Please see the [Transcriber’s Notes] at the end of this text.

THE NILE IN 1904

BY
Sir WILLIAM WILLCOCKS, K.C.M.G., F.R.G.S.

LONDON
E. & F. N. SPON Limited, 57 Haymarket.
NEW YORK
SPON & CHAMBERLAIN, 123 Liberty Street.

Price 9s/- net.

Printed at the National Printing Department of Egypt, Cairo

1904.

Dedicated to my old Chief and Master in Irrigation, Sir Colin Scott-Moncrieff, K.C.S.I., K.C.M.G., under whom I had the privilege of working for 20 years in India and Egypt.

PREFACE

The publication of Sir William Garstin’s monumental work on the “Basin of the Upper Nile” is an event of such importance in the history of the Nile that the occasion should not be lost of bringing Lombardini’s work on the Nile to date. The information utilised by me in this book as far as the Upper Nile is concerned is obtained from Sir William’s Report; for the Blue Nile and Atbara I am indebted to M. Dupuis’ interesting appendix at the end of Sir William’s Report; and for the river north of Khartum to my own studies and surveys. As Sir William employed Capt. H.G. Lyons, R.E. to collaborate with him, the references to the works of previous writers and geographical details may be accepted without any misgivings. To M. Chélu Bey, Director of the Government Press, I am indebted for his ever ready aid; to Mr. Hansard of the Survey Department for the plates accompanying this work; and to Mr H.G.F. Beadnell, F.G.S, F.R.G.S, for having kindly written the description of the Egyptian oases and the geology of Egypt which form the fifth chapter of this book.

W. Willcocks.

Cairo, 12-10, 1904.

THE NILE IN 1904.

SUBJECT MATTER.

LIST OF PLATES.

THE NILE IN 1904.

CHAPTER I.
The Nile.

1. Introduction.

—In the introduction to his brilliant essay on the Hydrology of the Nile[1], an essay, which, though written in 1865, foreshadowed much of what we know to day, Lombardini remarked, with much truth that, no river in the world lends itself to hydrological studies on so majestic a scale as the Nile. The most interesting river of the ancient world, it is still the most interesting river of our time; and, in spite of all that ancient and modern discoveries have unfolded, its discharges are to-day more difficult to unravel and weave together than those of any other stream in either hemisphere. These discharges are still a mystery, and it will need years and years of patient observation and study, at the hands of the Sudan Irrigation Department, to enable us to state with exactitude why its floods rise and fall with such regular and stately precision, why they are never sudden and abrupt, and why its summer supplies can never be completely cut off even in their traverse of over 3000 kilometres through the burning and parched Sahara. Though the mystery of the Nile is far from being solved to-day, still an enormous step in advance has been made by the publication of Sir William Garstin’s Report on the Basin of the Upper Nile[2]. This Report not only contains the results of three years’ observations of the Egyptian Survey Department in the Sudan, of Sir William Garstin’s own observations and studies, but also a mass of information of the Nile and its tributaries collected by Capt. H. G. Lyons. R. E., through four years of uninterrupted study. Those who know the intelligence and method with which Capt. Lyons works, will rate this information at its proper value.

[1] Saggio idrolico sul Nilo, by Elia Lombardini, Milan 1865.

[2] Report on the Basin of the Upper Nile by Sir William Garstin. Blue Book Egypt (2) 1904.

Lombardini gathered together all the information available at the time that Sir Samuel Baker announced the existence of the Albert Nyanza shortly after Speke and Grant had proclaimed to the world that the Victoria Nyanza was the true source of the Nile. From the information then available he deduced the laws and operations of the great river. About twenty years later, just before the rebellion in the Sudan closed the Nile to the civilized world, a German savant, Joseph Chavanne[3], in his book on the rivers of Africa, collected and tabulated on clear and methodical lines much of the information available in 1883. Though many of his facts are erroneous, his method is clear and his ideas just. Sir William Garstin, in his Report, has developed the information at his disposal on such practical lines as are needed to study the question of insuring an abundant supply of water to the Nile in Egypt during the times of low supply.

[3] “Afrikas Ströme und Flüsse” by Joseph Chavanne. Wien 1883.

Having myself studied the Nile for fifteen years in order to solve the problems of water storage and flood control on the Nile, and having devoted the whole of my life to this very science of Hydraulics, I have been encouraged to attempt the continuation of Lombardini’s work; and, to the utmost of my ability, to bring it to the level of the knowledge of our day.

2. Nomenclature.

—The nomenclature of the tributaries of the Nile is difficult to follow. In this book I shall call the river the Victoria Nile from Lake Victoria to Lake Albert; the Albert Nile from Lake Albert to the Sobat mouth (this reach is known generally as the Bahr el Gebel); the White Nile from the Sobat mouth to Khartoum; and the Nile from Khartoum to the sea. The Blue Nile stretches from Lake Tsana in Abyssinia to Khartoum.

3. Description of the course of the Nile.

—[Chapters II] and [III] contain detailed descriptions of the Nile and its main tributaries, and this paragraph is a short epitome of what is written there about the course of the Nile. The Nile drains nearly the whole of north-eastern Africa, an area comprising 3 million square kilometres. Its main tributary, the White Nile, has its furthest sources in south latitude 4°, near Lake Tanganyika. Known as the Kagera, it is one of the feeders of Lake Victoria, and has a course of 600 kilometres before it reaches the lake. Lake Victoria, covering 60,000 square kilometres, is the first reservoir of the Nile. The Victoria Nile leaves Lake Victoria by the Ripon Falls and after a course of 400 kilometres enters Lake Albert at its northern corner. At its southern end Lake Albert is fed by the Semliki river which has its sources in Lake Edward. Its own area is 4,500 square kilometres. The Albert Nile leaves Lake Albert at its northern end and has a course of 1280 kilometres to the mouth of the Sobat river. Of this length, the first 200 kilometres up to Dufile have scarcely any slope, the next 150 kilometres are down a series of severe cataracts. From the foot of these cataracts to its tail the Albert Nile has a gentle slope and traverses the Sudd region where the bed of the stream is often barred by blocks of living vegetation. In this latter region the stream divides into two, of which the right hand one is known as the Bahr Zeraf. After a course of 270 kilometres the Bahr Zeraf joins the Albert Nile again. In the interval the Albert Nile receives as a left-hand feeder the Bahr Gazelle. The Sobat river has its sources in Gallaland and joins the Albert Nile at the termination of the Sudd region. From the junction of the Albert Nile and the Sobat, the river is known as the White Nile, which, after a course of 840 kilometres, with an exceedingly gentle slope, joins the Blue Nile at Khartoum.

PLATE I.

Lith. Sur. Dep. Cairo.

[Larger map] (240 kB)

THE NILE

The Blue Nile is the true parent of the land of Egypt. The deposits of its muddy waters have made Egypt. The Atbara has added its quota, but the Blue Nile is incomparably the chief contributor; fed by the timely and plentiful rains of southern and south-eastern Abyssinia, it contributes 65 per cent of the waters which pass Assuân. The furthest sources are those of the Abai, which, after a course of 110 kilometres falls into Lake Tsana. This lake has an area of 3,000 square kilometres and lies about 1,760 metres above sea level. The Blue Nile leaves it at its south-eastern corner and hurries down to the Sudan, fed by numerous Abyssinian rivers. At Rosaires, after a course of 750 kilometres, it has fallen 1,260 metres; and below the Rosaires cataract enters the plain country south of Khartoum. For its remaining 615 kilometres on to Khartoum, where it meets the White Nile, it is navigable for the greater part of the year. North of Sennaar it is fed by the Dinder and Rahad rivers.

Between Khartoum and El Damer, on a length of 320 kilometres, the Nile has its even passage broken by the 6th cataract at Shabluka. At El Damer the Nile receives the Atbara as a right hand tributary.

The Atbara is a very muddy torrent fed by the rains of north-eastern Abyssinia. It runs for 4 months per annum and is dry for 8 months. Rising within a few kilometres of Lake Tsana, it falls 1500 metres in its first 300 kilometres, and is then joined by the Salaama, and, 100 kilometres lower down, by the Settit river. After the junction with the Settit, the Atbara flows for 480 kilometres and joins the Nile at El Damer, contributing a fair quantity of water and a very considerable quantity of Nile mud to the river.

From the Atbara junction to the sea, the Nile has a course of 2,700 kilometres. In its first length of 1480 kilometres to Assuân it traverses the 5th and 4th cataracts between Berber and Dongola, the 3rd and 2nd cataracts between Dongola and Wady Halfa, and the 1st cataract at Assuân. All these cataracts are navigable in flood, but not so in summer. From Assuân to the Barrage at the head of the Delta north of Cairo, the Nile has a length of 970 kilometres and traverses Egypt without a cataract or interruption of any kind. At the Barrage, the Nile divides into the Rosetta and Damietta branches, and after a further course of about 240 kilometres in either branch, flows into the Mediterranean sea. Its greatest length from the sources of the Kagera river to the sea is 6350 kilometres, constituting it one of the longest rivers in the world.

4. The Slopes and velocities of the Nile in its different reaches.

—[Table 2] of [Appendix B] and [Plate II] comprise all the information available under this head which I have been able to collect. For the slopes I have adopted the following data:

From Khartoum to Wady Halfa I have adopted the generally accepted levels of the original Soudan railway survey. From Wady Halfa to the sea I have levelled myself. Upstream and downstream from the adopted levels I have carried the levels by the aid of slopes calculated from velocity and hydraulic mean depth data. It seems to me absurd to adopt a level for Lake Choga 50 metres above that for Fowera, and then to add, that in the 140 kilometres between the two places the Victoria Nile has a gentle slope, wide bed and gentle velocity. By a strange fatality, this very error has crept into the figures under Lake Choga on [Plate II]. The error is noted in the corrigenda attached to the Plate. The Section is drawn correctly but these wrong figures have been interpolated by an oversight.

The Victoria Nile Falls 450 metres in 400 kilometres, but has four reaches; the first ¹⁄₁₂₀₀, the second ¹⁄₂₀₀₀₀, the third ¹⁄₁₈₀ past the Murchison Falls, and the fourth ¹⁄₁₀₀₀₀.

The Albert Nile falls 277 metres in 1290 kilometres. The first reach past Wadelai has a slope of ¹⁄₂₅₀₀₀, the second over the Fola and following cataracts has a slope of ¹⁄₇₀₀, the third ¹⁄₁₂₀₀₀, the fourth ¹⁄₂₀₀₀₀, the fifth ¹⁄₂₅₀₀₀, and the last below lake No of ¹⁄₇₅₀₀₀ in flood.

PLATE II.

Lith. Sur. Dep. Cairo.

[Larger illustration] (180 kB)

ERRATUM.
Erase the two figures under Lake Choga in the “Height” column.

LONGITUDINAL SECTION of the NILE,
THE BLUE NILE, THE WHITE NILE, THE ALBERT NILE, THE VICTORIA NILE & THE SEMLIKI RIVER

The White Nile falls 14 metres in 840 kilometres and has two slopes in flood; ¹⁄₅₀₀₀₀ in its upper reach, and then ¹⁄₁₀₀₀₀₀.

