The letter B is at a level about 25 feet above the foreground A. Coral colonies, in position of growth, bedded in a mass of loose coral débris, shells, &c. C. Hardened coral-mud. The weathered surface forms rounded masses in low relief. D. Gypsum strata, here steeply tilted, and upturned at their ends in the piece shown in the foreground. They are much folded in other parts of the cliff. E. Green and red shaly rock underlying and sometimes interstratified with the gypsum. It is here broken down into sand. This rock contains sheets of glass-like recrystallized gypsum.] Meanwhile organic remains were accumulating on ridge No. 2, and as elevation brought this within fifty fathoms or so of the surface, reef corals took possession and covered the summits one after another as elevation proceeded, so that [Illustration: Diagram 12] when ridge No. 1 emerged from the sea altogether, and its bases were surrounded by the gravel from the hills, No. 2 was a second barrier reef out at sea. The same process has been repeated, so that coral growth and levelling have made ridge No. 3 into the present barrier system and the maritime plain has reached the one-time barrier No. 2 and so made this the present coast-line. During the last of these elevations a good deal of breaking and cracking of ridges Nos. 2 and 3 took place. For instance, Rawaya was originally connected with the mainland, the proof being the presence upon it of scattered pieces of Archean rock which could not possibly have reached it unless a continuous surface stretched from it to the old hills. Dongonab Bay, and with it probably other parts of the channel within the barrier system, have evidently been formed or enlarged since the maximum extension of the maritime plain. The harbours of the coast, which are so interesting in themselves as to deserve separate consideration, were formed also at this time. Natural harbours, almost completely surrounded by land or reef, waveless in all weathers, more perfect than almost any made by man, abound throughout the length of this coast. In one part, just north of Rawaya, are ten of these strange inlets in a space of only 40 miles (see map, page 126). That of Suakin has been already described; on entering for the first time it is hard to believe that this long parallel-sided, deep channel, bounded by reefs covered only by a foot or two of water, and then by land only the same amount above the sea, is not an artificial canal. It leads nearly straight inland for two miles, but not _quite_ straight, indeed there is a bend that large steamers frequently fail to clear, and which led to the abandonment of Suakin as the Port of the Sudan.