The Course of the Oxus River

The Oxus River of the Greeks has been known to historical record for literally thousands of years. Its present-day name is the Amu Darya (or Amu river) and its course can be traced in any atlas. Turn to the map of Western Asia and find it: the river springs up in the Pamir mountains, runs west through what was once Baluchistan, and forms part of the border of what is now Afghanistan; here, along its banks, were once lapis-lazuli mines famous throughout the ancient world. It runs, gradually curving northward, between a desert marked Kizil Kum (or, the Red Sands) and another marked Kara Kum (or, the Black Sands). Along the way, it supplies irrigation-water for the ancient oasis kingdoms of Khiva and Khwarism. Eventually it empties itself into the southern tip of the Aral Sea.

That much is plain.

However there is a problem with ancient accounts of the river. Some of them say that the Oxus runs into the Caspian Sea, not the Aral.

Now, large rivers in alluvial plains are prone to shifting their banks. Anyone who's read Mark Twain's Life on the Mississippi knows this; Twain trained as a riverboat pilot, and it was a difficult job--because the Mississippi in his day changed its entire face from year to year. Its banks were constantly eroding, the river-bottom itself shifted depth, and farms along the shore could become islands without warning. Likewise, the Oxus' companion river the Jaxartes (now known as the Syr Darya) used to shift its banks up to a hundred yards sideways in the course of a single year. In the mid-eighteen-hundreds, Russian engineers cut the first railway across the deserts through which the Oxus and Jaxartes both run. They did not come near the Oxus, but they (and others) left records of the Jaxartes: its vicinity, they said, was filled with a constant noise and tumult of crashes, as pieces of the river's bank slid into the water. The river-bed moved even as the engineers tried to span it with bridges, and a railway bridge built one season could be left high and dry by next summer, while the Jaxartes itself proceeded on its merry way.

The bed of the Oxus is also variable. Looking back through Islamic and other documents, one finds that though it presently runs to the Aral Sea, it could well have been otherwise.

The Oxus has a complicated lower course. In ancient times, it was dammed by at least two large works: one lies near Amuy, at the beginning of the lower course, today called Carjuy, and the other was at Gurganj where the river bifurcated into two beds. (Al-Muqaddasi at about 985 AD describes the Gurganj dam as an amazing work of engineering, built all of wood and resembling sea-works in wood and wickerwork.) Just above this point, the river's course splits: a small branch runs north to the Aral Sea, but another larger branch runs west, past the town of Gurganj, into a deep depression called the Sarykamysh, about 155 miles SW of the Aral Sea. This Sarykamysh depression lies about 300 ft deeper than the river's bed at Gurganj.

From the Sarykamysh, a dried-up riverbed called the Uzboi runs through the Balkhan hills all the way to the Caspian Sea.

The Aral Sea's level is itself only about 60 ft deeper than river level at Gurganj. Everything depends on the river's level. If the larger branch of the Oxus empties into the Sarykamysh then in effect the Sarykamysh must be considered a lake--and hence, may even be called the true basin of the Oxus. Here the larger part of the river's waters end, like those of other desert lakes: evaporating, without exit. However if not too much irrigation-water is used upriver, the filled basin of the Sarykamysh overflows and discharges through the Uzboi into the Caspian.

Here is a history of the river's several courses:

Present-day back to about 1575 AD: both Jaxartes and Oxus rivers (the Syr Darya and Amu Darya) run to the Aral Sea, as they do today. Around 1575, the Oxus' course changed.

Evidence: Abulghazi Khan, born 1603 and ruled 1642-63, wrote that the course of the rivers changed thirty years before his birth. Jenkinson, the envoy of Queen Elizabeth, wrote in 1558-59 describing how the Oxus ran in the Uzboi channel but no longer reached the Caspian Sea, and predicting that the demands for irrigation would soon lead to a complete desolation of the region.

From 1575 to 1221 AD (the date of the Mongol invasion) the Oxus discharged into both the Caspian and the Aral Seas--usually the main branch ran to the Caspian, but during at least one period, a side branch also ran to the Aral. Thus the seas were linked. There are perhaps four courses which this river seems to have followed at various periods. At this time, the river split just above the town of Gurganj; the main branch ran south of the town (coursing westward) filled the Sarykamysh (about 150 miles SW of the Aral Sea) ran through the Uzboi bed SE to a gap in the Balkhan hills (over 100 miles SE of Krasnowodsk) and discharged into the Caspian Sea opposite the group of islands called Oghurtcha or Aghyrtcha.