The Blue Nile falls 1370 metres in 1370 kilometres, which may roughly be divided into three reaches. The first from Lake Tsana to Rosaires on a length of 750 kilometres ¹⁄₆₀₀, the second Rosaires to Sennaar ¹⁄₄₅₀₀, and the third ¹⁄₇₀₀₀. These are very approximate indeed.

The Atbara falls 1640 metres in 880 kilometres. In the first 300 kilometres the slope is ¹⁄₂₀₀; in the next 300 kilometres the slope is ¹⁄₂₅₀₀, and in the last reach of 280 kilometres it is ¹⁄₆₀₀₀. These are approximate.

The Main Nile from Khartoum to Assuan falls 295 metres in 1810 kilometres; the so-called six cataracts occupy 565 kilometres with a slope of ¹⁄₃₀₀₀; and the ordinary channel occupies 1245 kilometres and has a slope of ¹⁄₁₂₀₀₀. From Assuân to the Barrage, on a length of 970 kilometres, the Nile falls 76 metres with a mean slope of ¹⁄₁₃₀₀₀. The Rosetta and Damietta branches are each about 240 kilometres long and have a slope in flood of ¹⁄₁₃₀₀₀, and of ¹⁄₁₂₅₀₀ in extraordinarily high floods.

From the sources of the Kagera river to the sea, on a length of 6350 kilometres, the Nile falls 2000 metres, or has a slope of ¹⁄₃₂₀₀. From Lake Victoria to the sea the length is 5535 kilometres and the fall 1129 metres, or the slope is ¹⁄₅₀₀₀.

[Table III] of [Appendix C]. gives the velocities of the river in flood and low supply, in metres per second and kilometres per day, and also the time occupied in traversing the different reaches. There are two breaks. The first is at Lake Choga and the second is at Lake Albert. As the Victoria Nile traverses the eastern arm of the many-armed and peculiar Lake Choga with a perceptible current, and as, moreover, the lake is very shallow, we may give some figure to the velocity and make it half that of the Bahr Gazelle which is ·20 metres per second and is considered perceptible. With a velocity of ·10 metres per second or 8 kilometres per day, the 80 kilometres of the lake would be traversed in 10 days. The time of traverse from Lake Victoria to Lake Albert would be 15 days. With Lake Albert it is very different. A reference to [Plate V] will show that it takes the Victoria Nile 5 months to fill up Lake Albert before the Albert Nile can carry off the waters of the Victoria Nile, gauge for gauge. Under these conditions it will be wise to stop at Lake Albert and begin a new calculation from this lake.

The Albert Nile takes 22 days in flood and 25 days in low supply to traverse the distance from Lake Albert to the Sobat mouth. The White Nile takes 21 days in flood and 28 days in low supply to reach Khartoum. Consequently from Lake Albert to Khartoum we have 43 days in flood and 53 days in low supply.

From Khartoum to Assuân the Nile takes 11 days in flood and 22 days in low supply, and consequently from Lake Albert to Assuân we have 54 days in flood and 75 days in low supply.

From Assuân to Cairo we have 6 days in flood and 12 days in low supply.

[Table III] is very interesting and well worth study. Through the Sudd region we have a velocity of ·6 metres per second, but only of ·35 metres per second in the White Nile. In Egypt the Nile in flood has a velocity of 1·75 metres per second, and in low supply of ·85 metres per second.

5. Catchment basins of the Nile and its tributaries.

—[Table I], of [Appendix A] gives the areas of the catchment basins of the Nile. The total area according to the table is 3,007,000 square kilometres. The limits of the basin are depicted on [Plate I], and, with rare exceptions, they are now fairly well known everywhere. North of the 20th parallel of latitude the watershed on the west of the Nile is not far removed from the edge of the plateau skirting the Nile valley. The plateau falls away to the west, and occasional ravines find their way to the Nile down the reverse slope. On the east of the Nile the crest of the hills skirting the Red sea is the watershed. South of the 20th parallel of latitude the eastern watershed follows the crest of the hills on the west of the Red sea as far as Suakin. South of Suakin the watershed leaves the Red sea, to allow the Khor Barraka to flow into this sea. From the south east of Kassala, round by Addis Ababa, the watershed follows the crest of the high hills forming the eastern backbone of Abyssinia, and dividing the waters of the Nile from those flowing into the Indian Ocean. South west of Abyssinia the watershed travels in a south-westerly direction to the east of Gondokoro, and divides the Sobat from the rivers draining into Lake Rudolf. The watershed then moves due south to the western escarpment east of Lake Victoria. Mounts Kenia and Kilmanjaro are not within the basin of the Nile. Sweeping in a rough curve round Lake Victoria and nearly touching Lake Tangangyika in 4° south latitude, the watershed keeps close to the western shores of Lakes Edward and Albert to nearly opposite Wadelai. All the slopes of the Ruenzori mountains drain into the Nile.

From near Wadelai the watershed moves in a north-westerly direction along the hills dividing the waters of the tributaries of the Gazelle river from the Welle. Due west of the Sudd region the watershed has reached its most westerly position and from there turns northwards along the Marrah hills in Darfur, dividing the scanty waters of the Bahr-el-Arab and its tributaries from the rivers draining into Lake Chad. From the Marrah hills the watershed travels in a north-easterly direction to a point close to the Nile on the 20th parallel of latitude near Hannek.

Of the lands enclosed within this watershed, all that are drained directly into the Main Nile are desert. There are occasional showers, and some of the valleys and ravines carry water for a few hours every year, others every second, third or fourth year, but they contribute practically nothing to the volume of the Nile. The rains generally come in the winter when the Nile is falling every day, and the steady fall of the Nile is never arrested by the waters of any or all of these watercourses. The country west of the White Nile past Kordofan and Darfur to the Marrah hills is steppe land producing scanty grasses and forests of low accacias in the south, and rising to a general height of about 600 metres at the Marrah hills. The lands drained by the Gazelle river and the Albert Nile north of Gondokoro are flat plains or swamps in the north and east, and wooded and broken ground in the west and south-west, where the tributaries of the Gazelle river rise in the Blue mountains at a general height of 1500 metres. The upper waters of the Sobat and its tributaries drain the well wooded and cultivated mountain masses of Gallaland and then traverse the marshes and flat lands which lie east of the Sudd region. The Blue Nile and its upper tributaries drain the choicest portions of the high Abyssinian mountain plateau lying over 2000 metres above sea level, and rising in places to 2500 metres and upwards. The lower courses of the Blue Nile, the Rahad and the Dinder are through the black cotton soil plains of the eastern Sudan, which are either wooded or covered with dense grass in the south. The Atbara and its tributaries in their upper courses drain the northern slopes of the Abyssinian plateau, and traverse the level plains of the eastern Sudan in a direction parallel to the Blue Nile.

The Albert Nile and its tributaries between Gondokoro and Lake Albert traverse the broken and hilly country which is cut through by the Albert Nile at the Fola and succeeding rapids. The catchment basins of Lakes Victoria and Albert are the undulating hills, the flat marshy valleys, the great lakes and, in parts, high hills which constitute the highlands of Central Africa. The general level of the area may be taken as 1400 metres above sea level.

The area draining into Lake Victoria is 240,000 square kilometres. At the outlet of Lake Albert this has increased to 380,000, and at Gondokoro to 470,000. The Gazelle river drains 470,000 square kilometres, and the Sobat 160,000. The White Nile drains altogether 1,690,000 square kilometres, or more than half the total area of the catchment basin of the Nile. The Blue Nile drains 300,000 square kilometres and the Atbara 240,000. The Nile below the Atbara junction is draining 2,290,000 square kilometres. Between the Atbara mouth and the sea, the Nile drains whatever falls on a desert area of 720,000 square kilometres.

If we take 3,000 cubic metres per second as the average annual flow past Assuân we may say that the White Nile supplies 24% off more than half the area of the whole basin, the Blue Nile 65% off ¹⁄₁₀ the area, and the Atbara 11 % off ¹⁄₁₂ the area. The Gazelle river drains about ¹⁄₆ the total area and adds practically nothing to the discharge. [Table 24] should be very carefully studied by any man who wants to understand the Nile. It does not pretend to exactitude, but embodies the best information I have been able to obtain.

6. The climate of the Nile valley.

—This paragraph would have been much more complete if Capt. Lyons’ monograph on the Meteorology of the Nile valley had been published. In considering the climate I shall follow the subdivisions of the catchment basin of the Nile contained in [Table I].

In the catchment basins of Lakes Victoria and Albert, the mean annual rainfall may be taken as 1.25 metres, with great fluctuations between good and bad years. Neglecting here and through this paragraph, the light occasional falls of rain which are trying to travellers but which have no effect on the rivers, it may be said that in these basins there are two rainy seasons, the greater in March, April and May, and the lesser in October, November and December. The former are followed by dry southern winds, while north winds blow in the winter.

Along the whole of the Albert Nile, the mean annual rainfall may be taken as 1 metre, with severe famines in occasional years and heavy rainfall in others. The principal rains are between May and November, with the maximum between August 15 and September 15. In years of deficient rainfall, the June, July and August rains seem to fail. The catchment basin of the Gazelle river may be credited with a mean annual rainfall of 75 centimetres between May and October, while the mean annual rainfall on the Arab river cannot be more than 30 centimetres between June and September. The Sobat river in its upper reaches enjoys an annual rainfall of about 1.25 metres and of ·75 metres in its lower reaches. The time of rain is between March and September. The lands draining into the White Nile north of Tewfikieh have an annual rainfall of about 20 centimetres between June and September.

The Abyssinian part of the catchment basin of the Blue Nile enjoys a good rainfall throughout nine months of the year from February to October, with generally heavy rain between May and September, and very occasionally in October. The rainfall here may be taken as 1.25 metres per annum. In the plains of the eastern Sudan traversed by the lower reaches of the Blue Nile and the Atbara the rainfall is very much lighter and may be considered as 30 centimetres between July and September; fairly constant and heavier in the south, and very inconstant and lighter in the north. The Atbara and its tributaries in their upper reaches on the northern slopes of Abyssinia, have rain from May to the end of August and occasionally into September. There are great fluctuations in the rainfall. The mean annual rainfall may be taken as 75 centimetres.

The desert area between Khartoum and Cairo has occasional winter rains especially in the parts near the Red sea, but as these rains are nearly all soaked up by the desert, and very little, here and there, reaches the Nile, we may neglect them altogether. Railways have to be provided with culverts and bridges where they cross the terminal reaches of the khors and wadis which run considerable bodies of water for a few hours after rain; but the effect on the Nile is practically nothing. Along the sea-board of the Mediterranean there are a few inches of rain every winter, sufficient as a rule to raise poor crops of barley.

In the catchment basins of Lakes Victoria and Albert the direction of the winds may be taken as north-east in winter and south-east in summer. The maximum monthly temperature may be taken as 35° and the minimum as 12°, with a mean for the year of 21°.

Along the Albert Nile the north wind blows through the winter, and southerly winds prevail from about the 15th of April to October. The temperature may be taken as ranging from a monthly maximum of 38° to a monthly minimum of 16°, with a mean of 27°.

[Tables 75] to [81] give the principal meteorological data for many places in the Nile valley and compare the Bombay rainfall with the Assuân gauges. The latter show how closely the Assuân gauge in flood follows the rainfall at Bombay.

