Bactria, to the east of the Zagros mountains and the Iranian plateau, is connected to the Syro-Mesopotamian world through a continuous thread of ancient exchanges, and thus it also must be mentioned in this chapter. Bactria was a land of plenty and fertile val­leys as noted by Strabo:

“Man has only to take the trouble to irrigate, and all crops grown abundantly with the excep – 53

tion of the olive tree”.

In Chapter 1 we mentioned the appearance of the Bactria-Margiana civilization, or the Oxus civilization, for which the mastery of irrigation is a fundamental pillar. It is doubtful that Bactria ever really fell under the Assyrian yoke, as is suggested in certain ancient sources. To the contrary, it was clearly part of the Achaemenid Persian Empire, and therefore belongs on the list of conquests of Alexander – but not without combat. Indeed, it was at Bactria that Alexander married Roxanne.

Successive hydraulic developments in the region became more and more general­ized from the IVth millennium BC. In Margiana and western Bactria, rivers that used to disappear into the desert before reaching the Oxus River form deltas that are developed and irrigated in the IInd millennium BC. Among the oases on the deltas are those of Geoksiur (mentioned in Chapter 1), Merv (Marw) on the Murgab River (to be discussed further in Chapter 7), Sherabad, Ulambulak and Mirshada on the right bank of the Oxus in western Bactria, and Bactra (Balkh) to the south, capital of the region toward the end of the IInd millennium BC. Russian archaeological studies have shown that at Merv, [84]

habitation progressively migrates upstream as aggressive irrigation starves the down­stream extremities of the network.54 In Sogdiana to the north, one can also find rem­nants of early irrigation at Sarazm, on the Zeravchan River (the watercourse on which Samarcand is later founded). One also finds such traces on the lower course of the Oxus, to the southeast of the Aral sea, and at Dehistan, to the southeast of the Caspian (see Figures 1.3 and 7.1 for location maps).

Irrigated agriculture develops naturally in the valleys of the Oxus and its tributaries in western Bactria, at the foot of the great mountains of Hindu Kush and Pamir. French explorations conducted between 1971 and 1977 led to restitution of ancient canals, and helped establish a chronology of their evolution based on the dates of inhabited sites.[85] As agriculture spread to terraces situated higher and higher above the rivers, it was nec­essary to extend the canals so their intakes would be higher than the irrigated land, and to route the canals from one terrace to another. Often it was necessary to supply a canal from a tributary of the main river, sometimes far away, to bring water to the irrigated ter­races along the main river itself. For example, a plain that overlooks the Oxus at its con­fluence with the Kokcha was irrigated from the IIIrd millennium BC by a 25-km long canal fed by the Kokcha itself, and extending down to the Harappan settlement of Shortughai (Figures 2.20 and 7.2).

But even larger water resource developments appeared during the period from 1500 BC to the arrival of Alexander the Great – a period that also saw the development of major water supply in Assyria and Urartu. An artificial branch of the Taluqan River was excavated to develop the left-bank region of the Oxus, proceeding to the north against the natural drainage (Figure 2.20). This 50-km branch is today called the Rud I Sharawan. That the canal is artificial can be seen from its often perpendicular orienta­tion compared to the natural drainage. To the south, it follows the paths of old river arms or canals in the Taluqan plain; and to the north, it follows the valley of a small tributary to the Kokcha. The canal is deeply entrenched into unstable loess in its central portion that separates the drainage basins of the two rivers, attaining a depth of as much as 20 meters along one kilometer.

This is a significant operation of inter-basin water transfer, comparable to that implemented by Sennacherib to bring water to Khosr and Nineveh (Figure 2.15). Although it is impossible to say which of these developments came first, the necessary know-how clearly existed in both of these widely separated regions. What could have been the driving force for this project in Bactria: the decision of Persian leaders or the earlier work of a local Bactrian authority? Bactria had a strong cultural unity even before the arrival of the Persians. It may also have had strong political unity, but its his­tory is unknown.

When surface water cannot meet the needs of irrigation, one must tap groundwater. It was probably at the beginning of the Ist millennium BC, in Persia or in neighbouring lands, that a remarkable device for obtaining high quality water was invented: the qanat.[80]

This word means “reed” in Akkadian. The device comprises a gallery, or shaft, dug nearly horizontally from the flank of a natural slope back into the aquifer, but with a small slope (of the order of one or two per thousand) so that the water can flow out by gravity (Figure 2.19). In general, the construction of a qanat begins with the drilling of what will become its terminal well, called the “mother well”, through which the nature and level of the water table can be determined. Then the excavation of the gallery begins from the downstream end, making it possible to work in the dry until the aquifer is reached; this excavation thus proceeds all the way to the mother well. Intermediate wells, normally spaced at from 50 to 100 meters, provide for removal of the spoil from

the gallery, and provide for ventilation. The gallery can be several kilometers long, even reaching ten or more; the mother well can be the order of twenty to a hundred meters deep. The flow delivered by the qanat is generally from several, to several hundred, liters per second.

In 714 BC, Sargon II, king of Assyria (and the father of Sennacherib), is at war with king Rusa I of Urartu. He destroys the outposts of Urartu in the region of the lake of

Urmiah, as well as the qanats supplying water to the city of Ulhu, located to the east of this lake (60 km to the north of present-day Tabriz). The Assyrians, in a written account of this campaign, leave us an admirable description of the devices called “water outlets” that could comprise the first written evidence of the qanats:

“Ursa (i. e. Rusa) the king and lord of this land, pushed by his intelligence, showed his peo­ple how one manages the water outlets and creates a flow of water as copious as that of the Euphrates”.[81] [82] [83]

What is the genesis of this innovation? The Zagros mountains are a region of mines, especially in the area around the lake of Van. According to Henri Goblot, it was neces­sary to provide gravity drainage – to the surface – of certain mine shafts that were flood­ed after having pierced aquifers. In a country faced with the need to augment its water resources, the idea of making use of this drained water, and then to dig galleries express­ly for this purpose, surely picked up speed rapidly. The idea had a grand future: the Persians adopted it to develop the Iranian plateau, and in particular to provide water for their capital, Ecbatane. This is reported by Ctesias of Cnidus, a doctor of Xenophon’s expedition who was long held captive by the Persians:

“Having arrived at Ecbatane, a city located in a plain, she (again the legendary Semiramis! Here, it can only be a Mede or Persian sovereign) built a luxurious palace and she watched over the entire region with great care. The city was without water and there were no springs

in the vicinity; but Semiramis brought water to all the city, abundant and very pure water

52

thanks to her heavy investments.”

Cyrus brought the technique of qanats to Oman, and Darius brought it to the oases of Egypt. As we will see in subsequent chapters, the Romans developed the technique in all of the Near East, and as far as Tunisia, and the Arabs took it to Spain and Morocco. Migrants coming from the East brought it to the benefit of Saharan oases.