The Nile in 1904

While basin irrigation is followed by the winter crops of wheat, beans, clover, barley, flax, lentils, vetches and onions, perennial irrigation allows of all the above winter crops and in addition the summer crops of cotton, sugar-cane, oilseeds, gardens and orchards. It will readily be understood that all this double cropping necessitates a very free use of manures.

It would be a healthy innovation indeed, if the provision of suitable manures were to be considered as an essential part of a project for providing perennial irrigation. The day is not far distant, I believe, when governments which provide irrigation works will also provide manures, and sell the water and the manures together, one being as essential as the other; I know well, from observation, that a well-manured field needs only half the water that a poorly manured field does; and in years of drought and scarcity manures almost take the place of irrigation. Why should there not be a manure-rate as well as a water-rate? Here in Egypt, the numerous ruins of old-world cities have hitherto provided manure for a great part of the perennially irrigated lands; but these are being fast worked out, and other sources must be sought for. Farm-yard manure will never suffice for the intense cultivation in this country. In connection with this subject, I can recommend the study of a remarkably able paper on “Nile Cultivation and Nitrates,” read by Mr. J. B. Fuller, C.I.E., before the Agricultural Society of England, and embodied in the 3rd Series, Vol. VII., Part 4, 1896. Egypt possesses, in the vicinity of Luxor, natural beds of nitrates of unlimited extent, which come down to the river’s edge. These nitrate beds have been used from time immemorial, but were brought to the notice of the general public by Mr. Floyer. They contain only about 6 per cent. of pure nitrates, but as they are on the edge of the Nile, in a perfectly cloudless and very dry country, it might be possible, with the aid of the plentiful supply of water always at hand to profitably extract pure nitrates. The demand for nitrates is without limit in the Nile Valley, as Nile water, though rich in everything else, is exceedingly poor in nitrates.

The perennial canals and collateral works have cost £4·50 per acre, and the maintenance charges are £·10 per acre. The perennially irrigated lands are let at £5 to £8 per acre per annum as against £3 to £5 for the basins lands.

30. Flood protection in Egypt.

—The Nile during high floods is considerably above the level of the country, which is protected by embankments stretching from Assouân to the sea. In Upper Egypt, a very high flood is one metre above the country; in Middle Egypt it is 2 metres, and the same on the Rosetta branch of the Nile. On the Damietta branch it is 3·50 metres in places.

In parts of Upper Egypt, but nearly everywhere in Lower Egypt, the Nile on curves is protected by stone spurs. These spurs contain each from 4,000 to 250 cubic metres. They are very effective where the Nile bank has been well thrown back below them to a distance of some 50 metres on a length of at least 100 metres. This allows the waters of the flood to swirl harmlessly in whirlpools below the spurs while the banks are far removed from their action.

When we first came to Egypt, we found that the policy was to spread the flood into as many channels as possible and protect the whole of them with tens of thousands of corvée, in addition to the corvée on the Nile banks. We changed that and concentrated our energies on the Rosetta and Damietta branches.

In 1861, 1863, 1866, 1869, 1874, and 1878 the Damietta branch was badly breached, There has been only one serious breach on the Rosetta branch, and that was in 1863. The great breach of 1878 on the Damietta branch was attended with serious loss of life; but far more serious was the breach of 1863 on the Rosetta branch not far from its head. The whole western half of the Delta proper was swept by the river, and as the canals there have not got good high banks, the people had no place of shelter to flee to and were drowned in very great numbers. The same thing would happen again if a breach were to occur now, only the damage would be far more serious. The country is covered with villas and rich plantations and the low lands to the very edges of Lake Borrilos are being reclaimed and inhabited. The loss of life which would occur nowadays would be truly appalling. A breach anywhere within 100 kilometres of the Barrage on the east bank of the Rosetta branch or the west bank of the Damietta branch during a very high flood would be a national disaster.

The terror reigning over the whole country during a very high flood is very striking. The Nile banks are covered with booths at intervals of 50 metres. Each booth has two watchmen, and lamps are kept burning all night. Every dangerous spot has a gang of 50 or 100 special men. The Nile is covered with steamers and boats carrying sacks, stakes, and stone; while the banks along nearly their entire length are protected by stakes supporting cotton and Indian corn stalks, keeping the waves off the loose earth of the banks. In a settlement of a culvert in the Nile bank north of Mansourah in 1887 I witnessed a scene which must have once been more common than it is to-day. The news that the bank had breached spread fast through the village. The villagers rushed out on to the banks with their children, their cattle, and everything they possessed. The confusion was indescribable. A narrow bank covered with buffaloes, children, poultry, and household furniture. The women assembled round the local saint’s tomb, beating their breasts, kissing the tomb, and uttering loud cries, and every five minutes a gang of men running into the crowd and carrying off the first thing they could lay hands on wherewith to close the breach. The fellaheen meanwhile, in a steady, business-like manner, plunged into the breach, stood shoulder to shoulder across the escaping water, and with the aid of torn-off doors and windows and Indian corn stalks, closed the breach. They were only just in time. This is the way the fellaheen faced a breach. And this is how the old Governors of Egypt faced them. During the flood of 1887 I complimented an official on the Nile bank, whose activity was quite disproportionate to his apparent age. He told me that he was a comparatively young man, but he had had charge of the Nile bank at Mit Badr when the great breach occurred in 1878, and that Ismail Pasha had telegraphed orders to throw him and the engineer into the breach. He was given 12 hours’ grace by the local chief, and during that interval his hair had become white; subsequently he was pardoned. These were the senseless orders which used to petrify officials into stupidity.

The following estimate was made by me of the cost of protecting the Delta proper between the two branches of the Nile during the high flood of 1887:—

Cost of protection for 432 kilometres of bank or 1,200,000 acres of cultivation:—

This works out to £120 per kilometre of bank, or £·045 per acre of land protected. It is a very cheap insurance.

CHAPTER IV.
Projects.

31. Projects.

—No account of the Nile in 1904 would be complete without an enumeration and slight examination of the projects before the public for the provision of sufficient water to the Nile in times of low supply to insure the perennial irrigation of the whole of Egypt; to utilise these perennial waters by converting basin tracts into perennially irrigated ones; to protect the country from the dangers accompanying high floods; and to permit of the reclamation of the low salted lands of Lower Egypt which border the Mediterranean sea.

