Another condition which M. Linant’s lake is far from satisfying, is the depth, which Herodotus gives as 92 metres. Linant makes his lake depth 9·60 metres, assuming that his lake area occupies a plateau, which it does not. The greatest depth of his lake, according to the actual levels of the ground included in it, would be 18·60 metres, against the 92 metres of Herodotus. This condition therefore is not satisfied.

Faulty Foundations on which the Linant Theory was built.—Had Linant Pasha had before him a contoured map of the Fayûm, I believe he would never have enunciated his theory. The Minia wall made an undue impression on him and has been his stumbling-block. He clearly traced the remains of a large bank from Edwah to Biahmu, and less plainly to Medineh, but after Medineh he found no traces of a bank, but being desirous of connecting up with the big wall, supposed that it must have passed through certain villages leading to it. This wall, closing a valley encircled by contour R.L. 15·00, was probably constructed at a much later date, or at any rate independently of the bank of which the remains are found on the other side of Medineh, and for a different purpose. If this wall had been originally higher there would be remains of the high parts at each end, where breaches had not carried away the original wall. This we do not find, but on the contrary, the crest of the wall is at one uniform level from end to end, and appears to have been added to, instead of taken from. From an examination of the abutments of the bridge, built in the line of the wall, the original wall appears to have been constructed of stone, and to have been widened subsequently by an addition of coarse brick masonry of rough bricks, in mortar made of lime and clay, probably with the view of obtaining width enough to carry an aqueduct along the top of the wall. None of the masonry is sufficiently good for this purpose, and so, no doubt, the water, leaking from the aqueduct channel, gave rise to breaches in the wall, of which the signs are evident in the blocks of masonry lying scattered about on its down-stream side.

The cross-section of the wall, Plate X., gives its dimensions at a point near the bridge, where its height is greatest. Below this cross-section another of the Edwah-Biahmu bank is given for a comparison of the levels of wall and bank.[3]

Erroneous Data employed by Linant.—In Linant Pasha’s Atlas, published with his Mémoires, is to be found an extraordinary section of the Fayûm from Lahûn to Birket-el-Qurûn, in which the land from Lahûn to Medineh is shown as being higher than the land of Beni Suef on the Nile Valley side of Lahûn.

I reproduce his section on Plate XI., and below it I give a section showing the actual levels. As Linant appears to refer all his levels to the rock bed at Hawârat-el-Maqta, which he makes 32·80 metres above sea, whereas it is really 21·00, it is necessary before comparison to apply a correction of - 11·80 to all his levels.

Comparing the figures after correction with those of the “actual section,” it will be found that Linant puts the level of the Beni Suef lands 5½ to 7 metres too low, and that of his first plateau 6 to 8 metres too high, with reference to the rock bed at Hawârat-el-Maqta. According to the corrected figures his Lake Mœris level would be + 32·00, but how he gets it to that level it is difficult to understand, inasmuch as he says that his lake is filled by the Bahr Yûsuf, whose high-water level is shown 4½ metres lower. (The Birket-el-Qurûn level, after applying the correction, becomes 40·80, which must have been about its correct level in Linant’s time.)

I give another instance of error with reference to Linant’s conception of the first plateau. Writing of the bank from Edwah to el Alam, he states that the land to the south of this bank was about 2 metres below its crest, and to the north of it from 8 to 9 metres, which difference, he says, is explained by the deposition of silt in the interior of the basin formed by the bank, as is always seen elsewhere at all the banks of the inundation basins of Egypt. This great difference of level of the country surface on either side of the bank would have been very suggestive in a sense favourable to M. Linant’s theory had the difference of level been a fact instead of a fiction. The cross-section of the bank given on Plate X. shows its actual state with reference to the land on each side of it.

It seems scarcely necessary to discuss further a theory that was based on such erroneous data, but as the data were not known to be erroneous, and Linant propounded his theory with an air of authority, it has had considerable success in getting itself accepted. Guide-books, and even books used as school text-books on Egyptian history, show that his theory has been hitherto judged the correct one.

It is, however, satisfactory to find that in the fifth edition (1890) of ‘Ancient Egypt’ by George Rawlinson (The Story of the Nations Series) the exact size and position of “Amenemhat’s reservoir” is admitted to be sub judice, and it would appear that this desirable attitude is the result of a challenge of Linant’s theory by Mr. Cope Whitehouse, who is reported as believing that the water was freely admitted into the whole of the depression (i.e. the Fayûm), which it filled, with the exception of certain parts, which stood up out of the water as islands from 150 to 200 feet high. Nevertheless, in spite of this new attitude towards the Linant theory, the map representing Linant’s Lake Mœris is to be found at the end of the book, without any remark to prevent readers from being misled by it, the map being described as “Map of the Fayoum, showing the Birket-el-Keroun and the artificial Lake Mœris.” (Plate VII. is from an earlier edition.)

