20 PROCEEDINGS OF MADISON MEETING. 



until, as in the case of Bartlett deep, south of Cuba, the extreme depth of over 

 20,000 feet is reached. All of these valleys are joined by numerous tributaries. 



Such channels as the straits of Florida are only the junctions of valleys trending 

 in opposite directions, even the common watershed of which is submerged. While 

 the author regards all of these submerged valleys as originally terrestrial, yet he 

 accedes to any hypothesis that may call for their enlargement by marine currents, 

 which, however, do not appear to be scouring agents below the rims of the sub- 

 merged basins ; as, for example, the valley extensions of the Floridian straits, 

 below the depth of water in its shallowest portions (about 2,000 feet), which have 

 not been materially affected by the marine currents. This hypothesis is supported 

 by temperature soundings. 



The submergence indicated by the channels means great epeirogenic or extensive 

 continental land-movements. The oldest formations of Cuba, Haiti, Trinidad, 

 Barbados, etc, appear to be no older than the Cretaceous period. The oldest sedi- 

 ments are shales, sandstones and conglomerates (even at the base of the group in 

 Santo Domingo) showing former continental connections, which would demand the 

 enormous changes of level at an earlier period than that of which this paper treats. 

 With this connection or the mountain-building of the early Tertiaiy, or even that 

 after the Miocene period, the author is not concerned, but with those quieter ex- 

 tensive regional or epeirogenic movements which have slowly submerged great 

 areas without obliterating the former land topography. The time of greatest 

 land development was probably after the Pliocene period, if, according to McGee, 

 we regard the Lafayette formation of that age. Into the discussion of the exact 

 age, however, we cannot enter until further investigation. For the present, let 

 us say that this great continental extension was since the later Tertiary times and 

 was followed by movements in both directions, for modern deposits have been 

 lifted to considerable elevations on the islands. This late continental elevation is 

 not disproved by biologic evidence, but rather sustained. The drainage of this 

 extensive land area was largely into the Pacific, or its embayments extending across 

 portions of Central America and Mexico and indenting the western coast of the 

 Antillean lands. The watershed between the Atlantic and Pacific drainage is still 

 represented by the mountains of Cuba, Haiti and the Windward islands. They 

 are situated near the Atlantic border of the continent, just as are the Appalachians 

 of the southern states. 



The closing of the western drainage involves recent tilting or regional uplifts 

 forming the western rims of the Mexican and Caribbean basins. Similar move- 

 ments have been demonstrated by the author in connection with the barriers of 

 the Laurentian valley now producing the Great Lakes, which have been obstructed 

 by the land rising along an axis transverse to that of the outlet of the old valley. 



What has been the recent mean depression of the Antillean continental exten- 

 sion? From the depths of the fjords cutting through the continental plateau a 

 subsidence of from 8,000 to 12,000 or possibly 14,000 feet is apparent ; but probably 

 this epeirogenic ijiovement was not originally uniform, or, if so, it has been partly 

 deformed by more recent movements, so that the author regards the 20,000-foot 

 abyss (Bartlett deep), just south of Cuba, and even part of the Haitian channel 

 north of that island, as partly due to a downward local warping of the deeper 

 portions of the valleys. In the case of the Bartlett deep a basin has been produced. 

 This great continental depression diminished to the north, so that the .southern 

 states have only been partly submerged. The present elevation of Central America 

 represents complex movements. 



