120 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY, 
intelligent béche de mer fishermen, they do not think that corals flourish 
on the outer edge at a greater depth than on the faces of inner reef flats, 
six to seven fathoms. І have also been assured that the small outer 
barrier of reef flats which we examined south of the Lark Passage was 
typical of all the reef flats which form the line of the so called outer 
linear reefs (Jukes). The surface inside of the breakers of all the outer 
Barrier Reef flats are covered with dead corals, as those of the interior 
reef flats are. We may therefore safely conclude that the dead corals 
which cover the outer reef flats owe their origin to the same cause, — 
viz, to the disintegration of the former elevated coral reef which once 
was the outer barrier reef. 
That the outer reef flats all are arcs facing the breakers, and form so 
uniformly narrow shallow strips between deep inner channels and the 
outer sea-face, we can only explain by the peculiar action of the sea, 
which invariably on exposed shores erodes islets into a more or less cir- 
cular outline, and patches into convex faces turned about at right angles 
to the general trend of the breakers (Plates XXXII.-XXXIV.). 
The inner channel of the outside Barrier Reef patches ranges from віх- 
teen to twenty-eight fathoms, the average depth being about twenty 
fathoms ; from this it shoals somewhat steeply to ten fathoms, and then 
very gradually to the inner edge of the outer reef flats. Between six 
and seven fathoms coral masses appear, often forming large detached 
patches rising quite abruptly on all sides to within one or two feet of the 
surface at low-water mark. From there in on the slope, in four to five 
fathoms, the patches become more numerous, often forming long 
connected stretches of coral heads. The belt between five and three 
or four fathoms is the one in which the corals seem to thrive best. 
The masses become more or less disconnected’ again in from one 
quently determined by local causes that it seems impossible to assign their great 
development merely to their position on either the lee or seaward face at any 
locality. . Taking the Samoa Islands, the lee face of the islands is the very face on 
which the coral reefs of that group are developed. At the Sandwich Islands, where 
we have the northeast trades, while there is a lee face reef, we have an equally 
well developed sea-face reef. On the Windward Islands, the sea-face reefs are 
much more developed than those on the lee face of the islands. On Alacran 
Reef the lee reef is nearly overwhelmed by the sand driven from the islands over 
that face, while the sea face is a most flourishing mass of corals. While the lee 
faces of the Bahamas support a most luxuriant growth of corals, it does not com- 
pare with that growing on the weather side. On Hogsty Reef the corals facing the 
great swell of the northeast trades are far finer than those on the lee side. The 
same is the case for the Bermudas. 