At Addis Ababa in the highlands of Abyssinia, the mean monthly temperature ranges between 19° and 15°, with a mean of 17°. The winds are south-east and east through the year. In 1902 the rainfall was 980 millimetres, and in 1903 it was 1340 millimetres. ([Table 75]).

At Wad Medani in the Grezireh south of Khartoum, the mean monthly temperature ranges between 35° and 24°, with a mean of 30°. The winds from October to April are from the north and from May to September from the south. In 1902 there were 350 millimetres of rain and in 1903 there were 310. ([Table 76]).

At Khartoum the mean monthly temperature ranges between 34° and 19°, with a mean of 28°. The winds from October to April are from the north and from May to September from the south. In 1902 there were 120 millimetres of rain and in 1903 there were 70. ([Table 77]).

For Alexandria, Cairo and Assuân, representing the Nile valley in Egypt, I have prepared the following table:—

From the above table we may conclude that at Alexandria, Cairo and Assuân the absolute maximum thermometers may be taken as 40°, 45° and 47° Centigrade; or 104°, 113°, 116°, Farenheit. The minimum thermometers as 5·5°, -0·7°, 4·0° Centigrade; or 42°, 31°, 39° Farenheit. The rainfall at Alexandria, Cairo and Assuân respectively may be taken as 235, 31, and 0 millimetres; or 9, 1¹⁄₄, and 0 inches. The heaviest rainfall in any individual year at Alexandria and Cairo respectively has been 308 and 55 millimetres; or 12 and 2 inches. The lightest rainfall at Alexandria and Cairo respectively in any individual year has been 108 and 7 millimetres; or 4¹⁄₄ and ¹⁄₄ inches. Assuân is practically rainless. It does rain sometimes at Assuân, but there has been no rain during the last three years while meteorological observations have been taken.

7. The Geology of the Nile Valley.

—South of Gondokoro along the Victoria and Albert Niles, and at the lakes, the rocks are generally granites, crystalline schists and quartzites. The hills of Uganda are covered with red clay and marl on the higher lands, while the valleys consist of a rich black loam. All the cataracts are granites and granitic rocks or diorites. The Ruenzori range consists of lofty volcanoes. The surface of the ground is covered with a fine Kankar (nodulated limestone) in many places. North of Gondokoro the plains are formed of sandy deposits mixed with coarse peat in places. The hills of the Bahr-el-Gazelle and Arab river are all crystalline. Abyssinia is a volcanic plateau. It is the detritus of this rich volcanic soil swept down by the Blue Nile and Atbara which constitutes the richness of the soil of Egypt and of the water of the Nile. Those parts of the eastern Sudan south of Khartoum and El-Damer and at Kassala, which are the deltas of the Blue Nile, the Dinder, the Rahad, the Atbara, and the Gaäsh, are possessed of a soil in every sense similar to that of Egypt itself. At Khartoum and in the bed of the Blue Nile at Kamlin are extensive deposits of nodular limestone corresponding to the Kankars of India.

The main Nile from Khartoum to Assuân flows between low hills and tables of Nubian sandstone overlying crystalline rocks of gneiss, mica schists, hornblendic granite and red granite. Where the crystalline rocks come to the surface we have cataracts; where the Nubian sandstone is at the surface we have reaches of unbroken water.

From Assuân[4] to near Edfu the Nile flows between hills of Nubian sandstone, the best known of which is Gebel Silsila. From Edfu to near Luxor, the Nubian sandstone which overlies the crystalline rocks dips under the Nile and its place along the Nile Valley is taken by green and grey clays containing nitrate and phosphate deposits. The former are inexhaustible and have constituted the manure of this part of the valley for thousands of years. With these deposits are thick banks of soft white limestone.

[4] Condensed from a description of the geology of the Nile Valley in Egypt written by Capt. Lyons for the second edition of “Egyptian Irrigation”.

From Luxor northwards the clays dip under the Nile and the Nile Valley is bounded by the superposed white eocene limestone up to Cairo.

The Nubian sandstone is always soft and porous. The limestone is generally soft, though hard siliceous beds are sometimes met with. North of Cairo there is no building stone of any value except the siliceous sandstone of Gebel Ahmar near Cairo and the basalt of Abu Zabel, a recent outcrop furnishing a black rock of great durability. The area covered by this rock is small.

Thick deposits of sand and gravel underlie the Nile mud deposits of the Nile Valley. All along the Nile, but especially south of Luxor, river deposits of dark sandy mud exist on either side of the Nile Valley considerably above the level of the deposit of to-day. The best known of these is the plain of Kom Ombos. The thickness of the layer of Nile mud in the valley is as much as 18 metres in places, but the average depth is, I should say, 10 metres.

8. The discharges of the Nile and its tributaries.

—Reference should be made to [tables 24] and [25] which embody the results of an exhaustive examination of the observed discharges, the cross sections, the gauges of the Nile Valley, and the calculated discharge tables made for these gauges. Many of these tables are founded on only two or three discharges and some on only one, but they have been prepared with the greatest care and referred to all the existing gauge observations, and are good working tables, which can be modified and improved as time places more information at our disposal. Until then they may be used as about the best approximations available to-day.

In 1902 the Albert Nile discharged 600 cubic metres per second as against 520 discharged by the Victoria Nile. In 1903 the Victoria Nile discharged 730 cubic metres per second and the Albert Nile 800. Leaving a poor year like 1902 which was much below the average, and taking 1903 which was all round a good average year and only slightly below the mean, we have the following results:—

The Victoria Nile was at its highest in July with 840 cubic metres per second, while the Albert Nile at its head was at its highest in December with 1,060 cubic metres per second. Lake Albert took 5 months to fill up. At Gondokoro the Albert Nile was at its lowest in April when it discharged 700 cubic metres per second as against 550 cubic metres in the previous year. Swollen by timely and good rains south of Gondokoro the discharge at Gondokoro rose to 2,100 cubic metres per second in September after the river had scoured out its bed over a metre in depth. The mean discharge for the year at Gondokoro was 1,200 cubic metres per second.

The Gazelle river gave no discharge in the first half of the year and about 30 cubic metres per second in the latter half. Its mean discharge was 10 cubic metres per second for the year.

The Albert Nile at its tail above the Sobat junction gave as a minimum 350 cubic metres per second in March, which discharge rose to 430 cubic metres per second in September, but could not rise higher as the Saubat river was then in flood and the White Nile could not carry off much more than the discharge of the Sobat without putting the northern part of the Sudd region under 2 metres of water. This held back water helped later to maintain the discharge of the White Nile in January and February.

The Sobat river gave as a minimum a discharge of 40 cubic metres per second in April, and then rose to a maximum in November of 1,080 cubic metres per second.

At its tail, the Albert Nile gave a mean discharge for the year of 390 cubic metres per second and the Sobat of 550.

The White Nile at its head was at its lowest in April with 400 cubic metres per second and at its highest in December with 1,460 cubic metres per second, with a mean discharge of 940 cubic metres per second. At its tail near Khartoum the White Nile was at its lowest in May with 420 cubic metres per second and at its highest in October with about 1,700 cubic metres per second. As this latter figure was about 400 cubic metres per second more than it was receiving at its head, the additional water represented Blue Nile water which had run up the valley of the White Nile, been stored there while the Blue Nile was high and then been discharged into the Main Nile when the Blue Nile had fallen. The mean discharge at the tail of the White Nile was 830 cubic metres per second. This figure was much below that at the head and was due to the fact that in July, August and September the Blue Nile water was flowing up the White Nile.

The Blue Nile was at its lowest in April when it was discharging 120 cubic metres per second. During its maximum in August and September it was discharging 8,200 cubic metres per second. Of the discharge of the Blue Nile in July, August and September, a considerable part flowed up the White Nile which here has a slope of ¹⁄₁₀₀₀₀₀ and a bed from 3,000 to 1,500 metres wide. It is for these reasons that the Blue Nile water does not hurry on to Assuân in its full strength. The mean discharge of the Blue Nile for the year was 2,350 cubic metres per second. Gauges and discharge tables at Kamlin on the Blue Nile, and north of Omdurman on the Main Nile, would be very much better than the Khartoum or Duem gauges of to-day which are both in back waters.

The Atbara river was dry from January to May, in June the discharge was 200 cubic metres per second, rising to 2,300 cubic metres per second in August. In October, November and December it was dry. The mean discharge for the year was 380 cubic metres per second. When the Atbara river rises in flood it cannot flow down the Nile to Egypt in its strength until it has filled up the trough of the Nile as far as the 6th Cataract. Gauges up and down stream of the 6th Cataract and at Shendy would be interesting when compared with Berber.

The minimum combined discharges of the White Nile, Blue Nile and Atbara river were 540 cubic metres per second in April. The maximum combined discharges of 10,900 cubic metres per second were in August. The mean combined discharges for the year were 3,560 cubic metres per second.

The minimum discharge of the main Nile above Assuân was 440 cubic metres per second in May and the maximum of 8,600 cubic metres per second was in September. The mean discharge for the year was 2,650 cubic metres per second.

[Table 25] gives the actual daily minimum and maximum discharges during 1902 and 1903 for each stream, with their dates. For the Blue Nile in 1903 they were 100 and 9,600 cubic metres per second; for the White Nile 380 and 1,470; for the Atbara 0 and 3,100; and for the Nile above Assuân 420 and 9,000 cubic metres per second.

[Table 26] compares the discharges for a maximum year like 1878, a minimum year like 1877, and a mean year, at Khartoum, Assuân and Cairo. The maximum discharges in 1877 were 5,300, 5,900 and 4,400 cubic metres per second, at Khartoum, Assuân and Cairo. In 1878 they were 12,500, 12,100 and 10,300 cubic metres per second respectively, while for a mean year they are 8,500, 9,200 and 7,200 cubic metres per second.

The modulus of the river at Assuân is 3,040 cubic metres per second, and at Cairo 2,640. After the very high flood of 1878, the lowest discharge in May 1879 at Assuân was 1,500 cubic metres per second.

The behaviour of the Nile after passing Assuân and entering Egypt may be described as follows:—of the mean discharge of 3,040 cubic metres per second which passes Assuân 400 cubic metres per second are utilised in Upper Egypt in the irrigation of 2,320,000 acres and 2,640 cubic metres per second pass Cairo. Of these again 540 cubic metres per second are utilised in the irrigation of 3,430,000 acres in Lower Egypt, and only 2,100 cubic metres per second reach the Mediterranean sea.

CHAPTER II.
The tributaries of the Nile.

9. Lake Victoria Nyanza.

—Lake Victoria, the true source of the Nile, lies on the Equator, and fed by abundant rains and numerous streams, discharges its surplus waters over the Ripon Falls, and gives birth to the Victoria Nile. Its most important feeder, the Kagera, whose southernmost tributary rises in the Kangosi hills 2000 metres above sea level in south latitude 4°, has a length of some 600 kilometres. The direct line across the lake from the mouth of the Kagera to the Ripon Falls is 220 kilometres, so that in academical language the length of the Nile at the Ripon Falls is already 820 kilometres. Lake Victoria lies 1129 metres above sea level, and has an area of 60,000 square kilometres; though until the parallels of longitudes are definitely settled, the lake may be credited with an area of between 60,000 and 65,000 square kilometres, constituting it the largest fresh water lake in the old world. Its waters are beautifully clear and perfectly sweet. The depth of the lake is not great and it is covered with many islands. The greatest depth found by Commander Whitehouse in the northern half of the lake has been 73 metres, while the bays are shallow. The northern, southern and eastern shores of the lake, as described by Sir William Garstin, are generally clear, while the western shore, especially at the mouth of the Kagera, is flat, marshy and covered with papyrus. The country surrounding the lake is undulating or hilly and rises to a height of 700 metres above the lake at the south-east corner. The rocks are generally granites, chrystalline schists and quartzites, etc. The hills are covered with red clay and marl on the higher lands, while the valleys consist of a rich black loam.

