of Edinburgh , Session 1885-86. 
891 
situated towards the margin of the reef. On the contrary, I found 
them much larger and belonging to species of Porites ( P . tenuis, 
P. gaimardi, P. aspera , &c.) which must be of slow growth as 
compared with the species of the two genera, above mentioned, that 
are commonly found at the outer edge of reefs in these seas. Such 
facts are only to be explained on the hypothesis that the reef has 
gradually grown outwards as from a centre and quite independently 
of any movements of subsidence. 
( b ) In the cases of those partially elevated lines of barrier-reefs, 
which occur on the weather side of the principal island of the 
Shortland Islands, the present reef- flats are backed by a line of low 
coral limestone cliffs at the base of which is an old line of erosion. 
This mark might easily be taken as evidence of an upheaval, since 
the water coming over the reef-flat at high tide scarcely reaches it ; 
but, as shown in the accompanying diagram, it is only necessary in 
A 
B 
A. Here the waves are represented as breaking at the base of the cliff. 
B. A shore-reef has now been formed, against the outer edge of which the 
waves are breaking. 
imagination to remove the reef-flat in order to picture to oneself the 
rollers breaking at the base of the cliff and to perceive that this 
line of erosion has been formed at the present sea level. By the 
advancing growth of the outer margin of the reef, the cliffs have 
been cut off from the action of the waves. Such lines of erosion 
may be partially concealed by sand. They are commonly to be 
observed in cliff-girt coral islands, like Santa Anna, which have 
narrow shore-reefs extending seaward from the foot of the cliffs. 
(c) The characters of the wooded islets, that have been formed on 
