PACIFIC AND BEERINC'S STRAIT, 247 



CHAP. 

 VIII. 



Feb. 



by slow degrees. Thus the sand dispersed over 

 the lagoon indicates a period when the sea rolled 

 entirely over the reef, tore up blocks of coral jaJJ. 

 from its margin, and by constant trituration ground 

 them to powder, and finally deposited the par- 

 ticles where they now rest. The bank near the 

 lake must have originated at a subsequent period, 

 when the outer edge becoming nearer to the surface, 

 moderated the strength of the waves, and the wash 

 of the sea reached only far enough to deposit the 

 broken coral in the place described. At a still less 

 distant period, when the island became dry, and the 

 violence of the sea was wholly spent upon its mar- 

 gin, the coral, which had before escaped by being 

 beneath the surface, gave way to the impetuous 

 wave, and was deposited in broken masses, which 

 formed the high ridge. Here the sea appears to 

 have broken a considerable time, until a second 

 ledge gradually extending seaward, and approaching 

 the surface, so lessened the effect of the waves upon 

 this ledge also, that they were again only capable of 

 throwing up an inferior heap similar to the one first 

 mentioned. In process of time this outer ledge will 

 become dry, and the many large blocks of coral now 

 resting near its edge will, probably, form another heap 

 similar to the large one ; and thus the island will 

 continue to increase by a succession of ledges being 

 brought to the surface, while, by the same process, 

 the lagoon will gradually become more shallow and 

 contracted. 



The ridges are particularly favourable to the for- 

 mation of a soil, by retaining within them whatever 

 may be there deposited until it decays, and by pro- 

 tecting the tender shrubs during their early growth. 

 Near our observatory the soil had attained a depth 



