from the New England coast indicate only small elevations. 

 Moreover, the origin of the submerged Hudson River channel appears to have been of 

 much earlier date, as has been explained on page 744. 



J. W. Spencer has inferred from the Coast Survey maps that there are submarine 

 river channels off the mouths of several of the rivers of the coast south of Cape Hatteras, 

 and in the Gulf of Mexico, the Mississippi included. But no satisfactory evidence of such 

 channels exists on these charts, in the opinion of officers of the Coast Survey. 



G. M. Dawson states, with reference to the fiord region of western 

 America, that the land in the Pliocene stood relatively to the Pacific about 

 900 feet higher than now ; and he concludes that the fiords were shaped and 

 enlarged locally during the following Glacial period, when the amount of 

 elevation was still further increased. The submerged river channels of the 

 Pacific coast of North America, on the coast of California, as described by 

 G. Davidson (1887), descending to depths of 2400, 3120, and 2700 feet, 

 indicate a higher level of the region of 2500 to 3000 feet, and probably during 

 the Glacial period. 



