502 
Mr Crossland , The Coral Reefs of Zanzibar. 
The height of the raised edge is that of the original shore 
platform, the boat channel being a secondary formation. Darwin 1 
gives instances of sandstone formations which simulate coral reefs, 
e.g. a bar off Pernambuco, explaining the formation of their boat 
channels by the removal of the inner parts of reefs and sandbanks 
by the currents which return the water thrown upon them by 
the surf. The fewness of the depressions in so long a reef would 
make both these and the tidal current here exceptionally powerful. 
Near low tide while the edge of the reef is out of water the tidal 
current in the boat channel is running most strongly. Several 
times when travelling at spring tides where the boat channel is 
shallow or obstructed I have had to wait six hours for water to 
float the boat. This means that water (often muddy) is pouring 
over the reef flat for at least six hours longer than over the 
raised edge, and these hours are stated by Dana to be the period 
during which wave action is least powerful. 
This explains the absence of stones similar to those on the reef 
edge from the channel and flat. These stones must have originally 
been evenly distributed over the platform, as they are remnants 
of the hardest portions of the rock which has been removed. It 
is impossible to believe that such waves as occur on this coast 
could break into this rock, which can be quarried for building 
purposes only with difficulty. Their absence from certain areas 
of reef edge is only explicable on the former theory of their 
origin. It is hardly possible that numerous stones should be 
broken from one part of the reef while none at all are from 
others which are exposed to the same surf. But it is conceivable 
that some parts of the original reef should contain specially 
hardened portions of rock, which might be absent from other 
areas. 
It is remarkable that the reef edge should not be growing. 
All the physical conditions seem favourable to coral growths. 
Their absence from the Somali coast (north of the equator) is 
explained by the presence there of an off-shore current and a conse- 
quent rising of cold water from the depths to take its place. But here 
we have an on-shore current from the surface of the ocean under 
the equator. Mud is fatal to corals, unless kept in motion by 
a current. The bottom is visible here only to the depth of 
5 fathoms, so that the water is dirty compared to that around 
the Pacific Islands 2 . But the good effect of the current one 
would expect to neutralise the evil of the mud, as it does on the 
sand-banks at the south entrance to Kokotoni harbour. The' 
water here is positively thick, yet corals flourish, since strong 
tidal currents are continually flowing through these narrows. 
1 Corals and Coral Islands , pp. 72 and 73. 
2 Gardiner, on Funafuti, etc. 
