C. CROSSLAND—RECENT HISTORY OF THE CORAL REEFS. 21 
of water, perhaps one hundred yards wide, and apparently of uniform 
structure throughout. Its clear-cut edge meets the blue-black deep sea 
directly, without the intervention of a patch of lighter colour anywhere, giving 
the most forcible impression of the edge of a fathomless abyss. 
Scattered Reefs and the Shubuk Area. 
It will be convenient in considering the remaining reefs to divide the 
coast into two sections—the first from Ras Benas to Suakim, the second from 
Suakim to the boundary with the Italian territory, Ras Kasar. The former 
is characterized by the presence of well-developed barrier systems and deep 
water close to all reefs; the latter by innumerable small scattered reefs and 
sand and coral islands (the Suakim archipelago) in comparatively shallow 
water and a shelving bottom, and, in the south, a comparatively slight de- 
velopment of recent coral-growth. The contrast is thus a very strongly 
marked one and very significant also*. One of these scattered islands, Tella 
Tella Suraya, has already been described. 
The only other feature of interest in this neighbourhood is the Shubuk 
area, a hundred square miles of the most intricate passages among reefs, 
surrounded on the N.E. and H. sides by an unbroken reef similar to the 
ordinary fringing-reef of the shore, and on the 8. and W. by the mainland. 
Most of the reefs enclosed are merely sandbanks capped with a fathom or so 
of growing coral, but on the more exposed N. and W. sides the coral goes 
deeper and, where not growing coral, the bottom is largely composed of dead 
fragments of coral, shells, Kc. 
The land hereabouts is so low and so cut into by great shallow lagoons, 
that it is difficult to define a coast-line at all. It is easy to see how this very 
remarkable area has been formed, the shore-reef spreading seawards rapidly 
in the shallow water, where its débris would be at once available as a sub- 
stratum on which the growing zone of coral could advance seawards. Within 
the growing edge the accumulated mass of corals &e. would undergo decay 
and solution into the maze of reefs now found there; while others were being 
carved out of the low coast, and these, when their surfaces were covered with 
coral, living and dead, would become indistinguishable from those which had 
grown up in situ. 
The Barrier System. 
But little is to be added to the account given by Darwin in his * Coral 
Reefs, and for a description I refer to his work p. 102 & pp. 143-146 (the 
latter reference including a special section on the reefs between Suakim and 
Halaib), and to the map on p. 6 (supra). But the facts I have given 
above show that it is impossible to suppose a secondary subsidence of the 
* Note added Sept. 1907.—On Jand the contrast consists in the absence of sandstone 
coast-hills, the significance of which will be seen after reading the following pages and the 
postscript. 