The catchment basin of the lake is 244,000 square kilometres of which 60,000 are water. Most of the important streams feeding the lake traverse extensive marshes and swamps and must lose a great part of their waters by evaporation. According to Capt Lyons (Appendix III of Sir William Garstin’s report), the climate of the lake basin is typically that which is known as equatorial; two rainy seasons and two dry seasons make up the year, the rains coinciding more or less with the equinoxes and the dry seasons with the solstices, except that the second minor rains are delayed about 1 to 2 months after the autumn equinox. As Capt. Lyons hopes soon to publish a monograph on the meteorology of the Nile valley, I shall say little about the details of rainfall of the different catchment basins, contenting myself with broad principles and main features. March, April and May form the greater rainy season, and October, November and December the lesser. The rainfall of the former season may be considered twice as heavy as that of the latter, but it is the latter which practically decides the height of the lake in the following year. This, according to Capt. Lyons, is due to the fact that in the summer months, when the rainbelt lies to the north of the lake, the dry south winds must blow across the lake basin even though the diurnal reversal of winds on the lake is not mastered by them. These dry winds greatly increase the evaporation, and there is a marked diminution of the water between July and November, which must be primarily due to the increased evaporation.

PLATE III.

RIPON FALLS
Plan and Section
Scale 1 : 6.000

Lith. Sur. Dep. Cairo.

[Larger illustration] (210 kB)

Victoria Nile upstream of Ripon Falls

The rainfall in the catchment basin may be taken as 1250 millimetres per annum on the average. As the evaporation off the lake is probably the same, the area of the lake may be left out of the catchment altogether. The balance of the catchment basin amounts to 184,000 square kilometres, on which there is a mean annual rainfall of 230 cubic kilometres. The mean discharge of the Victoria Nile over the Ripon Falls appears to the approximately 580 cubic metres per second or 18 cubic kilometres per annum. This represents about ¹⁄₁₂ the mean rainfall. The greatest discharge of the lake seems to be about 850 cubic metres per second and the lowest 450. As the lake has risen in a single year 80 centimetres, which represents an increase of water of 48 cubic kilometres, and has fallen 60 centimetres which represents a decrease of water of 36 cubic kilometres, it will be seen that the discharges from the lake are factors of less importance in determining the level of the lake than the heavier rainfall and diminished evaporation in a year of good rain, and the lighter rainfall and increased evaporation in a year of poor rain. The great function of Lake Victoria in the economy of the Nile supply is the insuring of a nearly constant discharge of water into the Victoria Nile, and providing much of the evaporation which comes down in the catchment basin itself in the shape of rain.

The principal feeders of Lake Victoria are the following streams:—

The northern and western feeders are generally perennial streams, while many of the southern and eastern are torrents.

10. The Victoria Nile.

—From the Ripon Falls to Lake Albert, the Victoria Nile has a length of 400 kilometres. The first 64 kilometres are down a steep slope, in a stream varying from 300 to 500 metres in width. Any project for a regulator at the Ripon Falls should contemplate development of electricity for working a railway along these 64 kilometres. The next 237 kilometres are through a flat marshy land, partly lake, partly swamps, but with the water never more than 4 to 6 metres deep. In this reach the Nile is navigable. This many-armed swamp is known as Lake Choga, whose western end is traversed for some 80 kilometres by the Nile with a perceptible current. These large sheets of papyrus and water, which cover an area of over 2000 square kilometres, must cause as much loss by evaporation as they receive by direct rainfall. The Victoria Nile leaves the lake in a broad stream some 900 metres wide, past Mruli station, on to Fowera. In the longitudinal section on [Plate II], I have considered Fowera as 1060 metres above sea level, and the slope upstream of it as ¹⁄₂₀₀₀₀. From Fowera to the foot of the Murchison Falls, the Nile falls 377 metres on a length of 68 kilometres, and then in the next 30 kilometres reaches Lake Albert. Considerable quantities of floating pistea weeds pass down the Victoria Nile from Lake Choga. They are well churned up between Fowera and the Murchison Falls, but finally enter Lake Albert. It would be easy to develop electricity to work a railway along these 68 kilometres.

PLATE IV.

CROSS SECTIONS of the NILE & its TRIBUTARIES
Horizontal Scale 1 : 2.000
Vertical Scale 1 : 600

Lith. Sur. Dep. Cairo.

[Larger illustration] (300 kB)

The principal tributaries of the Victoria Nile are the following:—

On the right bank, the Gogonia, the marshy Kwania and Lenga, and the Duki. And on the left bank; the lake-like Sensiwa; the Kafu with its numerous feeders, large catchment basin but generally insignificant discharge; and the Titi. The Uganda rivers are more like sponges than rivers, and in all probability the tail of the Victoria Nile has a discharge only slightly in excess of that at the head.

11. The Semliki River.

—The Semliki river has its sources in Lake Edward just south of the Equator and flows into the southern end of Lake Albert, just as the Victoria Nile flows into the northern end. Lake Edward with the addition of Lake Dueru has an area of some 2500 square kilometres and lies about 965 metres above sea level. Its waters as described by Sir William Garstin, are green in colour and brackish. The Semliki river discharges the surplus waters of the lake, and may be considered as having a discharge of 100 cubic metres per second as a minimum and 400 cubic metres per second as a maximum. In some 120 kilometres of its length this river traverses the eastern end of the Congo forest, and the day may not be far distant when this timber will be floated down the Semliki, towed across Lake Albert and sent down the Albert Nile to Gondokoro and Khartoum. The fall of the Semliki, according to accepted levels is 285 metres in 260 kilometres. The feeders of Lake Edward are the following:—

On the west, none.

On the south, the Rutshuru and Ruendu.

On the east, the Msalala, the Wampuna, the Chambura, and the Mpanga which has its sources on the slopes of Ruenzori.

On the north, numerous glacier-fed torrents from the slopes of Ruenzori, the Nyamgasha and the Dibirra. The dry season discharges of all these rivers are small. The largest is under 10 cubic metres per second. The flood discharges are considerable.

12. Lake Albert Nyanza.

—Lake Albert has an area of 4500 square kilometres and lies about 680 metres above sea level. We have left the high lands of Uganda and are in the stifling heat of tropical Africa. There are considerable deposits of salt along the east shore of Lake Albert, as there are also around Lake Edward. According to Sir William Garstin, the waters of Lake Albert are brackish near the shores, but perfectly sweet and clear in the middle of the lake. It is due to the waters of Lake Victoria that those of Lake Albert are sweet. The catchment basin of the Albert Nile at its head where it leaves Lake Albert is about 379,000 square kilometres of which 244,000 square kilometres discharge into Lake Victoria. The rainfall over the additional 135,000 square kilometres may be considered as 1250 millimetres per annum, with seasons similar to those on Lake Victoria. The evaporation from Lake Albert must be greater than that from Lake Victoria, but if we exclude the joint areas of Lakes Victoria, Albert, Edward and Choga, and taking them as 70,000 square kilometres, consider that their rainfall is equalized by their evaporation, there remains from the 379,000 square kilometres of catchment basin about 309,000 square kilometres. An annual rainfall of 1250 millimetres means for 309,000 square kilometres a total rainfall of 380 cubic kilometres. The discharges of Lake Albert may be taken as varying from 500 cubic metres per second to 1100, with a mean of 800 cubic metres per second. This latter figure represents in one year 26 cubic kilometres or ¹⁄₁₅th the annual rainfall. We may compare with this the mean discharge of Lake Victoria of about 580 cubic metres per second and ¹⁄₁₂th of the annual rainfall of its catchment basin excluding the lake area.

A reference to [Plate V] will show how great is the regulating effect of Lake Albert on the Nile. Owing to the fact that an increase in the discharge of the Victoria Nile cannot pass down the Albert Nile before the whole area of Lake Albert has risen, the floods of the Victoria Nile are delayed nearly 5 months in their passage down the Albert Nile; a rise of 1 metre on Lake Albert meaning an increased cube of 4,500,000,000 cubic metres. If this takes place in one year, it represents an increased discharge of 150 cubic metres per second irrespective of what passes down the channel of Lake Albert.

If it were considered necessary to insure 1200 cubic metres per second as the discharge of the Albert Nile from the 15th January to the 15th May, it would mean adding 400 cubic metres per second to the mean discharge for 4 months, and deducting 200 cubic metres per second from the mean discharge for the remaining 8 months. By storing the surplus waters of good years by means of a weir capable of holding up 3 or 4 metres of water at the outlet of the lake, it would be possible to insure this discharge every year during the four months which decide the summer contingent of the White Nile to the Nile in Egypt during the months of April, May, June and July.

PLATE V.

GAUGE & DISCHARGE DIAGRAM of LAKES VICTORIA & ALBERT

Lith. Sur. Dep. Cairo.

[Larger diagram] (270 kB)

Luba (subsequently Jinja) gauge, near the Ripon Falls, represents Lake Victoria.
Wadelai gauge (60 km. below Lake Albert) represents Lake Albert.

The shores of Lake Albert are generally steep and barren, though in places they are shelving and covered with papyrus, notably at the inlets of the Semliki river and the Victoria Nile. The Lake is fairly deep and admirably suited for a reservoir. At the outlet of the lake enormous quantities of pistea weeds, especially in high floods, enter the Albert Nile. The passage of these weeds through the future regulator of Lake Albert will be an exceedingly interesting engineering problem.

The principal feeders of the lake are:—

On the north, the Victoria Nile.

On the east, the Waiga, the Wakki, the Hoima, the Wahamba, the Horo, the Ngusi and Msisi, discharging between them under 20 cubic metres per second in the dry season, though good streams in flood.

On the south, the Semliki; and no streams worth mentioning on the west.

13. The Albert Nile.

—The Albert Nile, or the Bahr-el-Gebel, has a length of 1280 kilometres from Lake Albert to the mouth of the Sobat river. For 218 kilometres, past Wadelai to Dufile, it has a broad stream with a sluggish current as a rule, and covered with islands and papyrus marshes. This reach of the Nile is navigable. The fall here must be very little, and it may be considered as 8 metres. In high floods enormous quantities of pistea weeds float down this reach of the Nile. Papyrus and ambatch are very common along the shores and on the islands.

At Dufile begin the Fola rapids followed by numerous cataracts up to Fort Berkeley. In this reach of 155 kilometres the river falls 223 metres. Sir William Garstin states that some of the falls have a width of only 12 metres. The depth must be extraordinary, while the velocity is terrific. The green water of the upper reaches is purified in its passage through these cataracts. The rock is granite. If the regulator for Lake Albert were constructed near Dufile, it should contemplate development of electricity for working a railway along these 155 kilometres.

