However, the amount of water in the Aral basin is governed not only by the volume of the Amu Darya, the Syr Dar'ya, and the subsurface water, but also by the direction of flow of these. As we have pointed out above, the Amu and Syr Dar'yas, and also the subsurface water, have a tendency to shift equatorward during periods of cold climate and a tendency to shift poleward during periods of warm climate. Obviously, these successive shifts of the surface and subsurface water in the basins of the two rivers will have a direct effect on the level of the Aral Sea: in cold periods it will be lower, and in warm periods it will be higher.
Very few factual data are available on the levels of the Aral Sea during ancient times. For various reasons, however, many investigators have concluded that this level underwent repeated, substantial variations. Fedorovich (1954) thinks that the small Paleo-Aral Lake which came into being during the Apsheron period did not exist throughout the entire Quaternary. Alenitsyn (1874), who discovered some winter-hardy animals native to a dry sandy desert on the islands of the Aral, concluded that they migrated there when the islands were joined to the desert by dry land. This could only have taken place at a time when the Aral, supplied just by the Syr Dar'ya, was a small, shallow lake, in comparison with the present day sea. "Whereas there are numerous indications that the level of this lake was very high in the preceding period, certain other facts show that there was also a time when the Aral did not fill up its entire basin. Its level and its area have become alternately greater and smaller ever since men began to dwell on its shores" (Reclus, 1898, p. 363).
During the first warm period the level of the Aral Sea was high. Yanshin (1953) assumes that the transgression of the Aral Sea which deposited the terrace with Gardium Edule L. shells was most likely related to the period when the climate of Turan was the most arid. This transgression may have taken place at the time of the Kel'teminar culture, since it occurred in the third millenium B. C. Consequently, this transgression is linked with a period when there was a general rearrangement (a poleward shift) of the drainage system. In particular, at this time the Amu Dar'ya made its maximum rightward deviation and flowed along the Akcha Dar'ya channel. This was the main reason why the Aral basin was then supplied with water.
During the second cold period (antiquity) the level of the Aral Sea was low. Tolstov and Kes', on the other hand, assume that at this time (middle
of first millenium B. C.) there took place a transgression of the Aral Sea which raised the level to 3.5 or 4 meters above the present level. Their assumption was based on the geomorphological features of the southeastern part of the Aral Sea, on the distribution of sea shells there, and on the fact that Iron-Age sites have been found on its shores. However, the assumption of Tolstov and Kes' is contradicted by the following statement of Kes' (1958, p. 95): "...The Aral Sea became a large basin only after the Amu Dar'ya and Syr Dar'ya began to discharge completely into the vast Aral depression. Both of these rivers have highly variable courses: during the most recent stages of geological history the rivers migrated repeatedly, sometimes discharging into the Aral Sea and sometimes turning aside and flowing, partially or completely, into other Central Asian depressions. Whenever the Amu Dar'ya turned aside for any long or short period (for example, toward Sarykamysh), the level of the Aral Sea dropped sharply".
Let us also quote the following statement of Reclus (1898, p. 363): "The Aral Sea depends entirely on the extent to which it is supplied by the two rivers discharging into it. If the Amu Dar'ya and Syr Darya move away from the Aral depression and flow into the Caspian instead, the "Sea of Khorezm" will inevitably become shallow and then dry up after some years. However, as we have seen, the Amu Dar'ya has already deserted the Aral twice during historical times, and in addition one branch of the Syr Dar'ya once flowed into the Caspian via the Oxus (Amu Dar'ya) channel". Thus, in antiquity, when all the tributaries of the Sarykamysh delta were flowing, the level of the Aral Sea could not have been very high. Most likely, this body of water was then just a small steppe lake.
Now let us consider the presence of Iron-Age sites on the shore of the ancient sea, the level of which was 3.5 to 4 meters higher than at present. We can assume that the settlements and the shore itself developed at different periods, since the shoreland had to form before dwellings could be erected on it. During antiquity people were probably forced to settle in elevated places, since at that time the Syr Dar'ya deviated to the left and abundantly watered the territory around the southeastern part of the Aral Sea. Thus, according to Tolstov, this entire region was a kind of "Central-Asian Venice" during ancient times.