Egypt has a total irrigable area of 6¹⁄₄ millions acres. Of this area, ¹⁄₄ of a million acres, which are to-day inundated in flood and lie along the edge of the deserts, must continue to be inundated in flood for all time, to prevent the sands of the desert from spreading over the Nile Valley. Their value is £5,000,000. Four million acres are perennially irrigated. They have a mean value of £55 per acre, and have a total value of £220,000,000. Of the remaining two million acres, two-thirds are irrigated only in flood and one-third is not irrigated at all. These 2 million acres have a mean value of £25 per acre, and are worth £50,000,000. The land of Egypt may be considered as worth £275,000,000 to-day. If it were possible to perennially irrigate the 2 million acres which are without such irrigation, their value would be increased by £30 per acre, or by £60,000,000.

The problem before us is how to provide perennial irrigation to these 2 million acres and so add £60,000,000 to the wealth of the country.

It has been calculated that each milliard of cubic metres of water stored in reservoirs situated in Egypt itself is sufficient to insure the conversion of half a million acres from flood to perennial irrigation. Egypt therefore requires reservoirs capable of storing 4 milliards of cubic metres of water.

In Mehemet Ali’s time, the great preoccupation of the Government was the pressing on of the cultivation of cotton, and as this crop needed perennial irrigation, the securing of an abundant supply of water all the year round was the problem of the day.

The fame of the ancient Lake Mœris had made a profound impression on the mind of Mehemet Ali, and he urged on his chief engineer the necessity of undertaking similar works. Linant Pasha first set himself to discover the site of the ancient lake, and then estimated roughly the cost of reconstructing it, but considered the cost prohibitive. He recommended Silsila as a suitable site for a weir and a canal head. The failure of the Barrage discouraged the Government from undertaking new works and the question dropped. In 1880 Count de la Motte proposed a dam at Silsila and a reservoir to the south of it. He also proposed putting a capacious depression to the east of Kalabsha in communication with the Nile by the aid of a dam at Kalabsha.

About two years later Mr. Cope Whitehouse suggested utilising the Wadi Rayan depression as a reservoir. This depression had been already mentioned by Linant Pasha in his book and located by him on his hydrological map. Financial difficulties prevented Sir Colin Scott-Moncrieff from immediately considering the question of reservoirs. The success of the Barrage repairs in 1887 however gave new life to the question of reservoirs and Sir Colin Scott-Moncrieff deputed Col. Western to give scope to the suggestion made by Mr. Cope Whitehouse, to make plans of the Wady Rayan and the deserts between it and the Nile, to find out the capacity of the reservoir, and see if it could be utilised. Col. Western’s report, plans and estimates were printed by the Egyptian Government in 1888. At the same time I was deputed to examine the other projects of Count de la Motte. In 1889 and 1891 I reported unfavourably on them, because I could find no depression near Kalabsha to put in communication with the Nile, and could find no rock at Silsila on which to build a dam. The Bergat Takham pan was the only depression near Kalabsha which could have been used as a reservoir and it was over 100 metres above the level of the Nile flood; while both in the Silsila pass and the Silsila gate I bored for rock and was everywhere still in sand 10 metres below the level at which the existence of rock was assumed by the Count’s engineers. On my report reaching Cairo, M. Prompt proposed using the trough of the Nile itself at Kalabsha as a reservoir in place of the depression which did not exist. Col. Western left the country in 1890 and I became Director General of Reservoir Studies. M. Prompt had supposed that rock could be met with at Kalabsha at a depth of 4 metres below low-water level. I could not find it at a depth of 26 metres. After sounding and boring at every possible site on the Nile and surveying, boring, and levelling in the desert between Wadi Halfa and the Fayoum, I submitted my report in 1894, proposing an open dam at Assouan of a type which I trusted would meet the requirements of a Nile reservoir dam. Sir William Garstin approved of the site and the design, and the dam was built between 1898 and 1902 with Mr. Maurice Fitzmaurice C.M.G, as resident engineer.

The Assouan Reservoir at its present level contains one milliard of cubic metres of water which will suffice for the conversion of half a million acres to perennial irrigation, adding £15,000,000 to the wealth of the country. But though the dam was only completed at the end of 1902, already the whole of the water has been devoted to special tracts, and the Government is reluctantly compelled to refuse all applications for water.

32. The raising of the Assuân dam.

—Egypt already possesses the germ of all the storage works she needs. Six years ago a few far-seeing men saw clearly what all of us understand to-day; but among the few, no man had greater faith in the future of the country than Sir Ernest Cassel. The Assouân Reservoir project had been lying buried for four years in official pigeon holes, when in 1898 Sir Ernest came forward with the funds, and with Sir John Aird & Co., as contractors, and Sir Benjamin Baker as Consulting Engineer, undertook to complete the Assouân dam and the Assiout weir by December 1903. The Egyptian Government, advised by Sir William Garstin, accepted his offer, and received the completed works by December 1902.

The Assouân dam is a granite structure 2,000 metres long which crosses the head of the Assouân cataract of the Nile in one continuous straight line. The dam is pierced by 140 under sluices of 7 metres by 2 metres for passing floods, and by 40 upper sluices of 3¹⁄₂ metres by 2 metres for passing the high level water of the reservoir. The sluices are regulated by “Stoney” gates worked by winches at the roadway level.

While the red, muddy waters of the Nile flood are pouring down the river, the whole of the sluices are open and the river discharges itself through them without parting with its silt. This is the real object of the sluices, for if the dam were solid and the river forced to flow over the top, the reservoir would soon be filled with deposit and obliterated, while Egypt, deprived of this rich mud, would be considerably the poorer. This is the great feature of the dam. While the dam holds together, the reservoir will be free of silt.

When the turbid flood has passed and the comparatively clear winter supply of the river has begun to arrive, the sluices are gradually closed and the reservoir filled. Beginning about the 1st December, the reservoir is filled in 100 days. It will ordinarily be full on the 1st March. No additional water is needed for irrigation between March 1st and May 1st as the river naturally has enough for the requirements of the area at present under crop at this season. As the area under perennial irrigation will be gradually increased, the demand for water for the new lands will begin about the 1st of April. The demand increases through May and June, and the reservoir will then be aiding the river with its supplement. If the flood is very late, water may be required from the reservoir till the 10th July; if the flood is early no water will be needed after the 20th of June, as in 1903, the first year of the reservoir. The earlier the flood the more effective the reservoir. By the time the flood has begun to get turbid, the under and upper sluices will all be open and the muddy waters of the Nile will sweep through the dam without impediment.

The dam has worked for two years and given satisfaction. When the Nile was at its lowest in May 1903, the natural discharge of the river, supplemented by all the subsoil infiltration water which enters the river between Assouân and the sea, was 400 cubic metres per second. The reservoir was adding 200 cubic metres per second to the supply, raising the total supply available for irrigation to 600 cubic metres per second. The reservoir was supplying one-third of the water which was being utilized in Egypt. This water will suffice for an increase to the perennially irrigated area of ¹⁄₂ million acres.

