Chapter IV: Topography

59.	A topographical map is the simplest and most efficient basic reference for other survey work. It should be an accurate geodetic map (i.e. points accurately marked both on the ground and on a map of the Earth's surface as a whole).
60.	Somaliland has not yet been included in the geodetic network of Africa, but some fairly accurate points have been fixed by astronomical means, and may be obtained from the Directorate of Colonial Surveys, Teddington. Unfortunately many of these points are not permanently marked on the ground, or refer to peaks and hill-tops from which temporary stone beacons have disappeared, and these points of course cannot be accepted as an accurate geodetic network.
61.	The most recent astronomical fix is that of Berbera Lighthouse, made by a surveyor of the Shell Company, and given in the Gazetteer (Table 3) (10^o 24' 49''. 2 N., 44^o 58' 42". 7 E.).
62.	The existing astronomically fixed points are, however, sufficient for the making of maps on the scale of 1 : 2500,000 (approximately quarter-inch to the mile), or maps on smaller scales, of limited areas in the coastal lowlands, and in sight of the Main Watershed Ranges.
63.	The existing published maps are of very variable value in different areas. Some of the most satisfactory are those made by Colonel Swayne at the end of the nineteenth century, by duration of camel-march and compass between points fixed by him astronomically during his hunting trips (Swayne, 1895). Some of the most misleading maps are those of the Railway Survey, 1906, and the Boundary Surveys of the 1930's, in which narrow strips of country have been accurately surveyed and others filled in, often extremely inaccurately. With the aerial photographs and the information given in this Report, an improved series of maps could now be made.

64.	It has therefore been necessary to improvise a topographical map from the existing published maps and by further amateur survey to cover as much of the area as possible. Topography -- popularly confused with the wider term "Survey" -- has only been sketched in where better maps are lacking, for the purpose of forming a basis for the General Survey of the Protectorate and Grazing Areas.
65.	In areas where there are well-mapped, visible landmarks, plane table surveys have been based on such points, e.g. Stock's Elayu-Heis plane table survey based on the N.E. boundary pillars of the international boundary with Somalia Italiana; Hunt's Golis-Guban plane table survey based on Somaliland Oil Exploration Co.'s fixed points.
66.	In other areas, e.g. Hargeisa Valley, Zeila Plain, and Onkhor area (Hunt, 1939 and 1945), base lines have been measured and extended by plane tabling: in the first two cases with telescopic alidade measuring heights, and in the last by ordinary alidade, plane tabling, and heights by aneroid with a few angle checks. Such maps are tied in to existing maps as well as may be; where possible to distant astronomically fixed points, but elsewhere merely by "fitting" into existing maps.
67.	In the Plateau area (e.g. the Haud and Sawl Haud mostly by Macfadyen, and much of the area south of the Main Watershed by Hund and Viney), the maps have been amplified for the most part by motor-vehicle mileometer, compass, watch, and aneroid (though in some areas of the central Haud Macfadyen also used aerial photographs). In some cases the mileometer was replaced by the duration of camel-marches on compass bearings, and by both these means a fair degree of accuracy can be obtained by paintstaking care.
68.	Diurnal hourly aneroid graphs were prepared before and after each such survey, and heights corrected from these graphs. If the results were not fairly constant before and after a short survey the heights were checked again or discarded altogether. In the course of a year the heights shown by an altimeter (aneroid) at one place, have been recorded as varying as much as 460 feet, and the diurnal variation is usually 150 to 200 feet. It is believed that the heights recorded by means of aneroids during the General Survey are mostly accurate within about 50 feet, and almost certainly within 200 feet.
69.	Small areas such as townships and air landing-grounds have been measured by tape and compass, or pace and compass, and sometimes extended by compass bearings. A series of such maps was included in the 1947 Annual Report of the General Survey, Illustrations XII to XXI, Townships 1 : 2,000, Stations 1 : 25,000, Environs 1 : 50,000, and Landing-Grounds 1 : 10,000. The series was not completed for the whole Protectorate, but copies of the completed maps are available for local use from the Stationery Officer, Hargeisa.
70.	As far as records show, areas as large as English counties remain practically unknown. The position of these areas is fairly obvious from the lacunae on the rainfall and tribal maps in this Report. In areas not personally known to the survey officers or to their Somali staff, unless shown on existing maps, no general records such as rainfall or tribal movements could be plotted. When reports about unidentified places were received, Somalis were interrogated, and in many cases the approximate positions of the places or areas named could be ascertained and plotted. At the end of the Survey, however, quite large areas of country remain almost blank on the topographical maps, and therefore very sketchy or blank on the rainfall and tribal maps.
71,72.	Illustrations 5 and 6.)

