ON TllK GEOLOGY OF BARnADOS. 215 



niiraorous and more completely calcified, so that they do not show 

 the open structure of those in the softer rock ; this is probably the 

 outer part of an actual reef. Still higher, at a level of 180 feet, 

 there is a cutting and a (juarry exposing rock which is rather harder 

 than that first seen, but not so hard as the intermediate beds. 

 Like that near Ceres, it consists of coral-debris with many broken 

 branches of coral, but few coral masses ; the rock is partially con- 

 solidated, but is still very porous, and can be worked as a free- 

 stone. 



At the next hill, near Jackson's estate (300 feet), the rock is 

 much harder again, and contains some thin beds of grey compact 

 limestone. Towards Exchange, near the 400-feet level, there is a 

 cutting in rough coarse-grained rock which resembles that at 186 

 feet, but shows many open hollows caused by the solution of some 

 of the pieces of coral, apparently the cylindrical branches of Forites, 

 while the other corals are altered and partially filled or replaced 

 by calcite. 



A similar change takes place about the same level on the road 

 near Endeavour and Locust Hall. At DayrcU's Hill (300 feet) 

 there is a deep cutting in rough rock like that above mentioned, and 

 many of the smaller cylindrical corals are dissolved. In the next 

 cutting (about 400 feet), near Locust Hall, this is still more marked ; 

 nearly all this kind of coral has disappeared, and the other corals 

 seem to be casts in calcspar, while the surrounding matrix is harder 

 and more com])act than the newer limestones at lower levels. 



Similar rock to the last is seen at intervals between 500 and 700 

 feet, as at Market Hill. In a quarry near Eisher Pond (just above 

 the 700-feet level) the rock exposed is very white and hard, and 

 traversed in all directions by ramifying perforations which are 

 doubtless the spaces left by the solution of the branching corals. 

 This rock had a certain resemblance to some parts of the Chalk 

 Rock of Enoland, in which similar ramifying perforations often 

 occur near the surface and are due apparently to the solution of 

 soft mealy chalk filling the spaces originally occupied by the roots 

 and stems of siliceous sponges. The other corals in the rock at 

 Eisher Pond were casts in calcite of the spaces between the coral- 

 lites, the coral itself having been dissolved, and in some the spaces 

 left by the solution of the original coral were also partially filled 

 with calcite, and the whole rock has been so calcified and hardened 

 that it is broken up and used for road-metal. 



Erom Eisher Pond and Cole's Cave up to the highest plateau of 

 coral-rock between Castle Grant and Bloomsbury similar hard and 

 altered coral-limestone is found. At some places, at a depth of 

 several feet below the surface, remnants of organic structure were 

 found in the holes, looking more like sponge-fibres than coral struc- 

 ture ; but the fibres are probably casts of the small tubular spaces 

 which can be seen in a broken branch of Forites^ and one specimen 

 sliows the cast of a Vermetus attached to the fibres. 



A small quarry close to Castle Grant exposed a rock which con- 

 tained a larger numl)er of corals than were seen in any of the places 