At the time of designing the dam it was intended that it should be of such a section that it could be raised 6 metres in height and hold up another milliard of cubic metres of water. Such an operation, if performed to-day, would mean: the whole length of dam being raised 6 metres, the winches working the sluice gates being raised 6 metres and provided with new wire ropes; and new copings being given to the parapets. It would necessitate two more locks and three more lock gates, and nothing else. The expenditure incurred would be about £500,000.

33. The Wady Rayan Reservoir Project.

—When the Assouân dam will have been raised, we shall be standing on the threshold of what it will be able to do. The projected Wady Rayan reservoir, or the modern Lake Mœris, will be well able to supply the two remaining milliards of cubic metres of water when working in conjunction with the Assuân Reservoir. The great weakness of this projected lake has lain in the fact that by itself it could give a plentiful discharge in April and May, less in June, and very little in July, and it was for this reason that in my report of 1894 to the Egyptian Government I had reluctantly to recommend that it be not carried out. But when the Assouân reservoir is capable of supplying two milliards of cubic metres of water it will be possible to utilise the Mœris Lake to its utmost capacity. The Assouan Reservoir, being high above the level of the Nile can give its supply at the beginning or end of the summer; it can give it slowly or with a rush; while the projected Lake Mœris, being directly in communication with the Nile, and only slightly above low Nile level, its discharge would depend entirely on the difference of level between it and the Nile, and consequently as the summer advanced its level would gradually fall and the lake would not be able to give at the end of the summer a quarter of the discharge it could give at the beginning.

But let us imagine that the reservoir and the lake are both completed and full of water, and that it is the first of April. Lake Mœris will be opened on to the Nile and give all the water needed in that month, while the Assouan Reservoir will be maintained at its full level. In May, Lake Mœris will give nearly the whole supply and the reservoir will give a little. In June the lake will give little and the reservoir much; while in July the lake will give practically nothing and the reservoir the whole supply. Working together in this harmonious manner, the reservoir and the lake, which are the true complements of each other, will easily provide the whole of the water needed for Egypt.

The Wady Rayan is a depression in the deserts to the south of the Fayoum and separated from the Fayoum by a limestone ridge. In 1888 Col. Western recommended it very strongly as a reservoir. In this he was supported by Col. Ross, the first Inspector General of irrigation. On Col. Western’s leaving Egypt, the study was entrusted to me, and Messrs. Hewat and Clifton deputed to make a final project. The Wady Rayan project, with its plans and estimates, was published by the Egyptian Government in 1894. As I said before, I was reluctantly compelled to reject its adoption owing to the one radical defect already described. That defect will have been completely removed by the completion of the Assouân Reservoir, when it will be possible to undertake the construction of the modern lake Mœris.

The question of Lake Mœris has interested the world for centuries. For the ancients it was one of the world’s seven wonders. Sir Hanbury Brown, in his book on the “Fayoum and Lake Mœris,” has collected all the information available about the lake, and after a thorough examination of the question has declared in favour of the Wady Rayan being converted into a modern Lake Mœris.

Herodotus, writing about B.C. 450, was the first to describe the lake: “Now the Labyrinth being such as I have described, the lake named that of Mœris, causes still greater astonishment, on the bank of which the Labyrinth was built.

“The water in the lake is not derived from local sources, for the earth in that part is excessively dry and waterless, but it is brought in from the Nile by a canal. It takes six months filling and six months flowing back. During the six months of the return flow, it yields a talent of silver every day to the treasury, and during the flow twenty minæ for the fish.”

Strabo, writing in B. C. 20, remarks: “It has also a remarkable lake called the Lake of Mœris, large enough to be called a sea, and resembling the open sea in colour.

“Thus the Lake of Mœris is, from its size and depth, capable of receiving the overflow of the Nile at its rising, and preventing the flooding of houses and gardens; when the river falls, the lake again discharges the water by a canal at both mouths, and it is available for irrigation. There are regulators at both ends for controlling the inflow and outflow.”

Diodorus Siculus, writing at the same time, says:—“King Mœris dug a lake which is amazingly useful and incredibly large. For as the rising of the Nile is irregular, and the fertility of the country depends on its uniformity, he dug the lake for the reception of the superfluous water, and he constructed a canal from the river to the lake 80 furlongs in length and 300 feet in breadth. Through this he admitted or let out the water as required.”

At one time there was much discussion as to what was Lake Mœris, but since the publication of Sir Hanbury’s book there can be but one opinion. The lake covered the whole of the modern Fayoum below the level of the contour which is 22¹⁄₂ metres above mean sea level. The common Nile shells are to be met in myriads at any point on this contour round the Fayoum that one cares to look for them. The ordinary high flood level of Kushesha basin to-day is 26¹⁄₂ metres above mean sea. In Amenemhat’s time, which was 4,000 years ago, the level was 4 metres lower, or at 22¹⁄₂ metres above mean sea. This was the highest possible level the lake could have attained in his day. In the course of time the level of the Nile valley rose by about 10 centimetres per century, but the frequent occasions on which the canal was kept closed during poor and low floods gradually silted up the channel and made it less capacious. As there are no Nile shells above the contour of 22¹⁄₂ metres above mean sea (except a few on the south side of the lake which have evidently been blown up by the north west winds in sand drifts) it is evident that the gradual silting up of the channel more than kept pace with the rising level of the Nile. Eventually the silting up exceeded the rise, and that at an accelerated rate, the canal became weaker and weaker, and the Fayoum Province gradually occupied the site of the lake. Lake Mœris had lasted over 2,000 years.

The connection between the Nile and the germ of the future Lake Mœris was in existence in King Menes’s time, as I have been informed by professor Sayce, but it was King Amenemhat, of the XIIth dynasty, who widened and deepened the canal, cleared away the rocky barriers, and converted the trifling lake of King Menes’s time into the mighty inland sea which controlled the highest floods of the Nile. Those ancient Pharaohs were giants in hydraulic engineering. They were, moreover, as wise as they were courageous.

Sir Hanbury Brown has well described the action of the lake. It had a surface of 2,500 square kilometres, and being drained back into the Nile and kept at a low level it was able to take from a very high flood 20 milliards of cubic metres of water. It was quite capable of reducing a very high flood to moderate dimensions; and if injudiciously or maliciously opened in a low flood, it was capable of depriving Lower Egypt of any flood irrigation at all; and in those days they had practically no irrigation except flood irrigation.