South of Dufile the principal tributaries of the Albert Nile are on the right bank viz: the Achua and Umi. In the dry season they are dry, but after rains they may add some 50 cubic metres per second to the river. Between Dufile and Fort Berkeley many rivers flow into the Albert Nile. The Asua, the Atappi, the Umi, the Karpetu, the Kweh and many others from the right bank bring down water in flood well laden with sand. In the dry season all but the Asua are dry. The left bank tributaries are numerous but insignificant. The volume discharged by these streams in the rainy season is very considerable, as will be seen if the discharges at Gondokoro north of Fort Berkeley are compared with those at Wadelai. They are between them capable of increasing the discharge of the Albert Nile for short intervals of time by 1500 cubic metres per second. All these streams are torrential. They rise and fall quickly.

From Fort Berkeley to Khartoum, the Albert and White Niles are navigable.

From Fort Berkeley to Bor past Gondokoro on a length of 206 kilometres the Albert Nile falls some 18 metres and has a good velocity and slope, and though the river is divided into two and more channels, it is within its banks and may be considered an ordinary river. The maximum discharge is about 2600 cubic metres per second and the minimum 550. In high floods the river scours out its bed and sides very considerably. The real flat lands begin at Gondokoro a little to the north of Fort Berkeley. The soil is light and sandy and capable of offering little resistance to the stream. Between Gondokoro and Bor the Albert Nile can lose some 30 per cent of its discharge when the river is high and capable of overflowing most of the islands and a great part of the valley. The main channel has a width of about 230 metres and depth of 3 or 4 metres when the discharge is below 600 cubic metres per second and well within banks. The main tributaries of the Albert Nile in this reach are on the east bank, the Kit, second only to the Asua, and the Lokadero; and on the west bank a number of unimportant torrents which however help to swell the river in flood.

From Bor to Ghaba Shambe on a length of 196 kilometres, the main stream of the Albert Nile flows between banks lower than those further south and more heavily inundated in flood, with a width of some 60 metres and depth of water of 5 metres. According to Sir William Garstin, the grass swamps end half way down this reach and the papyrus swamps begin. About 10 kilometres to the east of the main stream is another branch known as the Atem river fed by artificially maintained and natural spills from the main stream itself. These spills are kept open by the Dinkas living along the Atem river. They were noted by Werne in 1842. The Atem river at its tail apparently divides into two branches, of which one feeds the Zeraf river and the other returns to the Nile at Ghaba Shambe.

Capt. Lyons has pointed out to me that all this course of the river is extraordinarily like the course of the Mississippi south of Vicksburg, with its severe curves and oxrings which correspond to the mayahs of the Albert Nile. Such mayahs or lagoons can be seen in the last 20 kilometres of the Rosetta branch of the Nile. Placed as they are between Bor and Ghaba Shambe, they mean that while south of Bor, the Albert Nile has more or less formed its delta; north of Bor, the delta is in a more embryonic stage, with probably the Atem river the more ancient of the two streams. North of Ghaba Shambe, the Zeraf river leaves the Albert Nile, and fed by the Atem Nile, takes off its water eastwards through marsh and swamp to again tail into the Albert Nile below Lake No. About 30 kilometres further north is the cut made by Sir Samuel Baker to enable him to enter the Albert Nile from the Zeraf river.

The real Sudd region lies between Ghaba Shambe and Lake No, on a length of 380 kilometres. Between Fort Berkeley and Gondokoro, the flood as well as the summer supplies are within banks. Between Gondokoro and Bor, the summer supply is well within banks, but the floods overflow the valley. Between Bor and Ghaba Shambe the summer supply is just within banks, but the floods overflow freely. Between Ghaba Shambe and Lake No the summer supply overflows the banks, and hence there result the periodical barrings or sudds of the Nile by the floating vegetation so common in this region. While the summer supply is within banks it can insure a clear waterway; but when not only the floods but the summer supplies ordinarily overflow the banks, the stream must be aided artificially if it is to keep its waterway clear.

The ordinary width of the river south of Hillet-el-Nuer is between 50 and 60 metres, but in the reach of the old sudd blocks numbers 16 to 19 it diminishes in places to 25 metres, and in the reach blocked by sudd block number 15 increases to 200. North of Hillet-el-Nuer, the width varies from 60 to 150 metres, but the mean width may be taken as from 75 to 80 metres. The velocity is about 75 centimetres per second, which is a good velocity for clear water; and Sir William Garstin remarks at one place that since the sudds were cleared the channel of the Albert Nile seems to be deepening and widening itself, and he states in another place that the extent of overflow over the mayahs or side depressions is decreasing.

With proper training works and dredging it should be possible to reduce the length of the main channel of the Albert Nile between Gondokoro and Bor from 175 kilometres to 160; between Bor and Ghaba Shambe from 206 kilometres to 145; and between Ghaba Shambe and Lake No from 380 kilometres to 305. Or the length of channel from Gondokoro to Lake No might be reduced from 761 kilometres to 610. Some of the curves are nearly complete circles, needing but little work to cut off extensive lengths.

Between Ghaba Shambe and Lake No there were 19 sudd blocks of which all but one were removed by Major Peake, R. A., and Lieut. Drury, R. N., in 1900 and 1901. Their positions are indicated on [Plate VI]. South of Hillet-el-Nuer is sudd block No. 15, the only one remaining to be removed. Owing to this block, the Albert Nile leaves its true channel, which is from 5 to 6 metres deep, and, on a length of 43 kilometres, follows a series of mayahs, pools and marshes with a depth of from 1¹⁄₄ to 2 metres. The original channel was closed by a boat laden with ivory sinking in the true channel during the Dervish domination. The boat has been found, the ivory recovered and the work of sudd clearing commenced, but the work has not yet been completed.

The maximum discharge of the Albert Nile at Hillet-el-Nuer may be taken as 450 cubic metres per second, and into Lake No as 320 cubic metres per second. The water of the river is dark-coloured and contains no sediment. Very little ambatch is met with north of Ghaba Shambe.

Between Ghaba Shambe and Lake No on the left bank, in the first 120 kilometres, three channels carrying very appreciable quantities of water flow into the Albert Nile. These are considered to be the tails of the river Yei. At Hillet-el-Nuer, about 170 kilometres north of Ghaba Shambe, a branch takes off from the west side of the Albert Nile, known as Gage’s channel, with 35 metres of clear waterway, 1 metre depth and 0·60 metre per second velocity, discharging 20 cubic metres per second. This stream flows westwards and is lost in the swamps. On the right bank, downstream of Ghaba Shambe, are the two heads of the Bahr Zeraf, one natural and the other artificial made by Sir Samuel Baker. About 90 kilometres north of Ghaba Shambe a few spills take water towards the Zeraf river.

PLATE VI.

THE ALBERT NILE.
FROM GONDOKORO TO THE SOBAT JUNCTION
Scale 1 : 2,000,000

Lith. Sur. Dep. Cairo.

[Larger map] (200 kB)

Part of Albert Nile
South of Sudd No. 15

From Lake No to the Sobat mouth, on a length of 134 kilometres is the tail reach of the Albert Nile. The Albert Nile flows into Lake No in its south-eastern corner and leaves it on its east side. On the west side the Gazelle river flows into the lake. Lake No has an area of some 20 square kilometres in times of low supply and 100 square kilometres in flood. In low supply the depth is between 1·50 and 2·50 metres. In the reach between Lake No and the Sobat mouth, the river is very fairly straight; the summer channel is about 170 metres wide and as a rule about 5 metres deep upstream of the mouth of the Lolle, though in one place about 8 kilometres from Lake No it becomes only 50 metres wide for a short distance. Between the tail of the Lolle river and the mouth of the Sobat river, the summer width widens out to about 400 metres with 3 metres depth of water. However extensive the marshes may be, the high land forming the limit of the inundation is everywhere in this reach visible from the main stream. The Gazelle river, which flows into Lake No, has a discharge varying from 0 to 40 cubic metres per second. The Lolle river which flows in on the left bank upstream of the Sobat mouth is probably an arm of the Albert Nile or an overflow of the Gazelle and has a discharge varying from 0 to 40 cubic metres per second. On the right bank, the Zeraf river flows into the Albert Nile, with a discharge varying from 30 to 160 cubic metres per second. The Megahid river flows in 30 kilometres below the tail of the Zeraf river and adds an undetermined quota of water.

At its junction with the Sobat river, the Albert Nile discharges from 300 to 450 cubic metres in summer; it does not discharge much more in flood as the waters of the Sobat river in flood hold up the Albert River and make it overflow the low lands up to Lake No and beyond for a depth of 3 metres in high floods. This reach of the Nile is a flood reservoir, which empties itself ordinarily by the end of winter.

14. The Gazelle River.

—The Gazelle river, which flows into Lake No, has a catchment basin of 240,000 square kilometres, with an annual rainfall of 75 millimetres; and an additional catchment basin of 230 square kilometres with an annual rainfall of 30 millimetres; and yet the river discharges nothing in summer and about 40 cubic metres per second in flood. It is the most extraordinary river in the world, often blocked by sudd and invisible. It has practically no banks in flood or in times of low supply, while the waterway varies in width from 6 metres to 90 and in depth from 2 metres to 6 metres. It makes its way between interminable marshes of papyrus and water grasses. The Gazelle river is a deltaic river in a still more embryonic stage than the Albert Nile north of Ghaba Shambe. It, however, performs one good function. It keeps the swamps of the Sudd region full of water, year in and year out; and without it, the water of the Albert Nile would doubtless be lost by percolation in the Sudd region and the White Nile be left high and dry for 3 months every year.

Schweinfurth was the first to ascend and describe this river and its tributaries. Beginning from the east, its principal tributaries are:—the Rohl, the Dyow, the Tondy, the Kit, the Dyûr, the Dembo, the Humr and the Bahr el Arab. With courses of between 450 and 750 kilometres in length, the tributaries fall from 500 to 700 metres in their upper courses of from 300 to 500 kilometres, and in their remaining courses they traverse swamps. Certain of the tributaries may be discharging 10 cubic metres per second in summer and 250 cubic metres per second in flood, but the main stream of the Gazelle seems never to add more than 40 cubic metres per second to the Albert Nile. The whole of the water is evaporated from the weeds, papyrus, watergrasses and open sheets of water which cover a desolate area of 70,000 square kilometres. Evaporation and rainfall balance each other. If the «Singhara» or Indian water nut could be induced to grow in this waste of waters, some profit might be got out of them.

15. The Bahr Zeraf.

—The Bahr Zeraf is the right hand branch of the Albert Nile which leaves the latter river north of Ghaba Shambe and joins it again midway between Lake No and the Sobat river. It is fed largely by the escape water of the Atem river brought down in a traceable channel and in flood by torrents from the hilly region east of Gondokoro. Beginning with a series of marshes and swamps, it gradually establishes its right to be called a river and finally after a tortuous and meandering course of about 270 kilometres tails into the Albert Nile. In its lower reaches it is about 50 metres wide and from 2 to 4 metres deep in summer and 6 to 8 metres deep in flood. Its discharges vary from 30 cubic metres per second in summer to 160 cubic metres per second in flood. In its lower reaches the banks consist of solid earth, proving that at one time it carried water other than that brought down by the Albert Nile.

