The lake may have been chiefly filled by the Bahr Yûsuf and the flood waters inundating the Nile Valley, but, to fulfil the conditions of a six months’ flow-in and six months flow-out of the lake, under the new conditions supposed, and retaining a maximum lake-level of R.L. 22·50, the canal of supply would have to have its off-take from the Nile moved to a point about half-way between Beni Suef and Biba. Supposing the Bahr Yûsuf and the flood water of the Nile Valley filled this lake during the flood months and the Bahr Yûsuf ceased to flow with the end of the flood, the canal from between Beni Suef and Biba would have had to supply only about 10 million cubic metres a day to make good the loss by evaporation, if the lake-level was not to be allowed to fall below R.L. 22·50 till the return-flow to the Nile was required. But there is no reason to suppose this to have been a necessity. With a lowest level of 17·50 instead of 19·50, the problem of the lake as a relieving and supplementing reservoir to the Nile, with houses and cultivation above its highest levels, is much simplified, and a large margin is given between R.L. 20·50 and 22·50 for increasing the volumes given in my former calculations, to render the lake a more efficient safety-valve for excessive floods, and for moderating the fall of the Nile to low discharges by giving back to it a more abundant outflow.

Accepting this view of the range of the lake-levels, we shall have to look upon the Magnûnah Canal and its branches as channels of return-flow to the Nile for the commencement of the period of outflow, which would afterwards cease to carry any discharge in summer, when the lake-level had fallen below about R.L. 19·00. For the remaining period of outflow the Kosheshah Escape branch from Abûsir-el-Malaq to the Nile and the branch to Memphis skirting the Libyan Hills, would have carried all the discharge returning to the Nile Valley.

The peculiar isolated piece of Nile desert opposite Lahûn and the cultivated strip of land between it and the main desert, through which the Bahr Yûsuf flows into the Fayûm, seems to lend itself to the regulation of the entry and exit of the Nile waters. To control the entry of the waters a regulator A and cross bank a b from the island desert across the Bahr Yûsuf to the main desert on the west could have been made. (See map, [Plate XXI.])

The excess water, excluded from the lake by regulation on A, would have found its way along the east of the patch of desert as it does to-day.

To retain and to control the exit of the water, a regulator B and its bank c d might have been added, where shown on the map, or anywhere between B and the end of the narrow band of cultivation at C. There is, however, no evidence to show that such works did exist, but Strabo’s statement, the presence of the Lahûn pyramid and the situation of the villages Lahûn and Manshîyah make it perhaps probable that there were some important works connected with the lake in their neighbourhood.

The reason for the peculiar alignment of the present bank g D B c which closes the gap into the Fayûm, is difficult to imagine, as the bank is at least three times the length it would have been, if it had been formed in a direct line across the gap. But it has suggested itself to me, that the length B c may be part of the original bank d B c, that may have crossed from side to side of the valley of exit, and on which the villages of Lahûn and Manshîyah were built.

Trying to find some explanation for the alignment of the existing bank, it had also occurred to me, that the line of the bank may have followed the ridge of the bar, that would have been formed across the wider part of the entrance to the Fayûm by the high level water flowing in. This bar would, if it had existed, have been the first land to show above water on the subsidence of the floods, and may have been chosen, on the occasion of one of the repeated breaches at Hawârat-el-Maqta, as the most convenient line for forming a bank to shut out the Nile flood. But this would have been at a later date, after Lake Mœris had ceased to perform its functions of a Nile regulator.

However, I think the former supposition, that the bank B c was part of an old bank, formed for quite another than its present purpose, and that the bank g B was subsequently made between Lahûn and the desert (perhaps when the existing old Lahûn regulator was made), a more likely explanation. The length B d would have disappeared after it ceased to perform any useful function.

There may have been both, or one, or neither of the regulators A and B, but if there was a regulator at Hawârah at the head of the lake at F, there would have been little to be gained except additional security from the regulator A.

If then we suppose that the bank c B d and the regulator B only existed to collect the flood waters, and turn them into the lake, and that a regulator at Hawârah at F also existed to keep excess water out of the lake, such an arrangement would agree with Strabo’s statement that “when the river falls, the lake again discharges the water by a canal at both orifices, and it is available for irrigation. There are regulators at both ends of the canal for regulating the inflow and outflow.”