The Wady Rayan, as already stated, is a depression in the Lybian hills immediately south of the Fayoum. It has, at a level of about 29 metres above the sea, a surface of 700 square kilometres, or about one quarter the area of the ancient lake. Like the ancient lake, the lowest point of the Wady is 41 metres below sea level. When filled with water the greatest depth will be 70 metres. The uppermost four or five metres only will be utilised annually, or some 3 milliards of cubic metres of water out of a total volume of 20 milliards. Just as the great size of the ancient lake was of inestimable value to a work whose principal use lay in moderating high floods, so the smaller area of the modern lake will render it far more useful as a work for feeding the low Nile. This lake, will render no mean aid in time of dangerous floods, but, in its early years, its main use will be the provision of water in summer. It will supply the two milliards which are needed to convert the whole of Egypt from basin to perennial irrigation.

In my book on “The Assuân Reservoir and Lake Mœris” I have worked out the cost of the project and estimated it at £2,600,000.

The rates I have allowed for the excavation work are considered too low by some critics. If the earthwork in the Nile Valley had to be excavated within 30-day rotations as on the running canals, I should be the first to agree; but the work will last three years and the contractors will be able to concentrate all the spare labour of the country on the works when demand for labour is slack, and in this way the rate of P.T. 3 per cubic metre which I have allowed will be found to be ample. In the hill of salted marl it will be possible to employ the American system of excavating by the aid of water issuing from nozzles under pressure. By this method it will be possible to do much work at P.T. 2 and P.T. 3 per cubic metre as it is done in America. I have allowed P.T. 5 per cubic metre. To this hydraulic pressure work the salted marls will be specially suited, and indeed the recollection of the ease with which Amenemhat dug his canal though this very material lasted long in the memory of Egyptians. Some 1,600 years after the canal was excavated, Herodotus was informed that the excavated material was thrown into the canal and transported by the running water. A 12-inch pump on the Yusufi canal lifting water on to the top of the hill, a number of spade men helping the water as it coursed down the hill and leading the liquid mud along wooden troughs into side ravines and depressions, and a steep slope on the western half of the hill where the rock had been blasted away would soon remove all the material required at a very low cost. I have allowed P.T. 10 per cubic metre for the soft limestone. Here it will be easy to work on a vertical face of some 7 metres, blast out the rock, carry it away on four lines of railway running down hill, deposit the rubble on the desert; and as each 7 metres depth is completed, to begin the next 7 metres in depth in the same way.

In my 1894 Report I had anticipated difficulties with the canal running through the salted marl. Since then I have thoroughly inspected the ravines in the Fayoum and seen the El-Bats ravine where it cuts through many kilometres of this very salted marl. The sides are absolutely vertical and deposits of mud and self-sown tamarisk bushes protect the vertical sides at places where the running water is nearly touching the marl. Such natural protection will be far superior to the masonry lining I proposed and far more effective. It will moreover cost nothing.

To those critics who suggest that the waters of the lake might become salted or leak into the Fayoum I have to reply as follows: When the old Lake Mœris, or the present Fayoum, was full of water and 63 metres higher than the bottom of the Wady Rayan, and remained so for thousands of years, there was no question of the waters having become salted or having escaped into the Wady. The Wady was as dry as it is to-day and the great inland sea was always fresh. If there had been any serious infiltration from the ancient Lake Mœris into the Wady Rayan, there would have resulted a lake which could not have escaped the notice of the numerous travellers who visited the lake. No mention was ever made of such a lake. This body of water moreover would have been inhabited by fresh-water animals whose remains would have strewn its shores. No such remains are to be seen to-day. If therefore the ancient Lake Mœris with a head of 63 metres on to the Wady Rayan could not leak into the Wady, it is not likely that the Wady Rayan reservoir with a head of from 27 to 29 metres on the Garak side of the Fayoum will leak into that part of the area covered by the ancient lake. Any leakage into the Lake Kurun side is never contemplated by anybody, since many kilometres of compact limestone lie between the Wady Rayan and it, while about one or two kilometres of the same limestone lie between the Wady Rayan and the Garak depression.

34. Lake Albert reservoir project and project for training the Albert Nile and the Zeraf River.

—If we wish not only to irrigate the whole of Egypt, but to include the Sudan in the sphere of operations we must regulate the supply issuing from Lake Albert Nyanza and ensure its passage through the great swamp regions. To my mind no work in the Sudd regions will be of any substantial value unless the Albert reservoir dam is first built. Tabulating the information collected in the gauges and discharge tables we may state that the discharge of the Albert Nile in cubic metres per second between the 15th January and 15th May was as follows in:—

In 1861 the discharge at Gondokoro was as low as 500 cubic metres per second. It will be seen that, in spite of the great waste, there is an increase at the northern end of the Sudd region even under present conditions when the discharge at the south end is increased in the interval between the 15th January and 15th May. The water which enters the White Nile during these months represents the summer contingent of the White Nile to the Nile in Egypt.

Now though an increase in April at the south end of the Sudd region is felt at the north end, no such increase is felt in September and October, and the reason has been given in Chapters II and III.

In April the Sobat river is discharging practically nothing, and the whole supply available in the Albert Nile can pass down the White Nile past the Sobat mouth. In September the Sobat river may be discharging 750 to 1,000 cubic metres per second, and as the White Nile cannot discharge the combined waters of the two rivers, the water of the Albert Nile is headed up and accumulated in the lowlands between Lake No and the Sobat mouth. This is greatly to the advantage of Egypt, for it is the discharge at the head of the White Nile between January 15 and May 15 which decides the White Nile contingent to the summer supply of the Nile in Egypt, and the greater the quantity of water above the head of the White Nile, in the absence of a regulator or barrage at Wadelai or Dufilé, the better the summer supply of Egypt. After the abnormally high flood of 1878, when Gordon was up the Nile, so great was the accumulation that the discharge at Assuân never fell below 1500 cubic metres per second in the summer of 1879. The Barrage was not regulated upon and yet all the Lower Egypt canals were full of water, and the cotton crop of Egypt for that year was quite abnormal for the seventies of the last century.

Now an expenditure of between £400,000 and £1,000,000, say £800,000, could secure a regulator for Lake Albert at any point between the outlet and Dufilé. Such a regulator would insure 1200 cubic metres per second every year to the Albert Nile at Gondokoro between the 15th January and the 15th May. With this supply insured, the training works in the Sudd region would soon begin to affect the discharge at the north end of this region where the White Nile begins.

