FORMATION OF CORAL RFEFS AND ISLANDS. 211 



tions of the reefs, and being sometimes increased in force by 

 the contributions of island streams, which add to the detritus 

 and to the weight of accumulating waters. 



3. Harbours which receive fresh-water streams, or submarine 

 springs of fresh-water, are more apt to be clear from sunken 

 patches ; and the same causes keep open shallow passages to 

 the shores, where there are shore reefs. 



It should be remembered, that while the effects from fresh- 

 water streams are so trifling around islands, they may be of 

 very wide influence on the shores of the continents where the 

 streams are large and deep, and transport much detritus. 

 This point is illustrated further on. 



II. ATOLL REEFS. 



The remarks in the preceding pages respecting reefs around 

 other lands apply equally to atoll reefs. There are usually 

 currents flowing to leeward through the lagoon, and out, over, 

 or through the leeward reef, the waves with the rising tide 

 dashing over the windward side, and keeping up a large 

 supply, which is greatly increased in times of storms ; and this 

 action tends to keep open a leeward channel for the passage 

 of the water. This is the common explanation of the origin 

 of the channels opening into lagoons. These currents are 

 strongest when a large part of the windward reef is low, so as 

 to permit the waves to break over it ; and the coral debris they 

 bear along will then be greatest. When a large part of the 

 leeward reef is under water, or barely at the water's edge, the 

 waters may escape over the whole, and on this account large 

 reefs sometimes have no proper channels. When the land to 

 windward becomes raised throughout above the sea, so as to 

 form a continuous barrier which the waves cannot pass, the 

 current is less perfectly sustained, since it is then dependent 

 entirely upon the influx and efflux of the tides ; and the lee- 

 ward channels, in such a case, may gradually become closed. 



The action of currents on atolls is, therefore, in every way 

 identical with what has been explained. The absence of 



p 2 