16. The Sobat River.

—This river drains 156,000 square kilometres lying between the catchment basins of Lake Rudolf and the Blue Nile. The rainfall in the mountainous region of Gallaland is plentiful and especially heavy in autumn, and were it not for the extensive lakes and marshes in its middle course, it would be a torrent in flood. Regulated and restrained by the lakes and marshes, this river has an extraordinarily even rise and fall, as a reference to [Table 24] will certify. It is unfortunate that the Nasser gauge has been read so interruptedly. The Doleb Hilla gauge is in the back water of the Albert Nile and not very reliable. The discharges of the last four years have varied from 40 cubic metres per second in low supply to 1000 cubic metres in flood, though there have been years when the discharges have fallen to zero in summer and when the flood must have exceeded 1500 cubic metres per second. April is the month of low supply and November of maximum flood.

In its last 50 kilometres, the river has a deep, well defined channel between high banks, which are never topped in the highest floods. The width of waterway is about 110 metres and the depth 7 metres in summer and about 10 to 11 metres in flood.

The principal tributaries of the Sobat are the Baro from the north-east and east, and the Akobo and Pibor from the south-east and south. All the tributaries meet and form extensive swamps from which the Sobat has its origin. The village of Nasser is situated on the Sobat near its origin. A gauge has been erected here.

17. The Sudd region.

—The Sudd region of the Albert Nile lies north of Ghaba Shambe and corresponds to that part of the river where not only do the floods overflow the banks, but the summer supplies can do so in many places. It is the delta of the river in a very embryonic stage. There are two main branches to the river, the Albert Nile proper and the Zeraf, which have both been already described. Both these rivers are liable to be blocked by sudd or blocks of living vegetation. These blocks are sometimes as much as 5 metres thick and capable of turning nearly the whole supply of the river out of its course. They are formed of papyrus, weeds and watergrasses, which grow on the half sandy half peaty banks of the lagoons and marshes traversed by the river, and which, under the double action of a rising flood and strong winds, are torn up and driven into the channels wherever they are confined in width, and there jammed into solid masses of floating weeds, filling the whole width of the river, very nearly the whole depth, and sometimes over a kilometre in length. In addition to the local weeds and grasses, there are always at hand in high floods dense masses of pistea weeds which have come from the upper waters of the Albert Nile south of Dufile. While the sudd floats it is not so bad as when it sinks, as it has done at block No. 15 north of Gaba Shambe, where the Nile has left its course for 37 kilometres owing to sunken sudd. When the sudd sinks, its becomes putrid and especially loathsome.

The Sudd region is unmistakably, as Lombardini pointed out years ago, an old lake which has silted up and become full of peat and sand deposits. At one time the lake must have had an extreme length of 400 kilometres and width of 400 kilometres and been a larger sheet of water than lake Victoria. The Sobat river flowed into it, and the Blue Nile may have flowed backwards up the bed of the present White Nile for tens of thousands of years. The north-east corner has been better filled with deposit than any other part.

The dense masses of papyrus and water-grasses which shut out the horizon in every direction intimidated the expedition sent up the Nile by Nero, and it returned northwards without having accomplished anything. From Nero’s time to that of Mehemet Ali little was known of these regions. Mehemet Ali made a determined effort to discover what lay beyond these inhospitable regions, and sent up a well-equipped expedition under D’Arnaud.

One of the earliest descriptions of the Nile between the fifth and tenth parallels of latitude is by Werne, who accompanied D’Arnaud’s expedition sent by Mehemet Ali in 1840-1841. The expedition found the channel of the White Nile and Albert Nile easily navigable between December and March. The Albert Nile between 7° and 9° N. lat. had apparently a mean width of 120 metres, depth of 5 metres, and velocity of about 60 centimetres per second, giving a discharge of some 400 cubic metres per second. In this first description of the river the fact that strikes one most forcibly is the omission of the Bahr Zeraf. Neither the inlet nor the outlet are mentioned, though the Sobat, the Gazelle, and numerous insignificant streams are minutely recorded. Practically the whole of the water was confined to one stream, and that a good one. The water level in winter was found to be some 50 centimetres below the general level of the berm, and about 60 centimetres above this level in flood. The swamps contained offensive and fetid water, which mixed with the waters of the rising flood and helped to pollute the stream on the first rise of the river. Between the river and the swamps in its southern reaches were numerous cuts and openings, some natural and some artificial, made by the aborigines for fishing purposes. While traversing the swamps, the waters of the river in flood lost their silt and became quite clear. On page 100 of Vol. II of Werne’s work there is this significant sentence: “The report that the natives below (i.e. in latitude 5° to 7°) had blocked the river to cut off our retreat, turned out to be unfounded.” From the above it will be noticed that the aborigines in 1840-1841 spoke of their ability to block the course of the river, while the training works in the side channels and spills for fishing purposes were described as solid works regulated by rows of strong stakes driven into the ground.

Between 1841 and 1863 the expeditions up the Nile considerably increased, while the aborigines were being brutally treated by the slave traders. What could be more natural than that, as a measure of protection, the aborigines should have widened and deepened the side channels and spills which took off from the river between latitudes 5¹⁄₂° and 7¹⁄₂°, so that they might escape from the traders. Subsequently, when the main river was patrolled by Government boats, the slave-traders themselves used these side channels for prosecuting their traffic. All the channels and spills tailed into the Bahr Zeraf, which now began to form an appreciable stream, and which was navigated over the lower part of its course by Petherick between 1853 and 1862. The Bahr Zeraf was however always described as sudded, while the Albert Nile was open to navigation. This action of dissipating the waters of the river went on increasing till 1863, when there occurred a very high flood indeed; the floating weeds brought down from the south were excessive, the waters escaped everywhere from the main stream, while the floating masses of creepers were confined by the grasses and papyrus to the main channel, and sudded it downstream of Lake No.

On ascending the White Nile and Albert Nile in January 1863, Sir Samuel Baker found the passage clear to the south. On returning in April 1865, he found the sudd of the flood of 1863 still in the Albert Nile downstream of Lake No. The sudd was 1000 metres long and had a passage 3 metres wide cut through the middle of it, down which the river ran like a mill race.

In February 1869, Dr. Schweinfurth and his party, on their way to the Gazelle, took six days to get through this sudd, though the main obstruction was now only 200 metres long. In July 1872, Dr. Schweinfurth on his way back found the sudd to the downstream of Lake No as before, and described the opening through it as “a narrow stream of water which rushed along as a wild brook. The depth of the fairway varied from 2 to 3 metres, and the boat nowhere touched the bottom.”

In February 1870 Sir Samuel Baker found the sudd in the Albert Nile impossible for his expedition of heavily laden boats and steamers. He found the Zeraf sudded in its southern 100 kilometres, and tried to cut his way through but failed. And yet the slave traders had means of getting slave boats down the Bahr Zeraf (“Ismailia” pages 61, 62 and 29). Baker returned in January 1871 to the Zeriba Kutchuk Ali on the Bahr Zeraf and had before him the 100 kilometres of sudded channel. With the aid of 1200 men he completed the work by March 13. The final operation was a canal through stiff clay 600 metres long (known to-day as Baker’s cut). The fall from the Albert Nile into the Bahr Zeraf was so great that, in order to get the boats and steamers across the final distance, Baker made a dam 120 metres long across the Zeraf by means of a double row of piles, sand bags and fascines of the tall grasses. He thus secured the necessary depth of water, and the flotilla sailed into the Albert Nile. On his way back in June 1873, he thus describes the appearance of the head of the Ziraf river where he had made the cut in March 1871:—

“On arrival at the Bahr Zeraf cut, we found that the canals which we had formerly cut were much improved by the force of the stream. Although these passages were narrow, they had become deep and we progressed with comparatively little trouble.” The rest of the journey down the Bahr Zeraf was easily performed.

In January 1874, when the river was low, the sudd in the Albert Nile was removed by Ismail Pacha Ayoub, Governor General of the Soudan.

From 1874 to 1878, while Gordon was Governor General, the Albert Nile was clear of sudd, but the wide stream of 1840 had dwindled down to a clear waterway free of weeds on a width of 6 metres over long reaches. The escape of water down numerous spills had deprived the Albert Nile of the power of keeping its channel clear and when the heavy flood of 1878 came down, the river was sudded.

In 1880 Gessi was blocked in the Bahr Gazelle. The sudd in this Bahr was cut by Marno, who also cut the sudd in the Albert Nile in April 1880.

Emin Pasha mentions the fact that the Albert Nile was free of sudd and navigated from 1880 to 1883. In 1884 he states that no steamers reached Lado, but he attributed that to the Mahdi’s rebellion.

During the early years of the Mahdi’s and Khalifa’s rule there was no sudd. In the Khalifa’s time a boat laden with ivory sank in the stream where sudd block No. 15 is, south of Hillet-el-Nuer, and caused the block to form.

In 1898 Lord Kitchener found the Albert Nile sudded south of Lake No, and in March 1899 Sparkes bey, of the Egyptian Army, steamed up the Bahr Zeraf to within 30 kilometres of its head.

At the same time Sir William Garstin thought the Bahr Zeraf a stronger stream than the Albert Nile.

Descending the Albert Nile from Uganda, Colonel Martyr found the Albert Nile sudded 30 kilometres north of Ghaba Shambe.

In 1900 and 1901 Major Peake, C. M. G., R. A., and Lieut. Drury, R. N., removed sudd blocks Nos. 1 to 14, and 16 to 19, between Lake No and Ghaba Shambe. There now remains only block No. 15 south of Hillet-el-Nuer.

The condition of the channel to-day has been described under the heading of the Albert Nile. It is a very fair channel except at the diversion round block No. 15, which Sir William Garstin is very eager to see removed.

18. The White Nile.

—The White Nile stretches from the Sobat mouth to Khartoum and has a length of 838 kilometres and very little fall. It everywhere bears traces of having been the channel of the Blue Nile when in ancient times the Gebel-Royan hill had not been cut through by the Nile, and the Blue Nile itself flowed south into the great lake which is to-day the sudd region of the Albert Nile. If the Blue Nile discharged, as it does to-day, about 2500 cubic metres per second throughout the year, the Sobat 600 cubic metres, the Albert Nile 1000, the Gazelle tributaries 700, and the rainfall on the lake itself was 1 metre per annum, while the evaporation was 2¹⁄₂ metres, (all reasonable figures), the water entering the lake was 300 cubic kilometres per annum and the evaporation was the same, provided the lake had an area of 120,000 square kilometres, which is, moreover, reasonable when we examine the plan. During the whole of this period, the valley of the Nile in Egypt received its water from the Atbara alone.