The way in which this work of training should be carried out has been admirably laid down in page 174 of Sir William Garstin’s Report. “Alter the flood conditions of the Albert Nile (Bahr-el-Gebel) as little as possible, let the excess flood water escape on both sides, but keep the summer supply in its channel.” This is, to my view, the soundest statement from an engineering point of view in the whole report. Hitherto we have always assumed a vast expenditure for keeping the flood supply in one channel; but with our attention devoted to the summer channel, we should have before us all the advantages of summer training works without any fear of inundations. The very wildness of the regions would be in our favour. To be able to train a river in summer without any nervousness about floods, is given to few engineers. I had never thought that any good thing could come out of the sudd region, but looked at from this point of view, we can, even in this inhospitable waste of waters, confirm Shakespeare’s saying that “there is some soul of goodness in things evil.”

The Lake Albert reservoir could easily insure 1200 cubic metres per second every year between the 15th January and the 15th May. To pass through the Sudd regions as much as possible of the 1200 cubic metres per second received at Gondokoro, the following works would be necessary:—

The first work to be done would be the removal of sudd block No. 15 which for 37 kilometres south of Hillet-el-Nuer has turned the Albert Nile out of its course. The importance of this is strongly insisted on by Sir William Garstin on page 55 of Appendix VI of his Report, which is the very last thing he wrote. With this block removed, the Albert Nile would be given a good opportunity of working out its own salvation.

The next point is one to which I attach the greatest importance. Indeed, I look upon it as the key of the whole region. The Albert Nile enters the south-east corner of Lake No and almost immediately afterwards leaves its east corner. Now Lake No is the final evaporating basin of the Bahr-el-Gazelle, probably the most unsatisfactory river in the whole world; and it is open to doubt whether in a year of deficient rainfall on its own catchment basin and a year of good supply down the Albert Nile, it does not evaporate a considerable quantity of the water of the Albert Nile. If the discharge of the Albert Nile north of Hillet Nuer was brought to 450 cubic metres per second in April, this lake might waste much of it. Such being the case, a cut of a maximum length of some 5 kilometres should be dredged south of the south-east corner of the lake, and the waters of the Albert Nile separated from those of the Gazelle. A cheap wooden lock and regulator would allow boats to pass, and take in any water from the Gazelle river when it had it to give. The maximum water we have to deal with is zero in summer and 40 cubic metres per second in flood. If the Albert Nile were separated from the Gazelle, it might be found that the waters of the Gazelle river found their way into the Albert Nile through that mysterious and unsatisfactory river the Lolle. Some use might even be found for it. I have, since I wrote this, been informed by Capt. Lyons that Marno in 1881 proposed this cut round the south-east corner of Lake No. (Pet. Mitheil, 1881, page 425 and plate 20).

The Albert Nile to-day at Hillet-el-Nuer is capable of carrying 450 cubic metres per second. North of Hillet-el-Nuer, the Albert Nile has a mean width of 75 metres, but owing to want of training and papyrus swamps, the discharge dwindles down to 320 cubic metres per second. The Albert Nile has in this reach a very good section indeed and by beginning with the west bank and dredging round corners and closing spills by dredged earth, an improvement of section and slope, backed up by a permanent discharge of 450 cubic metres per second at Hillet-el-Nuer, might speedily result in this very discharge of 450 cubic metres per second being obtained south of Lake No. We are here in such swamped land that percolation would be practically zero. There are only 200 kilometres of channel with all its curves, or 160 kilometres of trained channel to work at, in which 75 per cent of the work is to hand; and we have to confine our attention to the low supply without worrying over the flood supplies.

Now the Bahr-el-Zeraf could be made capable of carrying 150 cubic metres per second by an improvement of the inlet at Baker’s channel and by dredging north of Ghaba Shambe, as will be seen by examining the measured discharges of the Bahr. With the Bahr-el-Zeraf carrying 150 cubic metres per second and the Albert Nile carrying 450 cubic metres per second, we should have ensured 600 cubic metres per second. As time went on, improvements in the channels would make themselves felt, as they are even doing to-day, and we might even have 700 and 800 cubic metres per second at the head of the White Nile below the Sobat mouth. We must always remember that it is assumed that the lake Albert reservoir dam has been constructed and a supply of 1,200 cubic metres per second assured at Gondokoro from January to May. If the Sobat river were capable of regulation downstream of the Pibor marshes, we might get an increase from that direction as well. The possibility of a good site for a reservoir upstream of Nasser is very great. The Nasser gauge rises and falls so slowly that there must be very great natural accumulations of water upstream of it, which might be improved. The question of quality needs study.

The discharge of 435 cubic metres per second, which the rivers in their present state, aided by the Lake Albert reservoir, could deliver into the White Nile, would thus see itself gradually increased to 500, 550, and 600 cubic metres per second and even more, and would enter the White Nile past the Sobat mouth in February, March, April and May. During these months the whole of the supply would pass down the White Nile without throwing backwater on the rivers. Later on, when the Sobat river came down in flood and filled up the channel of the White Nile, the Albert Nile would have its waters reduced at the Albert regulator, and the waters of Lake Albert would again be stored for the day when the Sobat floods had fallen and the White Nile channel was free and ready to take in the waters of the Albert Nile.

While I had contemplated the training of the flood waters of the Albert Nile through the Sudd regions I had estimated the cost at £100,000 per annum, for 25 years, or at £2,500,000. Now, however, that Sir William Garstin has shown how we need only attend to the summer supplies, the sum may be reduced to half the former figure or to £1,200,000, and be more than ample. With £2,000,000 devoted to the Lake Albert reservoir and the training works in the Sudd region, the summer supply of Egypt and the northern Sudan would be put on a firm base.

Dr. Schweinfurth, the eminent African traveller and savant, was the first to call the attention of the Egyptian Government to the necessity of closing the spills from the White Nile to the north of Gondokoro, and so beginning the training of the river. He very rightly said:—“Many years would elapse before the desired result would be obtained by the strengthening of the banks, but the works would be increasingly felt every year in Egypt as the works progressed.”

35. Flood protection for Egypt.

—In paragraph 30 it was stated that the floods in the Delta or in Lower Egypt can rise to a height of from 2 to 3¹⁄₂ metres above the level of the country. Such floods are really dangerous and means should be found for moderating them.

The Wady Rayan reservoir, when converted into the modern Lake Mœris and acting as a reservoir, will have one great advantage, it will be able to lower a high flood 30 centimetres for 50 days. This will give relief to the Nile, a relief which will be much appreciated by the whole country from Beni Suef to the sea, and especially by Cairo.

I have already stated that the Damietta branch is especially dangerous and unfit to act as an escape under existing conditions. That branch could be regulated on at its head and treated like a canal. Thanks to Sir Hanbury Brown’s initiative the Barrages can be regulated on in flood as well as in summer, and by lowering the supply in the Damietta branch and turning the surplus down the Rosetta branch, the latter would become the flood escape of the Nile. It might be trained as Mr. Eads suggested that rivers should be treated.

