state of things. . . . The instant the activities of man are suggested and his methods studied, there is not a phase in the rivers' behaviour that is not capable of a sound and reasonable explanation; but without this the caprices of the river[s] are inexplicable.'

Another officer of the Indian Public Works Department, Major Mackintosh, who was also employed for several years in Mesopotamia, gives in his report on the Hammar Lake area an admirable study of some of the methods employed in this 'very unscientific irrigation' and of the disastrous consequences. His facts are taken from close personal observation on the spot and extensive inquiries amongst the Arabs of the Muntafik confederacy. It is noteworthy that, though nearly a hundred years have passed since the Hammar Lake was first formed, the members of those tribes who were then flooded out still mark exactly where ran the boundaries of their now submerged tribal lands. To these they still refer as their dirah (tribal home), and they lead a semiamphibious life in as close proximity to them as possible, all ready to pounce down on their ancient heritages (and as much more as they can get), should the act of God or man cause the waters to recede.

'In 1830,' writes Major Mackintosh, the Euphrates river flowed as a navigable stream from Nasiriyah to Qurnah. In the low-water season the contrary current at high tide was felt to just below Nasiriyah, and the tide swelled the river as far as Darraji, 40 miles further up-stream. In those days the river was doubtless of the character of the Shatt-al-'Arab of to-day; broad, deep, and slow-moving. Not much silt would be found in the water, as the supplies, then mainly coming down the Hillah branch, would drop their silt in the marshes between Rumaithah, Erech, and Nasiriyah; the balance, or Hindiyah supplies, being similarly strained clear in the Shinafiyah lake and the marshes between Shinafiyah and Samawah. Irrigation would be easy by cuts made anywhere through the banks into the low-lying basins on either side. As the Turks had apparently no settled policy with regard to the rivers, Arabs were allowed to encroach on its (sic) flood bed, fortifying these increases to their holdings with bunds [earthen banks] and by the planting of willows. Subsequent floods could therefore only relieve themselves by scouring out and enlarging the existing irrigation channels,

which had a relatively steep slope, so raising the water-level in the basins on either side as to flood valuable land.'

'The inundation difficulty was settled for the time by the Sa'dun, Nasir Pasha, builder of Nasiriyah and Wali of Basrah under the Turks, who built a bund (circa 1870) along the right bank of the river from Qurnah to Suq. He wisely confined his building to the right bank; growing on the left bank only rice, which, planted as the floods went down, throve on lands soaked by their overspill. His bund was successful in protecting a large area of land sloping down to a salty lake in the centre which evaporated surplus drainage. On this land barley, wheat, and rice were grown for the ten years or so for which the bund lasted. . . . While Nasir Pasha's bund finally failed, partly through Turkish neglect of maintenance after he had been exiled to Constantinople, and partly by excessive reduction of the Euphrates bed by "reclamation," there can be no doubt that a contributing factor was flood-spill from the Tigris also caused by "reclamation" at 'Amarah and the Narrows.

...

'In Chesney's time the Butairah was only a small canal, and that much of the flood now pouring out into the marshes then passed down the Tigris is obvious from his description of the river at Ezra's Tomb as 300 to 400 yards wide and 6 to 10 fathoms deep now 400 feet wide and 12 feet deep. Arab opinion indeed insists not only that the Butairah was responsible for the failure of Nasir Pasha's bund, but that its increased draw has dried up several old canal systems between Kut and 'Amarah, and materially reduced supplies down the Hai. Whatever the cause, the effect is to-day a sheet of open water some 60 miles long by 10 to 15 miles wide, fringed on either side by miles of swamps expanding and contracting with the seasons and the intensity of the year's floods.

'The bank of the old Euphrates can still be traced across the lake in a series of islands, but almost the whole of the land cultivated in Nasir Pasha's time is now under water. Portions of his bund remain, a short length still in use protecting the date gardens at Medinah and other lengths near Bani Said. Running parallel to the latter can be traced the old bed of the Euphrates, 450 feet wide, now completely "reclaimed" by the victorious Arab with his spade and willows, a great river destroyed for the sake of a few acres of land.'

The practice of reclamation' which has called forth the irony of Major Mackintosh is, in reality, only

