542 PEOF. J. W. SPENCER ON THE GEOLOGICAL ANB [NoV. I90I, 



the old Oligocene limestone still remaining) as land-surfaces during 

 the greater part of this long era. The subsidence of the land being 

 renewed, its surface was covered by a mantle, from 15 to 30 feet in 

 thickness, of marly limestones, containing shells of living species, 

 but not until after the late Tertiary volcanic activity had com- 

 menced. These phenomena would not have been seen, had not 

 Brimstone Hill and The Quill been subsequently thrust up by local 

 eruptions of a still later date. While the fossils do not indicate 

 to which recent epoch they belong, their considerable age is shown 

 in the erosion-features of the surface of the banks, and their mar- 

 gins, which are indented by the very deep valleys described. These 

 marls are regarded as of the same age as the upper marls of 

 Anguilla, of Sombrero, of Guadeloupe, etc., which are correlated 

 with the Lafayette formation of the American continent, or belonging 

 to about the commencement of the Pleistocene Period. 



This epoch of subsidence was followed by that of the somewhat 

 long and stupendous elevation and erosion of the region, when the 

 deep valleys and cirques, indenting the margins of the tablelands, 

 and the relatively broad but shallow channels upon their surfaces, 

 cut back from their edges, were produced by atmospheric agents 

 during the early part of the Pleistocene Period. From the local 

 evidence, the elevation does not appear to have been more than 

 about 3000 feet above the present surface. But it should be noted 

 that a vastly greater altitude is indicated beyond this area.^ The 

 great mass of the volcanic ridges seems to have been built up or 

 elevated daring this long epoch, and largely completed before the 

 next epoch, which was one of subsidence. 



This depression does not seem to have exceeded 300 feet below 

 the present level, when the gravels, sands, and deposits con- 

 taining living species of shells, now found at this height, were 

 accumulated. These gravels and sands occupy the same position as 

 those of Cassada Garden in Antigua, the Upper Petit-Bourg Series of 

 Guadeloupe, and the gravels of St. Martin, all of which are correlated 

 with the Columbia formation of the American continent, which 

 belongs to the mid-Pleistocene Period. It was during this epoch of 

 subsidence that the domes of Brimstone Hill and The Quill of Statia 

 appear to have been formed. The succeeding upward movement 

 carried the land to 60 feet or more above its present height ; 

 then the ravines, now buried, and the small channels in the sunken 

 shelf were excavated. Again the land was slightly depressed, to 

 40 or 50 feet, and the ravines just referred to were refilled. There 

 is a corresponding formation in the other islands. 



The re-elevation which raised these sediments to the height of 

 40 or 50 feet, is the last change noted, but as the coral-reefs along 

 the coast are not elevated, it is possible that a downward movement 

 is in progress. 



All the oscillations of level are recorded in the adjacent islands, 



^ See ' Eeconstruction of the Antillean Continent' by the present writer, Bull. 

 Geol. Soc. Am. vol. y\ (1895) pp. 103-40. 