Mr. Eads’ argument is very clear. He insists that rivers eat away their banks in places, not owing to the direct action of the water but by the alterations in the velocity of the current. When the river water is charged with sediment to its full carrying capacity it cannot take up more unless the rate of current be increased. If the channel be nearly uniform the river water cannot eat away any of its banks. If, however, the channel is varying, the silt deposits in the wide sections, and the water, free of some of its sediment, is ready to eat more. It is this alternate dropping silt and eating away of earth which does the harm. To treat the Rosetta branch according to Mr. Eads it would be necessary to fix the top width to be worked to, at say 550 metres. The river could be brought to this uniform top width by building light inexpensive permeable spurs on the sandy shoals. The land between the spurs would become cultivated and such river training would pay the Government, which taxes all cultivated land; it would even pay handsomely for any company to undertake the work once the rule about foreshores was understood. The Government would, however, always succeed; when it could not sell, it could always tax. Such training would permanently lower the flood.

In addition to the above it would be necessary to complete the system of spurs begun in 1884, and to throw back the banks as already recommended. It has been estimated that the completion of this work would cost £900,000 for spurs and banks, while the training works would pay for themselves in addition to greatly improving the channel and lowering the level of the flood.

It may be humiliating to make the confession, but from B.C. 2,200 to the Arab invasion of Egypt in A.D. 640, while Lake Mœris performed its allotted task and the Nile possessed training works such as those we can see to-day in Nubia, Egypt was better protected from inundation, and the Nile better trained, than it is to-day. And yet we have many advantages which no Pharaoh possessed. By the aid of telegraphy we have knowledge of a coming flood a full fifteen days before it arrives in the Delta; the Khartoum gauge allows us to anticipate its very height. Meteorology is aiding us still further. In a paper I read at the Chicago International Exhibition I stated that years of heavy rainfall in India are years of high flood in Egypt, while years of poor rainfall in India are years of low flood in Egypt. Sir John Eliot, the Director General of the Meteorological Department of India, corrected this statement. He said that though this was not true of the Bengal monsoon, it was true of the Bombay monsoon. Years of heavy rainfall in Gujerat and Bombay are years of high flood on the Nile, and vice versà. As the rain falls in Bombay a month earlier than the Nile flood reaches Cairo, we have information of a high flood a month before it arrives, if we receive telegraphic information from Bombay.

36. Complete project for water storage and flood protection for Egypt.

—The complete project for water storage and flood protection for Egypt as proposed by me, contemplates the following works:—

To these have to be added the approximate estimates of the proposed works on the Upper Nile:—

The total expenditure amounts to £6,000,000 spread over 12 years.

The great advantage of undertaking all those works together may be thus summarized. The increased supply from the Assuân reservoir will be felt in Egypt after a period of two years. Five years later the waters of the Wady Rayan will be added to those of the Assuân reservoir, and it will be possible to increase the cotton crop of Egypt from 6 million to 10 million cwt. It will be possible to allow the Sudan to thoroughly develop its agricultural resources, and with the aid of the 25,000 horse power as a minimum which the 6th cataract near Khartoum can supply, to utilise for its own consumption the waters which can be stored at that cataract; and, in addition to those, the available supplies from Lake Tsana provided that that lake is furnished with an outlet tunnel.

While all this life and activity will be developing themselves in Egypt and the Sudan, the effects of the regulator of Lake Albert and the training of the Albert Nile in the Sudd regions will be gradually asserting themselves; and, if the works are being steadily and perseveringly carried out, it is well within the range of possibility that before 10 or 12 years will have passed, the additional supplies from the upper waters of the White Nile will have become so ample, that it will be possible to dispense with the Wady Rayan as a reservoir. When this will have happened, the Wady Rayan with its canal will become the true flood escape of Egypt, like the ancient Lake Mœris, and will, with the Rosetta branch, afford complete protection to Egypt against the dangers of a high flood. Egypt, in the fullest meaning of the term, will be enjoying perennial irrigation and flood protection.

In my book on the “Assuân Reservoir and Lake Mœris” I had recommended a more extended programme, but the reading of Sir William Garstin’s Report has convinced me that useful as the Lake Victoria reservoir dam may be, its postponement as recommended by Sir William is sound, until all the other works have been executed. The really essential work is the Lake Albert reservoir dam, of which the study might indeed be commenced immediately. Sir William’s proposal to train the summer supply of the Albert Nile and allow the overflow of the floods to find its way through the Sudd region is so sound and convincing that the necessary training works in the Sudd region are greatly reduced. With these reductions the estimated cost of the project for water storage and flood protection for Egypt is reduced from £8,200,000 to £6,000,000. I have left the Wady Rayan estimate as it was in my original programme. Mr. Webb’s criticism of the project was based on facts which are outside the project. He supposed that the Wady Rayan was to remain a reservoir for all time and that it was not to be aided by the works on the Upper Nile. Now the project I proposed and which I propose now, presupposes that the Wady Rayan will be a temporary reservoir and final flood escape for the Nile, and that it will be aided in years of very deficient flood by the gradual improvement of the Upper Nile owing to the works undertaken there.

37. Sir William Garstin’s programme for water storage and flood control.

—In the first appendix to his Report on the Upper Nile, Sir William Garstin, G.C.M.G., Adviser to the Ministry of Public Works, has drawn up a programme of works for water storage and flood control in the Nile valley. He approves of the raising of the Assuân dam for £500,000, and the conversion of the Rosetta branch of the Nile into a flood escape for £900,000. He then conditionally approves of a proposal suggested by Mr. J. S. Beresford, C.I.E., for making a straight cut from Bor on the Albert Nile to the mouth of the Sobat river at the tail of the Albert Nile. The line would be 340 kilometres in length and is estimated to cost £5,500,000, and carry 600 cubic metres per second in summer. In case of the line being found impracticable when it was surveyed and levelled, Sir William proposed abandoning the Albert Nile and thoroughly widening and deepening the Zeraf river for £3,400,000.

As a criticism of the Bor cut project I cannot write anything more convincing from my point of view than a letter written by me and published by “The Engineer” in October of this year.

“In your issue of the 16th September Sir Hanbury Brown has reviewed the scheme suggested by Mr. J. S. Beresford, C.I.E., and conditionally approved by Sir William Garstin, for diverting the waters of the Albert Nile (known as the Bahr-el-Gebel) from Bor to the mouth of the Sobat river, on a length 340 kilometres, and sending them down a canal capable of carrying 600 cubic metres per second, at an estimated cost of £5,500,000. In his review Sir Hanbury puts his finger on the weak point in the project, viz., the difficulty and loss of water entailed at the crossing of the Albert Nile just upstream of the Sobat mouth. The difficulty will be got over, as Sir Hanbury himself suggests, by an earthen embankment provided with a regulator. The loss of water cannot be got over.

