THE BUILDING OF AN ISLAND. 79 



The islands of St. John and Tortola, to the east of St. Thomas, much re- 

 semble the last named in their rock structure, only that the strike of the strata 

 lies east and west. Virgin Gorda differs mainlv in exhibiting a large area of 

 unstratified diorite, which is often mistaken by visitors for granite. 



Anegada, which ends the chain, is, however, quite different from the other 

 Virgin Islands, being a flat island entirely composed of limestone of quite re- 

 cent date, as shown by the fact that it contains onlv shells of species which 

 still live in these waters The Bahamas, which are commonly regarded as a 

 continuation of the Caribbean axis, are of similar structure. 



Before turning west to the large islands, it will be interesting to notice 

 that, judging from Professor Cleve's observations in Anguilla, St. Martin's 

 and St. Bartholomew, those islands belong, not as we might have expected, to 

 the great volcanic axis, but to our own east and west axis. He says, namely, 

 that the dips of the rocks in St. Martin's and St. Bartholomew are to the south ; 

 and about Anguilla he reports that in the abrupt cliffs of the northern coast 

 the rocks on which the limestones of the island rest may be seen ; it is a kind 

 of trap. From these observations it would appear likely that there has been a 

 continuation of the great axis eastward and that this has been connected with 

 the pressures which have raised those three islands. 



When, however, we leave the bank on which the above islands stand, and 

 arrive, going southward, at Antigua, the case is quite different.* 



In Antigua the tilt of the strata is from the southwest and they dip away 

 to the northeast. Antigua is, therefore, geologically a dependencv of the north 

 and south axis. The island shows three sets of rocks, two of them of ultimate 

 volcanic origin, the other consisting of limestones and marls originating in the 

 ocean. As in St. Croix, the older rocks have in a past age sunk into the sea 

 and have had the limestones deposited upon them. There are, however, con- 

 siderable differences. The older formations do not appear to have been raised 

 into ridges as in St. Croix, and the limestones have, therefore, not been laid 

 down on the edges of the older formation, but on the surface of their upper- 

 most beds. Another great difference is that the upper part of the older rocks 

 (the second of the two groups) is all clearly stratified, consisting of regular 

 beds, mostly of hardened clavs and sandstones, and has not undergone any 

 marked alteration as a mass. On the other hand, some of the beds have been 

 silicified in a remarkable degree. The conversion into silica has taken place 

 in some of the marine limestones interstratified with the clays, in which we 

 may find corals and shells partially or entirely converted into flint ; it has also 



* Reference to a chart of the West Indies shows that Antigua and Barbuda, its dependency to the 

 north, both stand on a bank lying north and south. Barbuda is a flat island, having an elevation along 

 the east shore known locally as the Highlands, the highest point of which is given in the chart as 115 feet. 

 The rocks in that part probably belong to the same formation as the limestone rocks in the north of Antigua, 

 all the rest of the island is made up entirely of a recent limestone, full of shells of the same species as now 

 live in the adjacent waters, and to be compared with the recent limestone of Fredenksted. The writer 

 has seen there the ruins of an old castle, all the stones of which had been cut from the limestone of the flat, 

 and were full of small shells which could be matched from the shore of the neighbouring lagoon. 