The Linant theory, examined in the light of the more accurate knowledge gained of the physical features of the Fayûm, and tested by the application of figures to determine its possible performances, can no longer stand, but falls to pieces; and the wonder is that, based as it was upon erroneous data and propped up by no solid support of facts, it stood so long. It may be said of it, to the credit of its author, that it was ingenious, but not that it was true.

Since writing the foregoing concerning the Linant theory, Mr. Cope Whitehouse has kindly lent me his first papers on the subject, the earliest paper, that I had previously seen of his writings, being that which was read by him at the Manchester meeting of the British Association, September 2nd, 1887. I now find that he has been before me in stating many of the arguments I have used against the Linant lake theory, but it is satisfactory to find that we have independently arrived at the same conclusions, though by no means surprising, as I believe that any one, with the same amount of personal acquaintance with the Fayûm, would be naturally led to hold the same views of this fantastic theory. As early as 1882, Mr. Cope Whitehouse pointed out that the Linant lake satisfied none of the conditions which a lake professing to be Lake Mœris must satisfy, and he concluded one of his papers with a remark expressing his conviction that, when Lake Mœris shall be recognised by the light of discoveries yet to be made through further research, the site of the ancient lake will in no case be found to be that of the reservoir of M. Linant de Bellefonds Pasha.

Mr. Whitehouse, at the same time as he lent me his papers of 1882, also lent me a copy of Dr. G. Schweinfurth’s letter to Paul Ascherson on a journey undertaken in the depression of the Fayûm in January 1886, in which letter I find I am anticipated again in a footnote on this theory, which gives an argument not given by Mr. Cope Whitehouse, and runs as follows:—

“Here it must be mentioned that one of the most important points on which Linant grounds his Mœris theory is the Dams[4] which has proved quite fallacious. The dam at Adwa (Edwah) geology shows to be layers of gravel; the stone dam at Minia is, on the other hand, a weir for the Bahr-el-Wady, and is evidently throughout its whole length of later date. Besides, it fills up only the deep curves of the ground, and has no continuation on the rising grounds.”

Dr. Schweinfurth, as well as Mr. Whitehouse, had thus pointed out the weakness of the Linant theory, but still we find it living and taught as a true theory so late as 1890; and this is my excuse for repeating the arguments which have not yet succeeded in overwhelming it, notwithstanding its feebleness, so much support does it derive from its parent being a reputed authority.

But we have public acceptance of the Linant theory so late as 1892, and by so eminent an Egyptologist as Brugsch Pasha, who communicated his views to the Société Khédiviale de Géographie in a paper read in Cairo, on the 8th April, 1892, the title of the paper being “Le Lac Mœris d’après les monuments.” A quotation from this will show that he accepts the Linant theory.

“De nos jours, les traces visibles de cet immense bassin d’eau (le lac Mœris) ont disparu et les savants les plus distingués se sont en vain efforcés pendant longtemps de retrouver ses anciennes limites sur le sol moderne de la province du Fayoum.

“L’opinion la plus généralement acceptée au sujet de sa position a Linant pacha pour auteur. C’est lui qui, le premier, a rejeté l’idée de reconnaître le bassin du lac Mœris dans le Birket-el-Kouroun de nos jours, c’est-à-dire ‘le lac des Cornes,’ situé comme on sait, à l’ouest du Fayoum. Suivant les recherches très minutieuses de l’illustre savant, il faudrait, au contraire, se diriger vers le côté oriental de la province susnommée, et, notamment, vers les plateaux bien connus de Hawara et de El Lahoun, où deux pyramides construites à l’époque de la XII^me dynastie (vers 2500 a. J.C.) excitent encore la curiosité des voyageurs.

“H. Lepsius, mon savant compatriote qui, il y a presque cinquante ans, a eu l’occasion d’examiner sur les lieux les résultats obtenus par Linant pacha, n’a pas hésité à déclarer dans un Mémoire spécial que le savant français avait fait la découverte la plus brillante et la plus indubitable quant à la véritable position topographique du fameux lac Mœris. Les doutes qu’il exprima à la même occasion ne s’appliquaient qu’à l’extension du lac vers le nord.

“Depuis Linant et Lepsius, aucun savant sérieux, du nombre des géographes et des Égyptologues, ne s’est opposé à l’opinion émise par ces deux illustres auteurs.”

