Reports and Proceedings — Geological Sociefij of London. 81 



occur on the slopes descending from high plateaux. From the ends 

 of the cirques, valley-like channels can be traced landward on the 

 submerged plateaux, or can be found to cross them in order to join 

 like features on the other side. The cols between the opposite valleys 

 vary in depth from about 2,000 to 3,600 feet, except that between 

 the Grenadines and the Trinidad banks, where the divide may not 

 be more than 750 feet below the surface of the sea, and one south 

 of St. Vincent (less than 1,300 feet). Some of the submarine 

 channels have remarkable tributaries. The drowned valleys, like 

 those about the islands to the north, assume two very different 

 forms — those with broad undulating outlines, such as characterize 

 the features produced during the long Miocene-Pliocene period of 

 erosion, when the surfaces of the land were at or near the base-level 

 of erosion, and other types where very deep valleys and gorges 

 incise the more rounded features of the drowned plateau, which in 

 the early Pleistocene epoch thus appears to have stood for a limited 

 time at an altitude of 6,000 or 7,000 feet, as shown within the 

 limits of the Antillean mass (and still higher from evidence beyond). 

 There are no coastal plains, strictly speaking; only to a very limited 

 extent are the islands surrounded by shelves submerged to a depth 

 of less than 200 feet. But the Grenadine banks are extensive. 

 One or two outlying remnants of the Antillean plateau occur south- 

 east of Dominica, and another about 60 miles east of Martinique, 

 all of which may be fragments of the old coastal plains. 



All the islands are underlaid by old Tertiary or pre-Tertiary 

 igneous rocks, as in Guadeloupe. Such, where exposed, are found 

 to be very much decayed. Elsewhere, they are covered by tuffs 

 with only angular breccia. Upon such surfaces, denuded, rest other 

 tuffs derived from older deposits, containing water-worn pebbles, in 

 lines of bedding. These last may be the equivalent of the Tertiary 

 tuffs and limestones of Guadeloupe. Upon their eroded surface 

 rests a gi'avel formation, which itself has largely been washed away. 

 In the hollows of its surface is found another formation composed of 

 coral limestone, containing a fauna which still survives, with one or 

 two possibly extinct forms. The deposit occupies a position similar 

 to one in Guadeloupe and another in St. Kitts. Its surface is also 

 greatly eroded, and then covered with another stratified sand-and- 

 gravel accumulation. The surface of the slopes is often covered by 

 a loam, which, in part at least, is a land-formation. The Pleistocene 

 Period is thus seen to have been one of changing physical conditions. 

 The older Tertiary history must be inferred from that of the 

 neighbouring islands. 



Lavas may be seen underlying the gravels, and accordingly we find 

 that the volcanic activity was renewed, after a very long Tertiary 

 quiescence, in the Pleistocene Period. 



The plains underlaid by the beds of old tuffs have been so raised 

 up as to give rise to sloping terraces dipping outward from the late 

 volcanic centres, showing that their elevation has been due to local 

 uplifts, and not to regional movements, and also that to igneous 

 centres alone are confined the volcanic uplifts, which do not extend 



DECADE IV. VOL. IX. NO. II. 6 



