140 REMARKS AND OPINIONS. 



RADACK, RALICK, REPITH-URUR, BOGHA, THE 

 CORNWALLIS' ISLANDS. 



We had opjDortunities at Radack to examine more 

 closely the formation of the low Coral islands, 

 and to illustrate and correct our previous observ- 

 ations on this subject. 



We conceive an island-group of this formation- 

 as a mass of rock, which rises with perpendicular 

 walls from the unfathomable depth of the ocean,, 

 and forms on the surface an overflowed i)lateau» 

 A broad dam, constructed by nature round the 

 edge of this plateau, changes it into a basin. 

 This dam, or reef, is generally on the side of the 

 circumference turned to the wind, and at the time 

 of ebb rises like a broad causeway above the 

 water. On this side, and particularly on the pro- 

 jecting angles, the most islands collect on the 

 back or ridge of this dam. On the lee side, on 

 the contrary, it generally keeps under the water. 

 It is there partially interrupted, and its intervals 

 are sometimes broad enough to admit even large 

 ships, which sail with the current into the internal 

 basin. Within the inside of this entrance fre- 

 quently lie single rocky banks, like fragments of a 

 broken wall, or indications of it. Other similar 

 banks lie scattered here and there in the interior 



