24 THE BERMUDA ISLANDS. 



An open continuation of this outer water is the Great 

 Sound, which is to an extent land-locked by the "hook" 

 of Main Island, and its disrupted continuation, Somerset 

 and Ireland Islands. At the eastern end of the archipelago 

 an incursion of the southern waters has formed, or helped to 

 form, the lagoon known as Castle Harbor, an extensive 

 body of water, with a depth of from 30 to 40 feet, whose 

 oceanic boundaries are well seen in Cooper's, Castle and 

 Nonsuch Islands, and their dissociated fragments. Castle 

 Harbor stands also in direct communication with the northern 

 waters by means of one or more channels, known as "The 

 Reaches," which are in part largely silted and coral-grown, and 

 consequently difficult of passage. 



Harrington Sound, which bites into tlie Main Island alone, 

 is the most nearly, land-locked of the inner waters, and at the 

 same time the deepest water in the archipelago. The average 

 depth is probably not less than nine or ten fathoms, and our 

 line frequently dropped to 12 fathoms. We sounded 14 

 fathoms in the southern bay opposite the Devil's Hole, and I 

 was informed that 16 fathoms had been obtained in the same 

 locality. Although two miles in length, and approximately 

 a mile and a half in greatest width, this extensive body of 

 water communicates with the outer sea by a channel not more 

 than 50 feet in width, the Flatts Inlet. The breaking action 

 of the waters, the undermined ledges, and the vertical cliffs all 

 clearly indicate that the Sound is still expanding, and it is 

 merelj' a question of time, if the present conditions continue, 

 when it will be more in the nature of an open bay than of a 

 land-locked lagoon. 



The reefs on the south side of the archipelago approach in 

 places to within a hundred yards or as many feet — or even less — 

 of the chain of islands, from which they are separated by a 

 belt of water of no inconsiderable depth, and always existent. 

 They are, therefore, more nearly in the nature of "barrier" 

 than of " fringing " reefs, if, indeed, they can be said to strictly 

 belong to either one of these two divisions. Opposite the open 



