Vol. 57.] PHYSICAL DEVELOPMENT OF ANGFILLA, ETC. 523 



main portion of the submarine platform is here sunk to a depth 

 of 3000 feet or more. The great depth of the channel west of 

 Sombrero entirely disjoins it from the Virgin-Island mass. 



At the opposite or south-eastern extremity of the St. Martin 

 plateau, separated from the rest of it by a deep channel, is a small, 

 flat-topped, insular plateau now covered by about 200 feet of water. 

 This is an interesting subject which will be considered under 

 the head of erosion-features (p. 530). 



III. The Older Geological Formations. 



St. Martin is underlain by an old eruptive basement of green- 

 stone or dioritic porphyry, in which crystals of triclinic felspar 

 are prominent.^ Where exposed at the surface, the rock is usually 

 very much decayed, so that the best specimens of it are seen in 

 the wave-rolled pebbles, where only the more compact material is 

 found. The surface of the basement-rock may be seen along the 

 southern shore, west of Philipsburg, where domes of it unconform- 

 ably underlie the altered stratified beds above. The old igneous 

 rocks occur at the surface along the road north of Philipsburg, 

 which passes over a divide at 240 feet above the sea. It is here 

 decomposed into a grey granular matrix, holding residual spherical 

 masses of the less decayed material. Along the coast, south-east 

 of Philipsburg, on the way to Point Blanche, is a light-coloured 

 greenish mass of decomposed volcanic rock of a somewhat different 

 appearance from the igneous basement-rock seen elsewhere. Mr. 

 Cleve regarded it as a dyke, or intrusion, beneath the overlying 

 formation. Whether a laccolite, or part of the older basement, it 

 is very old. 



The volcanic basement may be seen beneath some of the isolated 

 ridges rising out of the flats which separate Simson's Inlet from the 

 sea. Loose blocks of the same rock, occurring in the valley of Cul 

 de Sac and elsewhere, have been derived from the central mountains, 

 now mostly covered by another formation, or else by decayed 

 material forming the soil which is creeping down the hillsides. 



The overlying formation is complex, consisting of altered lime- 

 stones, volcanic ashes, tuffs, and breccias, which have often 

 been highly silicified or otherwise mineralized, presenting numerous 

 variations. As this formation has been dislocated by endless faults, 

 and the surface of the island has been subjected to excessive denuda- 

 tion, breaking it into disconnected ridges, the individual beds cannot 

 be continuously traced. Accordingly, where the characters change, 

 it is possible that I have included some tuffs and breccias belonging 

 to the basement-series with the tuffs of the altered limestone- 

 series. But whether this confusion has occurred or not, we find 

 the series capping the mountain-ridges and also forming sea-clifl's 



^ I have not attempted petrographical accuracy, as the igneous rocks were not 

 the primary object of the study. Mr. Cleve calls them ' syenite-porphyries,' 

 the same as he found in St. Bartholomevr ; they are probably closely related 

 to the old eruptives of Antigua, 



2n2 