“I have taken the following figures from Sir William’s report and from the gauge records of the Public Works Ministry:—

Discharge in Cubic Metres per Second During February, March,
and April of the Albert Nile.

“Now, in a year like 1901 or 1902, with 600 cubic metres per second passing Gondokoro, the diversion canal might be allowed to take in 500 cubic metres per second, leaving 100 cubic metres per second for the Albert Nile, Atem river, and all the Nuer, Dinka, and Shillook country between Gondokoro and the Sobat mouth. An allowance of 100 cubic metres per second would not be liberal, and would probably result in the water becoming stagnant and very impure; but we shall leave that alone. Starting with 500 cubic metres per second of clear water the high level diversion canal would never lose less than 50 cubic metres per second through percolation and evaporation before it reached the Sobat mouth. Many authorities would put the loss at 40 per cent., but we shall say 10 per cent.

“We should then have 450 cubic metres per second entering the White Nile at its head, just at the end of the Albert Nile and at the mouth of the Sobat river. At this point, however, under normal conditions the Albert Nile would have been discharging 300 cubic metres per second. This supply, after the opening of the diversion canal, would have failed utterly, as the waters of the Albert Nile would have been diverted down the diversion canal. Whatever water there was in the Albert Nile would, moreover, have been at so low a level that it could not have flowed down the White Nile together with the high level water of the diversion canal. We should therefore have had in a year like 1901 and 1902 a net gain of 450 less 300 cubic metres per second, or 150 cubic metres per second at the head of the White Nile. By the time this extra water reached Assuân it would have become 100 cubic metres per second.

“If this project, or any other project of any kind, is ever to be carried out on the upper waters of the White Nile, the very first thing to be done will be to construct a weir or barrage at the outlet of the Albert Lake, at Wadelai, or lower down at Dufile. I should say, judging from the map and the cross section, that Wadelai itself would be an excellent site for a weir. I have advocated this project in season and out of season these ten years, and now that actual discharges and figures are before me I am more than ever convinced that I was no untrue prophet when I wrote in my book on “The Assuân Reservoir Dam and After” that “the point where Lake Albert ends and the Albert Nile begins to have a rapid and contracted stream will be the site of the future great regulator or barrage of the upper waters of the Nile. This work will be here or at Dufile.” Such a work would cost anything between £400,000 and £1,000,000.

“If such a work were carried out it would be possible to insure every year a discharge of between 1,000 and 1,500 cubic metres per second at Gondokoro from the 15th of January to the 15th of May, i.e., during the months which determine the summer water supply of the White Nile for Egypt. Such a quantity of water would insure 435 cubic metres per second at the head of the White Nile, as it has done this year, even under existing conditions; while with training and dredger work in the Albert Nile and Bahr Zeraf between Gondokoro and Lake No, it might be increased to 600 cubic metres per second, and even more. The way in which this work of training should be carried out has been admirably laid down on page 174 of Sir William’s Report”.

There are moreover other reasons I think for condemning the excavation of a straight cut 340 kilometres in length across the eastern corner of the Sudd regions. The reasons are to be found in Sir William Garstin’s Report itself. One of the most interesting features of this report is the number of actually measured discharges at different sites. Of all these sites Gondokoro, the southern key of the Sudd region, is the most interesting.

It is very evident from an examination of Sections Nos. 18, 19, 26 and 27, Plate VIII of Sir William’s Report, that the Albert Nile at Gondokoro scours out its bed very severely after a high flood like that of 1903. The width of the section is about 230 metres with vertical sides, and yet while a gauge of ·50 metres on the 1st April 1903 (after the low year of 1902) gave a section of 615 square metres; on the 9th September 1903 (after a good year), the section was 1,347 square metres for a gauge of 2.33 metres. In other words, a rise of 1.83 metres gave an increased section of 732 square metres; while, if the bed had not scoured, it would have been 421 square metres. We have here an increase of 311 square metres, or more than 1 metre of scour. All this happened in 5 months, and proves that the clear water of the lakes, when in volume, has a fine cutting edge.

In footnote (2) of page 116 of his Report, Sir William Garstin says that in the parts of the river where the sudd has been cleared there are indications that a scour of the bed has set in. Again, on page 55 of the appendix, he says that the removal of the sudds has caused the levels of the shallow lakes to fall. All this proves that if the spills and escapes from the Albert Nile were closed with ambatch, as proposed by Sir William Garstin on page 175, and a few dredgers put into the Albert Nile and the Zeraf river the expenditure of a sum of money very moderate indeed compared with £5,500,000 would in all probability result in the two rivers being so widened and deepened that they could carry the full summer supply of the lakes, and so there would be a resulting economy of over £4,500,000 in the new channel from Bor which, when it began working, might introduce on an aggravated scale all the difficulties of to-day in the Albert Nile.

A good description of certain spills is given on page 112 of the Report, a good idea of scour in Plate XXIX., opposite page 110, while on page 181 Sir William Garstin makes the remark that the experience of American engineers has taught us that though in theory it may be possible to shorten or straighten a great river, in actual practice it is accompanied with almost insurmountable difficulties. If the new channel were dug and set working, in a few years it might be as crooked as the Albert Nile itself, unless it were protected with stone along its entire length.

In my project for deepening and widening both the Albert Nile and the Zeraf river to enable them to carry 600 cubic metres per second, I think I have given very solid reasons against abandoning the Albert Nile and sending the whole supply down the Zeraf river at a cost of £3,400,000.

I cannot but think that Sir William Garstin’s recent objections to the Albert Nile are founded on an oversight. He has, apparently, not kept the flood discharges of the Albert Nile at Gondokoro and above the Sobat mouth sufficiently apart from those of low supply. It is the Sobat flood, combined with the poor carrying capacity of the White Nile, which is the disturbing factor, and not any inherent viciousness in the Albert Nile itself. The Albert Nile has a good section, and, if it were trained in conjunction with the Zeraf river, would, I feel confident, discharge all the water required with a very moderate expenditure of money. This, I always understood, was Sir William’s own opinion. In such projects it is wise to remember Horace’s saying, “Naturam furcâ expellas tamenusque recurrit.”

38. Project for converting the basin irrigated lands of Upper Egypt into perennially irrigated lands.

—No consideration of this question would be complete without first examining into the changes which would be made in the regimen of the Nile flood by the contemplated conversion of basin irrigation into perennial irrigation. This question was examined very thoroughly by me in 1892 and 1893 and I give here my arguments for not anticipating any serious difficulties.