Evidence: 1) In 1392, Zahir al-din al-Mar'ashi a native of Mazandaran, describes in detail every waterway from Mazandaran to Khwarizm. 2) Shihab al-din 'Umar al-'Umari (d. about 1348) remarks on the authority of witnesses from Khwarizm "On Khwarizm borders ... a country ... Mangyshlaq, a steppe ... separated from the Djaihun (Oxus) by the Ak-Balkan hills north of Khurasan." 3) In the reign of Oldjaitu (1304-16) Hamdallah al-Mustaufi speaks of the caravan route from Gurgan by way of Dahistan (the modern-day Mashhah i Misriyan) to Gurganj, which had just been surveyed and was 110 fars (380 miles) in length between Dahistan and Gurganj--and says on the Uzboi there was a great waterfall, called in Turkish gurledi, roaring; and a side branch of the Oxus fell into the Aral Sea--and tells how the discharge of water of the Oxus into the Caspian had raised the sea's level so that the peninsula of Abaskun (Pliny's Socanda) was submerged. 4) Marino Sanudo (1325) and al-Bakuwi (beginning of the fifteenth century) confirm this description.

The river changed its course in 1221 because the armies of Genghis Khan destroyed the great dam at Gurganj. Evidence: record of ibn al-Athir, xii, 257.

From the tenth century to 1221, the Oxus seems not to have flowed to the Caspian. However the records are vague and contradictory. Evidence: the earliest is from ibn Khurdadhbih, but is vague because of variants in texts; Ibn al-Raqih at the beginning of the tenth century, tells a legendary and thus unverified story; ibn Rustah (writing between 903 and 913 AD) carefully describes the mouth of the river's main branch, showing that it discharged into the Sarykamysh. Istakhri (about 960 AD) says the river flows into the Aral Sea. Al-Ya'qubi (about 890) however says it falls into the Caspian.

Muqaddasi--the author of the description of the Gurganj dam--circa 985 AD, relates a legend according to which the Uzboi as the Oxus' main branch had once, long ago, carried water (ergo, it did not do so in Muqaddasi's day) and reached the town of Balkhan in Khurasan; however once the inhabitants of Kath (the pre- and early-Muhammedan capital, 30 miles east of Khiva) had diverted the whole course of the river to their easternmost branch for 24 hours, the river did not turn back to the Uzboi channel and the whole Balkhan region was laid waste by lack of irrigation.

According to Byzantine historians, this Balkhan region (then called Balaam) was settled before the Muslim conquest; ergo, it had irrigation then, and the Oxus flowed into the Caspian. (It is not possible to settle an entire region in the Middle East, without a major river to provide water.)

In high antiquity, Strabo (C.510) and Polybius (x,48) both mention a waterfall of the Oxus; this is probably the waterfall of the Uzboi mentioned by Hamdallah.

Evidence: Seleucus of Persia planned to connect the Caspian by a canal with the Black Sea, if a report of the exploration of the Caspian should be favorable. This took place not later than 282 B.C. Xenocles, the gazophylax of the archives of Babylon, had made accessible for study to Patrocles (the agent sent to study the problem) the original documents with observations by Alexander the Great's general staff concerning the problem. The purpose of the project was to develop trade. Strabo xi (c.509 and 518) quotes Aristobulus speaking of the transport of Indian merchandise on the "well navigable" Oxus (in the present day, the river is not navigable; the water-level fluctuates too much) to the Caspian and across the sea and thence through the Caucasus to the Euxinus or Black Sea . . . This plan of Seleucus' implies that the Oxus was known to discharge into the Caspian, else the whole project would fall to the ground. Hence, in the early third century BC the Oxus flowed into the Caspian.

Major source: Ernst Herzfeld, Zoroaster and His World, vol II, 1947. All this research and results are his, not mine; I reproduce them here without permission, without hope of gain, but with awed respect.

Return to the Asia page