522 
CCELENTERA TES. 
the steep slope, and surrounds the land which has risen above the sea, this terrace 
being exposed at low tide. 
We have still to mention some of the causes of modification in the form and 
growth of the coral-reefs. The presence of harbours in reefs and atolls can, as a 
rule, be traced to the tides and to local ocean currents. There is generally an 
outflow through the canals and openings in the reefs. This is apparently due to 
the fact that water is constantly being thrown by the larger waves over the lower 
portions of the reef into the canal or lagoon, and seeks either to escape as an under¬ 
current in opposition to the flood-tide, or else strengthens the ebb-tide. These and 
other similar disturbances of the water in the canals bring with them much coral 
detritus, and render the bottom altogether unsuitable for the growth of corals. 
Where such currents are strong, they keep the canals clean and open. The action 
of the oceanic currents is often increased by the fresh waters coming from the 
central islands, and harbours are therefore very often found at the mouths of 
valleys and of their small streams. The influence of the fresh water itself on the 
corals is not so great as is usually assumed, chiefly because it, being lighter than 
salt water, flows away on the surface of the latter and hardly touches the animals 
which grow below the surface. 
The form of the reef is again largely influenced by the form and constitution 
of the sea-bottom. Where deep submarine fissures occur, dipping down below the 
level at which the corals flourish, no reef can be formed, as also in places where 
firm ground alternates with sand and mud. All irregularities in the outline of a 
reef or an atoll, and the formation of harbours in coral islands, can thus be simply 
explained. 
The most important point which needs elucidation is why some reefs encircle 
islands as a fringe extending from the shore, while others run parallel with the 
land, no longer touching it; others, again, forming circular lagoons with no island at 
all in the middle. This was the question which puzzled the first discoverers of 
reefs, and at one time it was supposed that instinct guided the animals in giving 
their structures the form best suited to withstand the force of the waves. Accord¬ 
ing to another hypothesis, put forth by Steffen in 1822, the reefs represent the 
summits of volcanic mountains, the crater being filled by the lagoon, while the 
channels through the reef indicate the points at which the edge of the crater was 
destroyed by outbursts of lava. This superficially plausible view was disposed of 
by Darwin thirty years ago. He argued that the volcanic cone thus assumed must 
either once have stood upon dry land and then have been submerged, or else must 
have been formed beneath the sea. In the former case, the crater would in almost 
all cases have been destroyed during the gradual sinking; while the formation of 
craters by submarine eruptions and their subsequent elevation is hardly conceivable. 