unregulated riparian cultivation, perfectly legitimate up to a certain point, and not difficult, given a little supervision, to render wholly innocuous. Let the reader imagine himself a squatter on the banks of the Tigris or Euphrates and dependent for his livelihood on the capricious bounties of the great river. We will suppose him to be animated by the true Semitic desire for riches and virtuously resolved to create wealth for himself, instead of following a short cut and plundering his neighbour. We will further assume that he is in no fear of dispossession and that it is worth his while to improve. He will naturally decide to plant a date garden. Without a constant water supply the young suckers, which he will set out in quincunx formation, will perish miserably without taking root, and his labour will be in vain. The river is at his very door, but it is only for a few weeks in the late spring that the water lips his holding and irrigation by flow is possible. For the rest of the year some means of lifting the water to the land must be found. The squatter may be content with the primeval karad (bucket lift), worked by an ox or horse or mule, under his own immediate and unremitting superintendence. But if he is enterprising, as most Arabs are, he will mortgage his credit to the hilt and instal a pump worked by an oil-engine. In either case his procedure, so far as its effect on the river is concerned, is the same. He sinks a shaft somewhere on the riverward side of the flood-protection bank that guards his holding, and from it across the undulating silt and sand of the river bed he leads a channel along which the waters of the distant stream shall flow to his well-shaft. say 'distant' advisedly; first, because the undertaking must be carried out in the low-water season, when the river bed is a world too wide for the shrunk stream; and second, because the squatter, if he has any choice at all, will be foolish if he does not select for the scene of his paradise a convex bend, where the risk of loss by erosion is small and the chance of accretion proportionately great. For on a bend-and both rivers consist mainly of bends-the deep-water channel hugs the concave, shore. The spoil thrown up in connecting the shaft with the stream is naturally left in situ, where it forms a double rampart of varying, but often of considerable, Vol. 237,-No. 470,

I

F

height, athwart the trend of the river. The flood comes, and its waters, almost solid with suspended silt, run a bright copper red. The rampart is in places washed away, but where it stands it forms an efficacious silttrap. When the flood recedes, the level of the bed upstream of it will often be found to have risen by several feet. Each year the process is repeated, till in time the squatter pegs out and fortifies his acquisition with a row of willows, and so da capo. This constant interferencefor the squatter we have imagined is only one of many -interferes with the rhythm of the stream. What it gives to one bank it takes from the other. The opposite bank, ex hypothesi, is already concave; and on a concave bend the flood-protection bank, as is natural in a country where the population is sparse and distances are great, runs very near the edge of the river bed. With a little erosion it will be undercut, and the momentum of the stream, travelling in the direction of the last reach, will deepen and widen the gap with great celerity. A new door to calamity is opened on the opposite bank. But what of that? It is the affair of the people on that bank. Let them see to it! Our planter, with true oriental indifference, will quote piously from the Qur'an, Allahu khairu 'l razikin (God is the best of providers), and return to the watering of his plantation.

The question whether Mesopotamia from Hit and Samarrah southwards to the sea is correctly described as delta or not, is, therefore, no mere academic point of terminology. In so far as the rivers are in a true deltaic condition, any attempt to produce other conditions will have the forces of nature against it and will be prohibitively difficult and expensive. If, however, Major Walton and his supporters are right, Mesopotamian conditions, although they resemble deltaic conditions, are yet the outcome, not of natural causes, but of man's interference with nature acting through long ages and with the enormous leverage afforded by the difference in the rivers between low water and high flood. If this be so, the apparently deltaic conditions, not being due to natural causes, will disappear as soon as the interference ceases, or is scientifically regulated. In the one case, the scale is weighted against the engineer; in the other, in his favour.

Sir William Willcocks has rightly laid stress on the vital importance to Mesopotamia of protection against flood. But it is hard for any one without personal experience of the country to realise not only the importance but the immense magnitude of the task under present conditions. Some idea of it may be formed from the fact that, in the season 1918-19, more than two and a half million cubic metres of earth-work were put into flood-protection banks under Government supervision. Nevertheless, the floods of 1919 destroyed well over 100,000 acres of the spring harvest. To the toil of construction and the burden of vigilance during the flood season must be added the paralysing effect of uncertainty. A frail earthen bank is in many places all that stands between the harvest and ruin. On this bank, which may be many miles remote from the lands that it protects, wind and wave have full play. Rats burrow there and watchers sleep. Malice has to be reckoned with, as well as negligence. For by the furtive removal of a few clods an enemy can wreak irreparable mischief without serious risk of detection. Perhaps those who decry the Arab for a bad husbandman do not sufficiently take into account the conditions under which he labours.

One more quotation from the writings of Major Wilson, to bring home the difficulty, and we can then proceed to consider the remedy which that officer was, I believe, the first to suggest.

'Provided,' he writes, the Tigris channel were able to carry its floods at all places between Baghdad and Qurnah, the maintenance of the bunds along its margins would be greatly simplified, but unfortunately its flood capacity decreases as one descends from Baghdad. At 'Amarah its flood capacity is approximately 15 per cent. of that at Baghdad, whilst at Qal'at Salih it is only 3 per cent. This means that during a high flood 97 per cent. of the water passing Baghdad has to be turned out of the river, before Qal'at Salih is reached. Taking the high flood discharge at Baghdad as 250,000 cusecs (i.e. cubic feet of water per second), then the escapage amounts to 242,000 cusecs. In the provision of escapage from canals and rivers one expects and generally finds suitable works for controlling it, but in the case of the Tigris floods not 5 per cent. is under the effective control of suitable works; the remaining 99.5 per cent. is