The waters of the Sobat river in flood give its name to the White Nile. At Tewfikieh, near the head of the White Nile, is a gauge. For the first 500 kilometres the river is described by Sir William Garstin as having a waterway of from 300 to 500 metres in width with numerous islands. The depth of water in summer is 5 metres and 7 metres in flood. On either side of the waterway is a low ridge swamped in flood, and beyond that on either side is a deep depression, deep in the centre and rising to the ridge on one side and to the high land and forest on the other. Each depression may be 3 kilometres in width where it is wide and a few hundred metres where it is narrow, so that the flooded valley may have a width of 6 kilometres in places. The ridges are about 3 metres wide, and broken by openings through which the water passes in and out of the side marshy depressions. The depressions are covered by a dense growth of reeds and papyrus. When the reeds are burnt one can traverse the marshes on foot. South of Kaka (200 kilometres north of the Sobat mouth) the depressions are lower than they are further north, which, to me, goes to prove that the channel of to-day is formed within the channel of old days when the Blue Nile was flowing south into the Sudd region. At Gebelain (250 kilometres north of Kaka) the side depressions contract and the forests come nearer the river. At the Abu Zeid ford, 50 kilometres further to the north, is a serious obstacle to navigation when the river is low, in the shape of a very broad sheet of shingle studded thick with fresh-water oysters. This bar is 6 kilometres long, as hard as stone, and has in very low summer supplies a depth of water over it of only 50 centimetres. It is a wonder that a channel has not been blasted through it. The swamping now visibly decreases and the width of the river varies from 700 to 900 metres. Some 25 kilometres north of Abu Zeid the papyrus and sudd grasses disappear, and though there is flooding there are no swamps. We have now some well cultivated islands in the river for the negroes have come to an end and the Arabs inhabit the country. The summer channel may be now considered as 700 metres wide and the flood channel as 1300 metres. The summer depth of water is 4 metres. At Duem, 220 kilometres south of Khartoum, the width of the channel widens from 900 to 1000 and further north to 1500 metres, and finally to 3000 metres. We are in a lake rather than in a river, and in flood when the waters of the Blue Nile travel 300 kilometres up the White Nile, and wait for a fall in the Blue Nile to discharge themselves into the Nile, we are indeed in a pulsating lake and not in a river. It must have been in September, when the discharge of the Blue Nile had fallen from some 11,000 to 6,000 cubic metres per second, and the stored-up waters in the valley of the White Nile were forcing themselves down to take the place of those cut off from the Blue Nile, that Linant Pasha took his discharges of the Blue and White Niles and found them some 6000 and 5000 cubic metres per second respectively. The same remark may be made about M. Chélu’s discharge of the White Nile at Khartoum of 4000 cubic metres per second in September 1876 and mentioned in his book “Le Nil, le Soudan, l’Égypte,” page 17.

PLATE IX.

CROSS SECTIONS of the NILE & its TRIBUTARIES
Horizontal Scale 1 : 2.000
Vertical Scale 1 : 500

Lith. Sur. Dep. Cairo.

[Large ilustratration] (190 kB)

In 1903 the minimum discharge of the White Nile at its head was 380 cubic metres per second in April and the maximum discharge was 1470 cubic metres per second in December. [Table 24] gives the behaviour of the river. The minimum discharge of the White Nile at Khartoum may be taken as 300 cubic metres per second. The preceding paragraph will explain how difficult it will be to know its maximum discharge until a gauge and discharge table are established for the Blue Nile 100 kilometres above Khartoum well above back water, and a gauge and discharge table in the Main Nile north of Omdurman. The difference between these two discharges will be the true discharge of the White Nile which, with its slope of ¹⁄₁₀₀₀₀₀ in flood, is not a river but a flood reservoir. The discharges taken at Duem on the White Nile and at Khartoum on the Blue Nile in 1902 and 1903 are interesting, but of little value for anything except the very date on which they were taken. They were all in backwaters.

19. The Blue Nile.

—Compared to any river we have yet described, the Blue Nile is a true mountain stream. Draining the southern and more rainy half of Abyssinia, it is the principal source of the Nile in flood. Whatever waters it receives, it carries to the Nile and it is the true parent of the land of Egypt, for the deposit from its muddy waters is that Nile mud which has made Egypt. The Atbara carries waters which are probably more muddy than those of the Blue Nile, but compared to the Blue Nile the Atbara is a small river, and its quota is insignificant by the side of that of the larger stream. The principal tributary of the Blue Nile, the Abai, rises at a height of about 2,700 metres above sea level and after a course of 110 kilometres falls into Lake Tsana. Lake Tsana lies at a level of about 1,760 metres above sea level, and has an area of about 3,000 square kilometres and catchment basin of about 14,000 square kilometres irrespective of the lake area. On the 31st of January 1903, after a very poor rainfall, Mr. C. Dupuis found the discharge 42 cubic metres per second. Calculating from his cross section, it seems that the maximum discharge may be 200 cubic metres per second. As at Lake Victoria, possibly not more than ¹⁄₁₂th the rainfall finds its way into the lake; and, once there, the greater part is evaporated. Little seems to leave the lake, which would consequently make a very poor reservoir. The land rises from the lake in gently undulating downs as a rule. Wherever observed by Mr. Dupuis, the lake was shallow. There are many islands and some of them considerable ones. The rivers feeding the lake are the following:—the Abai discharging 9 cubic metres per second in January 1903, the Reb 2 cubic metres, the Gumara 2, the Magetch 0·3, the Arno Garno 0·3, the Gelda 0·5, the Unfraz 1·2, and many smaller streams say 1·7, or 17 cubic metres per second in all. This of course was in the dry season. Mr. Dupuis considers the evaporation as 4 millimetres per day.

Between Lake Tsana and Rosaires, on a length of about 750 kilometres the Blue Nile falls some 1260 metres; between Rosaires and Sennar, on a length of 270 kilometres, it falls about 60 metres, and between Sennar and Khartoum on a length of 345 kilometres it falls about 50 metres. The cross sections of the river at Wad Medani, 200 kilometres above Khartoum, and at Khartoum are given on [Plate IX].

In flood and early winter the river is navigable up to the Rosaires cataract. The width of channel may be considered as varying between 350 and 700 metres with an average width of 500 metres. The river rises from 9 to 12 metres in flood and has a velocity in high floods of 3 metres per second. Rivers with such velocities scour out their beds very severely in high floods and deposit silt in low floods, and for the discharges below 5 metres, cross sections should be annually established and discharge tables made depending on the sections. The summer discharge varies from 100 to 300 cubic metres per second and the flood from 7,500 to 12,500 cubic metres per second. The behaviour of the river is shown in [Tables 24] and [25]. The beginning of May is generally low water and the beginning of September high water. The winter discharge may be taken as 500 cubic metres per second.

The tributaries south of Rosaires are the following:—on the right bank, the Folassa, the Durra, the Fatsam, the Bir and the Temsha, veritable torrents; and on the left bank, numerous streams from north and north-west of Addis Ababa, the Anjur, and the Didessa, the latter from 100 to 150 metres wide. North of the last is the Tumat. The Didessa is about 350 kilometres long and the Tumat 200. It is not at all improbable that in the valley of the Didessa far better reservoir sites could be found than at Lake Tsana. Capt. Lyons tells me that there are important reaches here with very little slope. North of Rosaires there are two important tributaries on the right bank, the Dinder, north of Sennar, and the Rahad, just north of Wad Medani. Both these streams run only in flood and are dry in winter and summer. The Dinder has a bed width of about 120 metres, depth 4 metres in good flood and a velocity of 2 metres per second, which gives a discharge of about 1000 cubic metres per second in a high flood. The Rahad has a bed width of 60 metres, depth of 3 metres in a good flood and a velocity of 2 metres per second, which gives a discharge of 400 cubic metres per second in a good flood. The deltas of the Blue Nile, the Dinder and the Rahad are formed of the richest Nile mud. Such soil is rich in lime, potash and phosphates, but is poor in nitrates.

PLATE VII

Lith. Sup. Dep. Cairo.

[Larger map] (250 kB)

THE OUTLET
OF LAKE TSANA
From a rough Survey

The velocity of the Blue Nile may be taken as 75 centimetres per second in low supply and 3 metres per second in high flood.

20. The Atbara.

—The Atbara river flows into the Nile at El-Damer, south of Berber. It is essentially a torrent fed by the rains of north-eastern Abyssinia. The rains here begin early and end early, so that the Atbara is in high flood in August and falls quickly through September. Its floods last from June to October and the river is dry for the remaining months of the year. By dry it is meant that there is no running water, for the bed of the river contains numerous pools of water, which are nearly always deep and often very extensive.

Mr. Dupuis has given a rough longitudinal section of the Atbara river. Rising within 16 kilometres of Lake Tsana, at a height of about 2000 metres above sea level, in its first 300 kilometres it falls 1500 metres to 530 metres above sea level, where it is met by the Salaam river. In the next hundred kilometres it falls 40 metres and is joined by the Settit river, a larger and more permanent stream than the Atbara itself. Sixty kilometres lower down is the Khasm-el-Girba gauge, just upstream of Fasher and about 420 kilometres from the Nile. Two hundred kilometres below the Settit junction and about 280 kilometres from the Nile is Gosrejeb, and 150 kilometres lower down Adarma. Finally, after a total length of about 880 kilometres, the Atbara flows into the Nile.

The Settit junction is about 490 metres above sea level, Fasher 470 metres, Gosrejeb 410, Adarma 380, and El-Damer about 365 metres above sea level. In the last 280 kilometres there is therefore a fall of 45 metres or about ¹⁄₆₀₀₀. In this reach the river has a width of about 330 metres and depth in flood of 6 metres.

[Tables 24] and [25] give the behaviour of the river. In 1902 and 1904, two very low years, the maximum discharge was about 2000 cubic metres per second, and in 1903 about 3000 cubic metres. In high floods the Atbara can discharge 5000 cubic metres per second.

The principal tributaries of the Atbara are the Salaam and Settit already mentioned. On its right bank between Gosrejeb and Adarma it is joined by the Gaash river, which flows past Kassala and loses itself in the deserts. In years of extraordinary rainfall the Gaash reaches the Atbara. The Gaash at Kassala has a width of 150 metres, depth of 1 metre and approximate discharge of 300 cubic metres per second in an ordinary flood. It has a course of about 160 kilometres before it disappears in the desert.

21. The Nile from Khartoum to Assuân.

—The Nile begins its course without any gauge to record its varying height. A gauge north of Omdurman, another upstream of the 6th cataract and a third downstream of the cataract are badly needed. Until these three gauges are erected and recorded, and another erected and recorded on the Blue Nile at Kamlin, about 100 kilometres above Khartoum, the behaviour of the Nile and its tributaries at their junction will never be exactly understood. Making use of the information which is obtainable, we may say that the Blue Nile is generally at its lowest between the 15th April and 15th May with a mean low-water discharge of about 200 cubic metres per second, falling to nearly zero in certain years; it is at its highest between the 15th August and 15th September with a mean maximum discharge of some 10,000 cubic metres per second, rising to 13,000 and falling to 6,500 in maximum and minimum years. If the larger figure is correct, the Blue Nile bank at Khartoum is over a metre too low, and the town is liable to be flooded out. If reference is made to [Plate VIII] it will be seen that the flood of the Blue Nile in July, August and September travels up the White Nile, holds back its waters and converts the valley of the White Nile into a flood reservoir. When the Blue Nile falls rapidly in October and November, the discharge of the Nile is maintained by the stored-up waters in the White Nile and by the White Nile flood which has slowly travelled down its almost level bed. [Table 24] shows this more clearly than any description could. I do not think that the maximum discharge of the Main Nile on any given day is ever equal to the maximum discharge of the Blue Nile.

