40 Mr Crossland, The Coral Reefs of Pernod Island 



explained by the fact that Zanzibar is itself a part of the 

 barrier system, whereas Pemba is independent of the mainland 

 formations. 



The forms assumed by portions of the barrier are interesting. 

 In several cases an approximation to an atoll is found, as in the 

 following sketches of reefs from Zanzibar channel. The fact that 



MBWAKUNl 

 10 





Kl KWXRO 



solution alone has formed these rings from originally plane areas 

 supports strongly Gardiner's theory that it is to solution that 

 atolls owe their lagoons. 



In two places, viz. to the north of Mombasa, and at Chale 

 point, near Wasin, the barrier for many miles becomes nearly as 

 straight, narrow and uninterrupted as an artificial breakwater, 

 enclosing an equally regular channel two or three fathoms deep at 

 low tide and about one-third of a mile wide. I had no opportunity 

 of visiting Chale, but I examined the exactly similar reef situated 

 immediately to the north of Mombasa harbour. Its surface, 

 alike on both sides, is very regularly rounded, loose stones and 

 hummocks continuous with the rock beneath occurring rarely. 

 Coral is entirely absent, and nullipores are almost so, the surface 

 being covered by brown filamentous Phaeophyceae. The reef is thus, 

 as in the other cases examined, formed entirely of dead rock, and 

 probably at one time it extended above the level of high tide. 

 The channel contains muddy water with blocks of living Porites 

 at long intervals. The bottom is largely covered by a marine 

 phanerogam which usually grows on rock. 



