50C 



BARRIER-REEFS 



building corals, which cannot li\-c at great clcptlis, based their 

 encircling structures? This is a great apparent difficult)-, 

 analogous to that in the case of atolls, ^^•hich has generally been 

 o\-erlooked. It \\-\\\ be perceived more clcarh" by inspecting 

 the following sections, -which are real ones, taken in north and 

 south lines, through the islands \\'ith their barrier -reefs, of 

 \'anikoro. Gambler, and i\Iaurua ; and the}' are laid down, both 

 \'ertically and horizontall_v, on the same scale of a quarter of an 

 inch to a mile. 



It should be obser\-ed that the sections might ha\'e been 

 taken in an\- direction through these islands, or throu""h man\- 



WB/m 



I, Vanikorn. 2. Orimbier Islands, 3. I^faurua, 



The hori/oiital shading shtn\s the barrier-reefs and higoon-channels. 'rlie inclined sli.ading abo\"c 

 the level of the sea (AA) shows the actual form o( the land ; the inclined shading below this liiic 

 shows its probable prolongation tnider water. 



SECTIONS CiF L..\KRIEK-KhEFS. 



other encircled islands, and the general features would hax'e 

 been the same. Xo\v bearing in mind that reef-building coral 

 cannot lix^e at a greater depth than from 20 to 30 fathoms, and 

 that the scale is so small that the plummets on the right hand 

 show a depth of 200 fathoms, on what are these barrier-reefs 

 based ? Are we to suppose that each island is surrounded by 

 a collar-like submarine ledge of rock', or b\' a great bank of 

 sediment, ending abruptly where the reef ends ? If the sea had 

 formerl}' eaten deeply into the islands, before the\' were 

 protected b\' the reefs, thus having left a shallow ledge round 

 them under water, the present shores would ha\-e been 

 in\-ariably bounded b\- great precipices ; but this is most rarely 

 the case. ^loreovcr, (ni this notion, it is not possible to explain 



