The calculation, giving these figures, is made assuming that regulation on the Barrage below Cairo is not called in to assist; the discharges allowed to Lower Egypt, after abstracting the above quantities, being sufficient to give a water surface at the level required for irrigation without any heading-up. Now, the quantity required to fill the lake to R.L. 25·00, as found before, is 11,800 million cubic metres. This, let us suppose, would be made up thus:—
| September | 30 | days at | 112 | millions | = | 3,360 | millions. |
| October | 31 | „ | 192 | „ | = | 5,952 | „ |
| Parts of August and November | 29 | „ | 86 | „ | = | 2,494 | „ |
| Total | 11,806 | „ | |||||
These discharges could be abstracted without affecting Lower Egypt irrigation, as it exists now, for eleven years out of sixteen. This, however, is not a quite correct statement as applied to each of the eleven years, since the discharges given in the table are the averages of those years in which a surplus discharge is available. Among these would be some very high years, in which it would be possible to fill the lake to a higher level than 25·00, and years when this level would not be reached.
Besides the years, for which the averages of surplus discharges are taken, there would be five years out of sixteen when the supply would fall short in September, the month of highest level.
There is no necessity in this stage to seek for a connection between the lake and the Nile for filling the lake, as the water would find its way in large volumes across all the low parts of the valley into the drainage depression along the edge of the Libyan Desert, now known as the Bahr Yûsuf.
The levels at which the Nile deposits are found in the Fayûm, the discharges which might be drawn off from the Nile, and the area of the Fayûm Lake are thus all in agreement with the supposition that the level of the Fayûm lake was yearly raised from about R.L. 20·00 to 25·00, and that the level attained was never sufficiently high to cause an overflow into the Wadi Raiân.
THE FAYÛM AS LAKE MŒRIS.
Judging then from the evidence furnished by Nile deposit and fresh-water shells, there is nothing to support the theory that there has been any great change in the Nile levels since the waters first found their way into the Fayûm. But whatever conditions of levels and volumes of Nile discharges we start with, we must at some date arrive at the period of present conditions. According to Mr. Petrie, however, there is good evidence (which I will give later on) to support the theory that in Herodotus’ time the Nile levels were 2 metres lower than now, and it is further probable that, at the time of the transformation of the Lake Fayûm into Lake Mœris, the Nile volumes were what they are now. I shall, however, discuss the subject, assuming that present conditions as to Nile levels also existed at the formation of Lake Mœris, and point out afterwards how the difference of 2 metres in the level at the time of Herodotus affects the conclusions.
The formation of Lake Mœris is credited to Amenemhat III. of the XIIth Dynasty, who gained a reputation for making great improvements in the Irrigation Department, and carrying out hydraulic works of immense benefit to the country, about 2500 B.C., or within 5000 years of to-day. Now, 5000 years, geologically estimated, is a very short time, and we may assume, without much chance of error, that he had practically the same general conditions to work with as regards relative levels of land and water and Nile discharges as we have to-day.
The Fayûm Lake would, in his time, have filled and emptied itself, and low Niles would now and then have occurred. But even without the occurrence of a low Nile it would have been observed that during the summer, when the surface of the lake in the Fayûm had reached its lowest level, there was a considerable area of land, formed of Nile deposit, laid bare. A shining light among the king’s subjects may have conceived a project for reclaiming this land from the annual inundation, submitted his project to the king and obtained his approval to its execution.