884 
Proceedings of the Royal Society 
the breakers at the margin of the reef, collect at the foot of this 
declivity in depths generally of from 15 to 20 fathoms, but some- 
times in rather greater depths as at Port Mary. However, since 
the growth of coral is suppressed by this accumulation of sand and 
gravel that collects at the foot of the declivity, which I have 
termed the growing edge of the reef, it follows, as I have already 
pointed out in the instances of the Onua, Choiseul Bay, and Port 
Mary reefs, that the apparent lower limit of the zone of corals 
will be determined in the case of each reef by the depth at which 
the base of the declivity lies. The “ raison d’etre ” of the declivity 
is to be found in the circumstances that the corals flourish only on 
the outer slopes of reefs, in depths beyond the first line of breakers, 
i.e ., below a depth of 4 or 5 fathoms. Here they are exposed to 
the strength of the tidal currents, and, as clearly shown by 
Professor Semper, they would tend to form precipitous or wall-like 
declivities growing outwards, as Mr Murray holds, each on its own 
talus. 
With regard to the disposition of the sand and gravel on the 
slopes below the growing edge of the reef, my soundings showed 
that where the descent is at all rapid (i.e., more than 10° or 12°), 
as is generally the case, this detritus extends down far beyond the 
depths in which corals thrive, but that where the slope is gradual 
(i.e., less than 5°), as in the instance of the barrier-reef of Choiseul 
Bay, the lower limit of the sand and gravel lies within the coral 
zone which is, in point of fact, divided into two sub-zones by a belt 
of detritus. 
The results of my examination of the seaward slopes of reefs 
have supplied me with an explanation of the formation of barrier- 
reefs, which I will briefly review in the light of numerous observa- 
tions I have made in this group, both on existing and elevated 
coral reefs.* 
precipitous slope in depths of 15 to 20 fathoms, the rock overhung to such an 
extent that he was able to get underneath the projecting portion. This was 
evidently due to the outward growth of the corals on the face of the declivity. 
* I first described this view of the formation of barrier-reefs in a short paper, 
entitled “ Suggestions as to the Formation of Barrier Reefs, ” &c. , which was read 
before the Linnean Society, New South Wales, in October 1884 ( Proc ., vol. 
ix. part 4). At that time I was in ignorance of the fact that substantially the 
same explanation had been proposed many years before by Professor Joseph 
Le Conte in the instance of the Florida barrier-reefs. In 1856 Professor Le Conte 