We have to consider the effect of the introduction of perennial irrigation on the regime of the Nile. The perfection of the perennial irrigation of the Delta north of Cairo will in no way affect the Nile in flood. The canals will continue to run as they do at present, and the question of conversion in Lower Egypt is therefore quite independent of the subject of flood protection. In Upper Egypt, however, we have 1,460,000 acres of basin irrigation; and as each acre receives in a low flood 80 cubic metres of water per day, in an ordinary flood 130 cubic metres per day, and in an extraordinary flood 170 cubic metres per day, while the demands of perennial irrigation are only 25 cubic metres per acre per day, it will readily be understood that we are dealing with a quantity of water which demands the greatest attention.

To foretell with exactitude the anticipated changes in the regime of the Nile, it is necessary to know first the daily gauges of the Nile at Assuân and Cairo for a period of at least twenty years, and the discharges corresponding to these gauges. The difference between these discharges represents the consumption of water. We have next to determine the amount of water which passes into the canals, the amount utilised in filling up the trough of the Nile and covering the berms, and the amounts evaporated and absorbed. We know that the last three items are constant while the canal discharges are variable and depend on the system of irrigation and, if our data are correct, we can tell with moderate certainty what changes in the level of the Nile will follow certain changes in the system of irrigation.

Tables 65 and 67 of Appendix L contain the Assuân and Cairo gauges for typical years in a period of twenty years from 1873 to 1892, and the mean gauges of these twenty years. Finding it impossible to understand the Nile without first referring every gauge to some uniform standard, I have had to choose the line of reference. The mean high water level and the mean low water level are both available. In Egypt the mean high water level varies very considerably whether we take it in August and the early part of September when the basin canals are running full supply, or in the latter half of September when the canals are running only half supply, or in October when the basins are discharging back into the Nile. Early and quick rising floods have a different series of levels from slow and late floods; while again the recent works carried out in Upper Egypt by Col. Ross have so increased the discharging capacity of the canals that the flood gauges have been appreciably affected. All this points to the conclusion that the mean high flood is no satisfactory standard. The mean low flood on the other hand is much less liable to change and is very fairly constant from year to year. High floods are certainly followed by scouring out of the bed, and low floods by a silting up of the channel, but the changes are very moderate compared to those in high flood. I have chosen the mean low water level as the line of reference, and referred all gauges to it. From the mean of twenty years’ observations, this level at Assuân is R. L. 85 metres. By observations along the Nile generally, and by calculations at Cairo, I have fixed it at all important places north of Assuân. Table 46 of Appendix K gives the Reduced Levels at different places, while it is also drawn on the longitudinal section of the Nile in Plate XII. It was on this system that the ancient Egyptian engineers worked the Nile. They however chose the mean high water level during the early part of the flood as their standard of reference and consequently made the so called cubits in the flood reaches of the Cairo gauge half cubits. This means a discharge of 1600 cubic metres per second and fairly represents the discharges of the basin canals in flood. When it is considered that the level of the Nile valley is raised by about 10 centimetres per 100 years it will be seen that the old Cairo gauge, which was a living record 1500 years ago, is to-day a meaningless anachronism. It has also to be compared with the Assuân gauge which was erected in Ismail Pasha’s time with an arbitrary zero some 90 centimetres below mean low water level, and which may be reading 17 cubits while Cairo may be recording 25 cubits. The Cairo gauges in winter and summer are no records of discharge as the afflux from the Barrage affects them. To find the discharge at Cairo during these months, I have added those of the Rosetta and Damietta branches and the Delta canals upstream of the Barrage. When the Nile falls below mean low water level, the gauges are recorded as minus quantities.

Discharge sites having been chosen for the Assuân, Assiout and Cairo gauges on the Nile, a continuous series of surface velocity observations, cross sections and slope measurements were made during 1892 and 1893 and the resulting discharges recorded. Curves of discharge have been drawn and referred to the gauges of twenty years and modified until finally a curve has been found which will suit any year whether it is a maximum or a minimum. In connection with this subject, it must be remembered that the Nile bed is raised by silt during low floods and scoured out during high floods and that consequently August and September discharges vary considerably at times from October and November discharges for the same gauge. In addition to this, it must also be borne in mind that the slope of water surface and that consequently the discharge of a flood during the rise is far greater than during the fall for the same gauge reading. Indeed the Nile often discharges more when it is 30 centimetres below its maximum and rising fast than when it has reached its maximum and begun to fall. It is owing to this fact that we often see the discrepancy of the Assuân gauge reaching its maximum a day before Halfa which is 350 kilometres higher up the river. The discharge depends on gauge and slope, and the gauge only records one element. Keeping these facts in my mind, I saw that it was of no use recording the gauges to two places of decimals and covering paper with useless figures, and consequently I have chosen the higher unit for a rising gauge and the lower for a falling gauge when I have been dealing with discharges.

Flood discharges have been taken of all the canals in Upper Egypt through 1892 and 1893 and have been recorded in >Tables 48 and 49 of Appendix K. From these tables, Table 47 has been compiled which gives rough approximate discharges of the canals corresponding to the Assuân gauges in the first half of the flood.

To obtain information about the trough of the Nile, the area exposed to evaporation and the area of absorption, a longitudinal section of the Nile from Assuân to Cairo has been levelled, and cross sections taken at every 3 kilometres. The kilometrage on Plate XII counts from the Assuân gauge and is measured down the centre of discharge of the flood, since it is with flood discharges that we are principally dealing. As the Nile winds about considerably and is often broken into numerous channels, the areas of the cross sections vary very appreciably according as they are taken at right angles to the centre line of discharge or of the deep channel of the river. The former gives the more reliable results. I have taken 8 millimetres per day as the evaporation during flood in Upper Egypt. The absorption has been calculated from the water consumption during the floods of 1892 and 1893, and found to be about 300 cubic metres per second between Assuân and Assiout, where there is practically no perennial irrigation. Between Assiout and Cairo, where there is a considerable length of perennial irrigation on one bank and limestone rock on the other, the absorption is about 100 cubic metres per second.

When perennial irrigation has once established itself in Upper Egypt, we may assume that the absorption during flood will be halved in quantity for the reasons given above, and become 150 cubic metres per second between Assuân and Assiout, and 50 cubic metres per second between Assiout and Cairo. The amount of water expended in irrigation will be about 700 cubic metres per second. The evaporation during flood will be approximately 120 cubic metres per second. The quantity of water needed to fill the trough of the Nile will depend on the gauges and may be calculated from table 43 of Appendix K. The last item will be the only variable one and the others may be approximately tabulated as follows:—

Expenditure of water in flood in cubic metres per second: