2i 4 THE POPULAR SCIENCE MONTHLY 



have cut three ravines down the slopes, and in former times, before 

 the coral reef grew seaward to protect the island, the sea commenced to 

 cut into the shores, forming a precipice about 20 feet high along the 

 southeastern side of the island. 



But, after the volcanic fires had ceased, corals began to grow along 

 the shores and soon the island was surrounded by a fringing reef. The 

 steep seaward slope and outer edges of this reef provided the best foot- 

 hold for the growth of corals, and no sooner did the old ones die than 

 new growths took possession of the coveted space and thus the reefs 

 pushed seaward from the shores and the old volcano was protected from 

 the attack of the breakers which now break impotently upon the outer 

 edge of the ever-widening reef. 



As is always the case under these conditions, the reef advanced sea- 

 ward most rapidly against the wind, for corals thrive best in agitated 

 water which is free from silt. Thus the reefs are three times as wide 

 on the southeastern as on the northwestern sides of the island, for not 

 only is the water too calm for the most luxuriant coral growth on the 

 leeward side, but the mud which is washed over the reef-flats during the 

 ranry season interferes seriously with coral growth. It is due also to 

 this fact that the reef — or platform — is narrow wherever the silt from 

 the streams washes over it, and wide wherever it is covered with pure 

 ocean water and exposed to the full force of the breakers. 



Maer Island is oval, about three miles long and one and a half wide 

 and the southeast trade-wind causes the water currents to sheer in oppo- 

 site directions from the middle of the southeast shore, and eddies are 

 formed at both ends of the island, and these deposit water-washed sand 

 along the leeward side, thus forming dunes at the ends and a sand 

 beach along the middle of the leeward side, and converting the originally 

 oval shape of the island into a crescent with its horns and concavity 

 directed down the wind. Indeed Hedley and Griffith-Taylor, and Wood- 

 Jones have shown that the crescentic shapes of coral atolls are formed in 

 this manner in obedience to the direction of the prevailing winds and 

 currents. 



In short, the width and character of the reefs surrounding Maer 

 Island are determined by the shore conditions ; and it is quite clear that 

 the island has not developed in the midst of a preexisting reef-flat, but 

 that the volcano was formed before the modern reefs began to grow 

 around it. Moreover, they are all fringing reefs that have grown out- 

 ward from the shore and thus the " lagoon " of the great southeast reef 

 is only 18 inches deep and its bottom is of hard coral with none of the 

 mud and occasional reef patches seen in the bottoms of all lagoons be- 

 tween barrier reefs and the shore. 



Indeed according to a theory which has been put forth by Penck, 

 and more recently by Professor R. A. Daly, barrier reefs originated 



