This description is based on observations made during several months’ work during 1938 and 1939, undertaken for the Government of the Somaliland Protectorate in connection with water-supplies in the barren Zeila Plain.

Geology Of The Zeila Plain, British Somaliland


Geological Magazine, Volume 79, Number 3, pages 197–201 on 1 June 1942

Article contents








The plain was formed from detritus from the mountains bordering it on the western and southern sides in Tertiary times. Near the coast these terrestrial deposits merge into littoral marine deposits. Stratigraphically the plain is divided into four concentric zones, truncated by the sea coast: (i) the Older Boulder Beds; (ii) the younger gravels; (iii) the Zeila plain “laterite”; and (iv) the Coastal Marine beds. These are described and the possibility of oil deposits discussed.

Geology Of The Zeila Plain, British Somaliland
Zeila old buildings – Somaliland © Eric Lafforgue


This description is based on observations made during several months’ work during 1938 and 1939, undertaken for the Government of the Somaliland Protectorate in connection with water-supplies in the barren Zeila Plain.

The plain is bounded on the west by basalt flows overlying igneous and metamorphic basement rocks, and terrestrial sandstones presumed to be Cretaceous Nubian, but possibly in part of Tertiary age. These western mountains of the Protectorate have been visited by Macfadyen (Macfadyen, 1932).

To the south, the plain is bounded by a series of step-faulted basement rocks (mostly acid), Jurassic, and Cretaceous “Nubian”, partly surveyed and described by Wyllie and Smellie (Gregory, et al, 1925).

To the north and north-west the plain slopes gently down to the shore of the Gulf of Aden.

The plain itself consists of the detritus from these western and southern mountain ranges, decreasing in coarseness towards the Gulf of Aden, and upwards in the geological column, and merging near the coast into littoral marine deposits.

Stratigraphically the plain has been divided into four concentric zones, truncated by the sea coast: (i) the Older Boulder Beds (Macfadyen, 1932 and 1933); (ii) the Younger Gravels (op. cit. sup.); (iii) the Zeila Plain ”Laterite”; and (iv) the Coastal Marine Beds.

(i) The Older Boulder Beds consist of conglomerates and soft buff calcareous false-bedded sandstones, made up of boulders of Basement rocks, Jurassic limestones, and “Nubian” sandstones. No volcanic boulders or pebbles have yet been observed in these conglomerates, even close to the basalt flows of the Western Mountains, and vulcanicity in the Zeila Plain is therefore at present believed to have originated after the Older Boulder Bed age. In the foothills, boulders of older rocks up to two or three meters in diameter occur in this series. At Geriado, on the Jireh-Geriso fault, these Older Boulder Beds are cut by a basalt dyke, feeding a flow which overlies and bakes them. This whole complex of Older Boulder Beds, basalt flow, and some overlying Younger Gravels at Geriado are up-tilted by an acid volcanic vent of later age, consisting of rhyolites, quartz porphyry, and volcanic ashes.

(ii) The Younger Gravels, overlying the Geriado basalt flow, consist of coarse conglomerates, gravels, sandy clays, and clayey sands, all more or less calcareous, and typically set in the reddish matrix of the calcareous sandy clay. In addition to the pebbles found in the Older Boulder Beds, there are pebbles of basalt, and in the greater part of the series pebbles of rhyolite, locally quartz porphyry, and of other acid volcanics. These acid volcanics are correlated with the Younger Gravels, a belt of which they penetrate in a series of vents, from the Djibouti border on the coast, southwards to Geriado. The spherulitic rhyolite flow of unproved thickness, at a depth of 92m. in a bore-hole at Zeila (Parkinson, 1932), is tentatively correlated with this intra-Younger Gravels age.

(iii) The “Laterite” is really only a zone of the Younger Gravels in which matrix preponderates over gravels and conglomerates. It is typically red and consists of red limestone residue, volcanic ash, red sandy clays, and clayey sands. The name “Laterite” has been chosen for convenience to distinguish this volcanic bearing deposit from seemingly similar red sandy clays which are so widespread throughout the Protectorate, but which normally consist largely of limestone residue without any volcanic material. Lenses of conglomerate and sandstone, usually with calcareous cement and impermeable, occur at least as far out into the plain as Ameruta and Sillil boreholes, but a rough boundary has been drawn between Younger Gravels and “Laterite”, where matrix preponderates, where pebbles are smaller than fist-size, and where there are no acid volcanic extrusions.

At Ameruta (altitude 152m. above sea level), a dry borehole of 128m shows an increasing predominance of calcareous conglomerates in the “Laterite”, and the basal 15m of this bore-hole are correlated with the Younger Gravels. At Sillil (alt. 74m) bore-holes have shown a similar sequence of ”Laterite” with some gravels, and with freshwater sands at 68m (i.e. 6m above sea level).

The only fossils found in the ”Laterite” are some beds of a thick-shelled helicoid snail near the top of the series.

This “Laterite” zone extends towards the coast from the belt of Younger Gravels and acid extrusions to about 15m contour, where there is a marked flattening of the already low gradient towards the sea.

(iv) Coastal Marine Beds are presumed to begin about this 15m contour, but they are overlain by salt silts and loess, with a typical vegetation of salt-bush called “harun”, which stops abruptly above this contour, indicating a close conformity between topography, stratigraphy, tectonics, and vegetation.

The actual sub-surface boundary between the “Laterite” and the Coastal Marine Beds is obscure, the nearest bore-hole at Zeila, having proved 92m of littoral deposits, shelly sands of Recent, Pleistocene (or possibly late Pliocene) age overlying the spherulitic rhyolite mentioned above (Parkinson, 1932). The proved marine deposits above this acid extrusion are therefore correlated with the “Laterite”.

