INTRODUCTION. 95 
Now, while this is the way in which the theory accounts 
for the formation of barrier reefs surrounding islands, and 
of atolls where no island remains, it is equally applicable 
to barrier reefs extending longitudinally for many miles 
along the coast of a continent at a considerable distance from 
the shore, for at first there were fringing reefs close to the 
land, but when by upheaval the shore became a raised beach, 
the reefs, when built up by the polypes, were far out at sea. 
The theory accounts satisfactorily also for those gaps or 
gateways, one or more of which are found in every barrier 
reef or atoll. The existence of such gaps might at first 
seem fatal to the theory, for it might be said, had the po- 
lypes reared the reef from the foundation, they would have 
carried on their work uniformly, and would have left no 
gaps; and yet had there been no gaps no vessel could ever 
have fled for refuge into these lagoons of broad, peaceful, 
sheltered waters. Wherever there is a high-peaked island, 
the clouds, attracted and caught by the peaks, are condensed 
into water; the water forms rills, and the converging rills 
in many cases become rivulets flowing into the sea. In an 
island of considerable size we may well suppose two or three 
of these mountain streams at different places will enter the 
deep. Wherever a stream enters there is no fringing reef 