The Nile between Khartoum and Berber has a channel wider and deeper than that between Wady Halfa and Assuân and a gentler current. I have not taken, or seen any discharges which have been taken in this reach, but judging from what I saw I should say the channel was 800 metres wide on the average. At a distance of 86 kilometres from Khartoum is the Shabluka or 6th cataract. Here the Nile descends 6 metres on a length of 18 kilometres. Two hundred and twenty kilometres below the cataract the Atbara flows into the Nile and repeats on a very small scale what the Blue Nile does at Khartoum. The Atbara is a flood torrent and is dry from October to May. In flood it discharges from a low maximum of 1,700 to a high maximum of 5,000 cubic metres per second, with a mean maximum of 3,500 cubic metres.

PLATE VIII.

BLUE & WHITE NILE GAUGES
FLOOD OF 1903.

BLUE & WHITE NILE GAUGES
FLOOD OF 1904.

Lith. Sur. Dep. Cairo.

[Larger graphs] (190 kB)

In this reach the Nile has a maximum range of 8¹⁄₂ metres and an ordinary range of 7 metres.

Twenty-four kilometres downstream of the Atbara junction is Berber, and 45 kilometres downstream of Berber is the beginning of the 5th cataract, which has a length of 160 kilometres and a drop of 55 metres with three principal rapids, the Solimania, Baggâra and Mograt. The village of Abu Hamed is situated at the foot of this Cataract. Between Abu Hamed and Dongola is the 4th Cataract, which begins at a point 97 kilometres downstream of Abu Hamed, and has a length of 110 kilometres with a drop of 49 metres. In this series of rapids are the Um Dâras and Guerendid. Between the 4th and 3rd Cataracts is a reach of 313 kilometres on a slope ¹⁄₁₂₀₀₀. On this reach is the town of Dongola. The 3rd Cataract has a length of 72 kilometres and a drop of 11 metres with the Hannek and Kaibâr rapids, surveyed and levelled by De Gottberg in 1857. Upstream of the Hannek rapid, on the left bank of the Nile, is the termination of the long depression in the deserts which goes by the name of the Wady-el-Kab and is considered by many as lower than the Nile valley. Between the 3rd and 2nd Cataracts is an ordinary reach of 118 kilometres. West of this part of the Nile are the Selima Wells and according to some travellers an old abandoned course of the Nile slightly above the present high level of the river. This waterless river is said to terminate in the Oasis of Berys which is separated from the Khargeh Oasis by a limestone ridge.

The 2nd Cataract, known as the “Batn-el-Haggar” or “Belly of Stone,” has a length of 200 kilometres and a drop of 66 metres with the rapids of Amâra, Dal, Semna and Abka. At Semna are the rocks where Lepsius discovered the Nile gauges cut by one of the Pharaohs some 4,000 years ago. The Nile flood recorded there is 8 metres higher than any flood of to-day. As the Nile at Semna could be very easily barred by a dam, it struck me when I was there in 1892 that probably King Amenemhat (of Lake Mœris fame) had tried to bar the river with a dam in the hope of creating a reservoir. At Wady Halfa, near the foot of the 2nd Cataract, a masonry gauge divided into metres has been erected and read since 1877. Its accidental zero is R. L. 116.69 and the mean low-water level, or true zero, is R. L. 117.89. Between the 1st and 2nd Cataracts, the Nile has a length of 345 kilometres and a slope of ¹⁄₁₂₅₀₀. The mean width of the river is 500 metres, and the mean depths in flood and summer are 9 and 2 metres. The velocity in summer falls to 50 centimetres per second and rises to 2 metres per second in flood. The river in this reach is generally within sandstone, and the greater part is provided with gigantic spurs on both banks. These spurs perform the double work of collecting soil on the sides in flood and training the river in summer. They were probably put up by the great Rameses 3,000 years ago, as some of the most massive of them have evidently been constructed to turn the river on a curve out of its natural channel on to the opposite side in order to secure deep water in front of Rameses’ temple of Jerf Husain (“Jerf” means steep, scoured bank). The spurs have been constructed with care, and as the courses of roughly-dressed stone can be examined at fairly low water (I have never seen them at absolutely low water) it is evident that there has been no great degradation of the bed during the last 2,000 or 3,000 years. The first, or Assuân Cataract, has a drop of 5 metres on a length of 5 kilometres.

From Khartoum to Assuân, on a total length of 1809 kilometres, there are 565 kilometres of so-called cataracts with a total drop of 192 metres, and 1,244 kilometres of ordinary channel with a total drop of 103 metres.

At the head of the 1st Cataract is the Assuân dam, regulated on for the first time in October 1902. It has 140 openings of 2 metres × 7 metres and 40 openings of 2 metres × 3¹⁄₂ metres.

At the foot of the 1st Cataract, opposite the town of Assuân, on the Island of Elephantine, has stood a Nile gauge from very ancient times. An officer belonging to the Roman garrison in the time of the Emperor Severus marked an extraordinarily high flood on the gauge. The maximum flood-mark at the time of the visit of Napoleon’s French savants was however 2.11 metres higher than the above. As the middle of Severus’ reign was A.D. 200, and the visit of French savants A. D. 1800, they concluded that the bed and banks of the Nile had risen 2.11 metres in 1600 years or 0.132 metres per 100 years. The new gauge divided into cubits and twenty-fourths was erected in 1869 and has been recorded daily since then (a cubit = 54 centimetres). The accidental zero of the gauge is R. L. 84.16. The mean low-water level or true zero is R. L. 85.00.

PLATE X.

LONGITUDINAL SECTION of the NILE from WADY HALFA to GEBEL SILSILAH

Lith. Sur. Dep. Cairo.

[Larger longitudinal section] (270 kB)

[Larger cross sections] (90 kB)

CROSS SECTION OF THE NILE VALLEY NEAR IBRÎM
CROSS SECTION OF THE NILE VALLEY NEAR ASSUÂN
TYPICAL CROSS SECTION UPSTREAM OF ASSUÂN DAM

22. The Nile from Assuân to the Barrage.

—From Assuân to the Barrage, the length of the river is 973 kilometres in summer and 923 in flood. The slope in summer is ¹⁄₁₃₀₀₀ and in flood ¹⁄₁₂₂₀₀ The mean fall of the valley is ¹⁄₁₀₈₀₀. The slopes vary in the different mean reaches, the least being ¹⁄₁₄₈₀₀ in the Kena Mudiria and the greatest ¹⁄₁₁₄₀₀ in Beni Suef. In a high flood with a rise of 9 metres at Assuân, the rise in Kena will be 9.5 metres and only 8.2 in Beni Suef. [Table 42] gives the mean areas of cross sections of the Nile, while [table 44] gives the mean widths. Neglecting spill channels, we may state that in a high flood the mean area of the section of the Nile is 7,500 square metres and the mean width 900 metres. In the Kena Mudiria, the area is 7,000 square metres and the width 800 metres, while in Beni Suef the mean area is 8,000 square metres and the mean width 1,000 metres. Speaking generally it may be stated that where the Nile valley is narrow the slope of the river is small, its depth great and width contracted; while where the valley is broad the slope is great, the depth small and the width enlarged. The mean velocity in flood ranges between 2.0 metres and 1.0 metre per second, while the velocity in summer varies from 0.5 to 0.9 metre per second. We may say that the Nile in soil has a natural section whose width in flood is 110 times its depth, while its mean velocity is 1.50 metres per second.

The natural canals, which take off the river and which never silt, have a mean velocity of some 65 centimetres per second, while the proportion of width to depth is about 12 to 1. Artificial canals of this section do not silt if their velocities are 80 centimetres per second, while silting takes place as readily when the velocity is greater as when it is less than the above. In muddy streams, like the Nile in flood, certain velocities demand certain proportions of width to depth, and if these are not given to it, they will make it for themselves by eating away the sides if they can, or, if they cannot eat away the sides, by silting up and raising the bed.

To the north of Assiout is situated the Assiout weir or barrage across the Nile with 111 openings of 5 metres and 10 metres depth of water in high flood. It was regulated on for the first time in August 1902.

On Roda island, opposite Cairo, has stood a gauge from the earliest times. It has been frequently reconstructed. The present gauge is reputed to have been erected in A.D. 861 with its zero at the same level as a more ancient one whose readings have been preserved since A.D. 641. When the gauge was constructed, a reading of 16 cubits meant the lowest level at which flood irrigation could be insured everywhere. The level to-day is 20¹⁄₂ cubits on the gauge and the difference between them is 1.22 metres. As 1,026 years have elapsed since the construction of the gauge it means a rise of 12 centimetres per 100 years. This is slightly under the rise calculated at Assuân by the French savants.

The following table gives the means of the maximum flood and low water levels per century:—

It is evident from the above that the head of the Delta, or the bifurcation of the Nile, was much nearer to Cairo in early days than just now, and the last three centuries have seen great changes. The fall of watersurface is very considerable at every bifurcation, and the difference between mean high and low supply at the Barrage to-day is 6.0 metres against 7.2 metres at Cairo. Judging from the above figures, I should say that from the 7th to the 13th century the bifurcation was gradually approaching Cairo, while since the 13th it has been receding.

PLATE XII.

LONGITUDINAL SECTION of the NILE from ASSUAN to CAIRO
ALONG CENTRE LINE OF FLOOD

Lith. Sur. Dep. Cairo.

[Larger longitudinal section] (460 kB)

[Larger cross sections] (250 kB)

CROSS SECTIONS of the NILE VALLEY IN EGYPT

The following table gives the highest and lowest floods at Cairo in intervals of 25 years from A.D. 639 to A.D. 1904.

The gauges are in pics and kirats and are referred to mean low water or R. L. 12·25 metres above mean sea.

[5] There is an error here in the records.

At Assuân the Nile has a mean range of 7.90 metres between high and low supply, with a maximum of 9·80 metres and a minimum of 6.40 metres. The high supply varies between 13,200 and 6,500 cubic metres per second, with a mean of 10,000 cubic metres per second, while the low supply varies between 350 and 1400 cubic metres per second with a mean of 590 cubic metres per second. September is generally the highest month and May the lowest. The mean low water level is R. L. 85.00.

At Cairo the Nile has a mean range of 7·00 metres with a maximum of 9·6 metres and a minimum of 5·3 metres. The high supply varies between 12,000 and 4,800 cubic metres per second with a mean of 7,600 cubic metres per second, while the low supply varies between 1,300 and 250 cubic metres per second, with a mean of 500 cubic metres per second. October is the highest month and June the lowest. The mean low water level is at R. L. 12·25.

PLATE XI.

CROSS SECTIONS of the NILE & its TRIBUTARIES
Horizontal Scale 1 : 2.000
Vertical Scale 1 : 500

Lith. Sur. Dep. Cairo.

[Larger cross sections] (140 kB)

[Tables 41] to [52] refer to the Nile between Assuân and the Barrage at the head of the Delta proper.

[Table 46] gives the Reduced Level of the mean low water level of the Nile at various points between Assuân and Cairo. If, for example, it is known that the water surface at any time of the year at Assiout is R. L. 50.80, we know the mean low water by the Irrigation Department levels is 45.05. The gauge is therefore 5.75, and by turning to [Table 37] we know the discharge.