110 THREE CRUISES OF THE ^' BLAKE." 



The islands to the north of Guadeloupe form two parallel 

 chains, the western consisting of Saba, St. Eustatius, St. Kitts, 

 Nevis, Redonda, and Montserrat, all of which are volcanoes of 

 postpliocene date ; while to the eastward is a chain of volcanoes 

 of tertiary age, — Sombrero, Anguilla, St. Martin, St. Barthe- 

 lemy, Barbuda, and Antigua. At Guadeloupe the recent islands 

 are directly united with the volcanic chain, and the still more 

 modern limestones are found on its western shores. 



The miocene rocks of the greater West India Islands^ have 

 been, as far as observed, but little disturbed. The cretaceous 

 beds, however, are found greatly modified. 



The geological history of Florida is in striking contrast with 

 that of the greater and lesser West India Islands. The share 

 which modern limestones have had in building up a portion of 

 the peninsula during the most recent geological period has been 

 described in the chapter on the Florida Reefs. From the obser- 

 vations of Conrad, of E. A. Smith, and of E. Hilgard, it would 

 appear, as is stated by Smith,^ that up to the end of the eocene, 

 the great Florida limestone plateau was still submerged ; that 

 during the time of the upper eocene, Florida was elevated 

 nearly to its present height. The axis of elevation did not 

 coincide with the present dividing ridge of the peninsula, but 

 occupied probably a position to the westw^ard. From the sub- 

 sequent deposition of sand, clay, and pebbles over Florida and 

 parts of the adjacent States, they must have been slightly sub- 

 merged again, and subsequently elevated (during the Champlain 

 period) to about their present configuration. 



In attempting to reconstruct, from the soundings,^ the state 



1 The most striking feature of the limestone was probably deposited much 



West India Islands as a whole is the axis as was the limestone that forms the back- 



of eruptive rocks, flanked by sedimentary bone of the Yucatan and Florida penin- 



formations and terraces of elevated coral sulas. This limestone was subsequently 



reefs and recent limestones. On the capped by reef-building corals during 



northern coast of Cuba the mode of for- periods of rest, followed by the elevations 



mation of the harbors within the fringing to which the successive terraces are due. 

 reef, inside of the great outside l)arrier ^ Smith, E. A. American Journal of 



reef, seems to prove that the rivers cvit Science, April, 1881, p. 292. 

 their way out as fast as the laud was ele- ^ See the accompanying maps (Figs, 



vated and the successive terraces formed. 57, 58), for which I am indebted to the 



It must be remembered that coral reefs Hon. Carlile P. Patterson, Superintendent 



have no such thickness as is seen on the U. S. Coast Survey, 

 terraces, but that the greater part of this 