74.	Searching for places on the existing maps resulted early during the Survey in the necessity for a gazetteer. Fortunately the R.G.S. II system was used from the outset. The "First List of Names in the Somaliland Protectorate" (Permanent Committee for Geographical Names: P.C.G.N. 1928) was not available, and as the Gazetteer (Table 3, para. 78) has been found much more useful in practice, it is suggested that it should supersede the 1928 publication.
75.	In the Gazetteer positions are given as far as possible to the nearest minute of H. latitude and the nearest minute of E. longitude. The retention of the R.G.S. II system of spelling names is strongly recommended, since it has become a universal system of spelling for British soldiers, geographers, geologists and other explorers and scientists who use it throughout the world on their travels, and cannot always learn local languages before they visit a country.
76.	A place name sketch map on the scale of 1 : 3,000,000 was included in the 1945 Annual Report. This has been amended and reduced to a scale of 1 : 10,000,000 (illus. 8 below) for the purpose of this Report, as an illustration of the Gazetteer, and to give a general idea of the main areas of the country referred to in the following chapters.
77.	(Illustration 8.)

79.	In view of the wide areas covered by the Survey, both in surveying and in patrolling the observer posts, it was necessary to know the mileages of roads for the calculation of petrol supplies and programmes of travel. These mileages have been revised (from the 1945 Annual Report, General Survey) and are given in detail to the nearest mile in Table 4.
80.	Illustration 9 (in pocket) shows the roads referred to in Table 4 with the names of corners and road junctions. The mileage lists of Table 4 have been divided into patrols based on Burao, and the patrols are marked on this map (illus. 9) as a general guide to the use of the Table.
81.	(Table 4., Table of Road Mileages - Somaliland Protectorate and Grazing Areas) [Editorial Note: This table appears sequentially in the original printed text, on 12 pages listing the individual patrols. We have not yet included it as a separate webpage in this later online edition, because this information does not appear to be currently relevant because roads and routes have changed.]
82.	A diagram on the scale 1 : 3,000,000 was given in the Annual Report, 1945 (illus. xiv), and has been enlarged by the Public Works Department for their use. It has now been corrected up to the end of 1950, and reduced to 1 : 10,000,000 for this Report (illus. 10) to show only the main route mileages.
83.	(Illustration 10.)

84.

Great care has been taken in the compilation of mileages, which were measured by mileometers checked over measured distances, and with constant tyre pressure and diameter. The short cuts made by drivers, and variations in road routes made from time to time during repair of improvement work, ensure, however, that few mileage records will remain constant in open country. The correction of the enlarged diagram printed by the Public Works Department is recommended, according to Table 4 above.

85.

In the course of measuring roads a good deal of reconnaissance mapping, both topographical and of a general nature, was carried out.

86.

Camel track in the Plateau area and in the Haud were there are few hills, are often better than the main roads over much of their length, if they are not too much used by motor transport.

87.

In spite of the popular statement that modern transport can go anywhere (in the Somaliland Protectorate), there are in fact a number of very definite barriers to motor vehicles.

These may be divided into:

(i)	Barriers which could be made passable by ordinary labour and tools and supervision.
(ii)	Barriers which must remain impassable unless expensive engineering works are carried out.

Some of the latter include cliffs and narrow gorges impassable to camels and even to donkeys. Barriers to a determined man are very few and can usually be circumvented by a detour of a mile or so.

89.

Illustration 12 shows the area already voered by the R.A.F. with vertical air photos.

This, with lacunae, covers approximately the areas:

(i)	Coast of the Gulf of Aden to 10^o 00' N. by 45^o 30' E. to 47^o 30' E.
(ii)	10^o 00' N. to 8^o 00' N. by 44^o 30' E. to 46^o 00' E.
(iii)	Coast line of Gulf of Aden to 10^o 00' N. by 45^o 00' E. to 45^o 30' E.

The first two, known as R.A.F. Air survey SOM 13, and the third as R.A.F. Air Survey SOM 13B, were flown at an average height of 20,000 feet, and photographed with focal length 6 in. There are unfortunate gaps in the mountainous areas of the Main Watershed. Three sets of these photographs have been transferred at the completion of the General Survey to the Director of Agriculture and Veterinary Services (D.A.V.S.).

90.

The whole country was also covered by U.S.A.S. Trimetrogon air photographs, the average vertical scale of which is 1 : 40,000. A nearly complete set of these photographs of the Protectorate was also handed over the the D.A.V.S. in 1951.

91.

The Trimetrogon photographs were obtained by the General Survey in November 1946, and the R.A.F. photographs in 1947 and 1948, and were therefore too late to be of use in all areas covered by the Survey. Details of drainage of the central Haud were, however, obtained for some areas from these photographs by Dr. Macfadyen, who by their study also first discovered the existence of "Vegetation Arcs" (Macfadyen, 1950).

92.

In 1948 a ground survey party of the R.E.'s fixed some control points for use with the R.A.F. airphotos in map-making. The maps have not yet been made as far as is known, but is it hoped that they will be completed by the Directorate of Colonial Surveys soon, together with maps of the areas of the Protectorate not yet photographed.

93.

Much can be learnt from the interpretation of air photographs, and maps can be made more quickly by means of such photography than by any other method (if there is sufficient trained staff available). It is essential, however, that the area should also be covered on the ground especially for the purpose of fixing control points, identifying peaks and other landmarks from the ground, finding water, passable routes, and, of course, all available information for the plotting of which the topographical map is produced, including place names.

Notes for Writers: Bookmarks are included for each section, i.e. "A." - "H.", and will later be included for each numbered paragraph, i.e. "443." to "555.", each table and each illustration.