(v) The superficial most recent deposits are red and white sand-dunes of the coast south-east of Zeila (Macfadyen, 1933), the coral and shell-rag reefs of Zeila and the coast towards Djibouti, modern river alluvium, and dark brown carbonaceous clays, the sand and loess dunes of from a few cm. to 2m or more in thickness drifted sporadically across the whole plain, and the most recent gravels.

These most recent gravels consist of dark grey, brown, cream, and pink limestones, probably all Jurassic, the brown being commoner south of Asha Ado, and the pinks south of Sillil. Igneous and metamorphic pebbles with a few of the harder “Nubian” sandstones are universally distributed, but the vesicular basalts from the western flows of the Djibouti border are very rare south of Geriado Tug (dry river bed), whilst the quartz porphyries, and pink and purple ashes, believed to be derived only from the Geriado vent described above, have only been found in the Geriado, Sillil, and Odowa Direh tugs.

In the terrestrial most recent deposits inland, broken ostrich shells, and a present-day fusiform terrestrial snail shell indicate the very recent age of these deposits.

Geology Of The Zeila Plain, British SomalilandTECTONICS

Whilst the main faulting trends of the Somaliland Protectorate are those of the Gulf of Aden (east-west), and the Red Sea (N.W.-S.E.), in this area the latter trend is not so common as a direction at right angles thereto, as shown on Wyllie and Smellie’s map of the southern mountains (Gregory, et al., 1925). This is believed to be due to the compression at the junction between the Red Sea and Gulf of Aden troughs, and is analogous to cross-folds so frequently developed bisecting the angle at the intersection of two major fold directions. The east-west Jireh-Geriso fault at the foot of the southern mountains marks the northern boundary of the Basement rock-Jurassic “Nubian” block-faulted complex. North of this line no Jurassic rocks have been seen, and to the south there are no volcanics, except that the Geriado vent seems to be on the actual fault.                In the southern mountains the Red Sea faults are well-developed, with some N.E.-S.W. cross-faults. In the western mountains (which have only been cursorily visited at a few points) the N.E.-S.W. cross-faults appear to be, in part at least, of post-basalt age. The general dip of the Jurassic and the “Nubian” sandstones south of the Jireh Geriso fault is south to south-west.

The tectonics of the Zeila Plain itself are largely hypothetical. There is a dip, towards the sea only slightly greater than the topographic gradient, and in the main, the topography reflects the dip. As a result of detailed survey of some areas on a large scale, and subsequent reconnaissance over wider areas, it is believed that the plain is underlain by block-faulted rocks similar to those of the southern mountain range. These faults have been thrown down towards the north at intervals up to the present day, the last recorded earthquake shocks at Zeila having been in 1930 (Taylor, 1931). The Recent and Quaternary (and probably Pliocene), are believed to reflect these faults faintly in a series of monoclinal tendencies, folded down towards the northwest and north. This hypothesis is borne out by (a) topography, (b) distribution of river gravels and drainage, and (c) the known shallow water deposits which indicate a gradually sinking littoral about the present coast-line, which has only recently been slightly elevated (Parkinson, 1932). The monoclinal trends seem to be for the most part from S.W. to N.E., but towards the coast, they tend to curve round more to the Gulf of Aden direction towards the west, where they have deflected or ponded back most of the old north-east flowing rivers.


The southern and western mountains, now bounding the Zeila plain, were terrestrial areas in Tertiary times, and it is believed that the drainage from these areas flowed east, north-east, and north into a faulted synclinal into which the Tertiary seas may also have penetrated from the east. Some fluviatile and lagoonal beds might be expected locally under the “Laterite”, of the same facies (though not necessarily of the same horizon), as the Daban beds farther east (Macfadyen, 1933). The importance of these hypothetical Tertiary and known Quaternary littoral deposits, coupled with the absence of known vulcanicity in the east of the plain, the known impermeable cover of at least 100m of “Laterite”, and the tendency to monoclinal folding even in the upper parts of this, lies in the possibility of oil being found under this barren plain towards the coast. In spite of rumors of oil seepages, only sulphur springs have been seen, and those near the Jireh-Geriso fault at Jireh. The plain and surrounding hills, however, have not yet been thoroughly surveyed, and the conditions are similar to those of some of the intra-continental oil-fields in the East Indies, where in some cases no oil seepages were discovered before drilling proved rich deposits. The fact that the East Indian oil-fields are clothed in rich forests and the Zeila plain is an unattractive desert is misleading but quite immaterial.


Gregory, J. W., et al. , 1925. The Collection of Fossils and Rocks from Somaliland made by Messrs. B. K. N. Wyllie, B.Sc, F.G.S., and W. R. Smellie, D.Sc., F.R.S. Ed. Mon. Geol, Dept. Hunterian Museum, Glasgow University. Google Scholar

Macfadyen, W. A., 1932. The Late Geological History of British Somaliland. Nature cxxx 433 CrossRef Google Scholar

Macfadyen, W. A. 1933. The Geology of British Somaliland, pt. 1. (Published by the Government of the Somaliland Protectorate.)Google Scholar

Parkinson, J., 1932 a. Climatic changes in British Somaliland. Nature, cxxix, 651. CrossRef Google Scholar

Parkinson, J. 1932 b. A Preliminary Note on the Buramo Schists, British Somaliland. Geol Mag., lxix, 517–520. CrossRef Google Scholar

Taylor, R. S., 1931. Earthquake shocks in Somaliland. Nature, cxxvii, 34. Google Scholar

This site uses Akismet to reduce spam. Learn how your comment data is processed.