Apparently Brugsch Pasha does not class Dr. G. Schweinfurth, Mr. Flinders Petrie, and Mr. Cope Whitehouse as “savants sérieux,” for they have expressed themselves as opposed to the Linant theory. Lieut.-Colonel J. C. Ross, C.M.G., late Inspector-General of Irrigation, Egypt, justly renowned for his power of comprehension of the levels of any part of the country, which he studied professionally, at one time gave much of his attention to the Fayûm, and especially to that part which was known as Hod-el-Tuyûr, and which is the depression embraced in the area which Linant calls a plateau and in which he localised his supposed Lake Mœris. I think I may say, without fear of contradiction, that Colonel Ross’s examination of the ground in question was much more thorough and more prolonged than that of M. Lepsius; but whereas the latter did not hesitate to accept Linant’s theory and to style it the most brilliant and certain discovery as regards the position of Lake Mœris, Colonel Ross on the contrary rejects Linant’s theory and thinks that the lake was north of the Edwah bank and not south of it.

Mr. Petrie has also clearly expressed the same views as Colonel Ross on this point.

I too have had advantages of studying the ground itself and the way the water runs, such as few have had, and have come in for the legacy of Colonel Ross’s levellings and maps.

The names of Linant and Lepsius do not therefore carry with them sufficient authority to override the facts, which are ascertained to be such by a more thorough examination of the country and a better knowledge of the physical features of the province.

Irrigation officers may, perhaps, not be classed as “savants,” but they have at least as much right to be heard as any other body of experts on such a subject as Lake Mœris, which is more than anything else an irrigation question, and one that has especial interest of a more or less practical nature at this time, when the question of the construction of Nile reservoirs for the storage of the surplus waters of the Nile is under consideration.

COPE WHITEHOUSE THEORY.

Mr. Cope Whitehouse, who has a personal acquaintance with the Fayûm, and has studied the question of Lake Mœris, though with a prejudice in favour of giving the Wadi Raiân a leading part, would possibly have been inclined to hold the same views as I do, had it not been for his anxiety to recommend the Wadi Raiân for future use by magnifying its imaginary past performances. But as his views about the Wadi Raiân are the essential and distinctive part of his theory, we do not agree.

For ten years Mr. Whitehouse has been brooding, as the faithfullest of mothers, over his theory, looking for a practical project to be hatched therefrom, but, as time passes, he begins to show signs of impatience, and fears lest his egg be addled. The possibilities as to what the chick may be when it appears, are set forth in Chapter V. of this paper, for I cannot but think that the egg is a good one.

Mr. Whitehouse believes that in prehistoric times, before artificial works of control were made, the Nile flowed into and submerged the whole Fayûm, which was filled at high Nile, and that when the flood subsided, the return flow, that took place from the Fayûm to the Nile, prolonged the period of inundation by at least two months. He also believes that the river flowed in a single channel along the eastern desert.

So far most of us who have our theories about the Fayûm travel together, with but small differences on the way. But after this our roads diverge, and each thinks the road he has selected leads to Lake Mœris. But they cannot all go there.

Leaving prehistoric times, and coming to the period of ancient history, Mr. Whitehouse holds that there were two lakes. At first the northern lake, the Fayûm, was a lake and marsh serving as a backwater to the Nile, while the southern, the Wadi Raiân, was dry. So far I agree with him, but now we part company. Subsequently, he imagines, engineers of an alien race diverted the flood waters into the dry Wadi Raiân to the south-west and evaporation dried up the Fayûm, which was then irrigated by a system of canals. The Wadi Raiân basin becoming, full served as a reservoir, and was, according to Mr. Whitehouse, the “Mœridis Lacus” of Ptolemy.

Later on, in 1890, Mr. Whitehouse explains his views in terms which are not quite in agreement with the foregoing, for he then supposes that the natural backwater of the Nile included the Fayûm and Wadi Raiân (with its minor basins Wadis “Safir” and “Lulu”) at one and the same time, and that these combined basins were filled to the level of high Nile, which he puts at R.L. 30·00.

I think, however, his present views are, that at first the Wadi Raiân formed part of the Lake Mœris of Herodotus, of which the Fayûm was the main part; and that afterwards the Wadi Raiân alone formed the “Mœridis Lacus” of the Ptolemaic maps, the Fayûm having been brought under cultivation, after its waters had been dried up by evaporation.

As the Wadi Raiân is the prominent feature of this theory, I will give here in full Colonel Western’s official description of it:—

Description of the Wadi Raiân.—“This valley or depression in the Libyan Desert, discovered by Mr. Cope Whitehouse in 1886 (really three or four years earlier), lies immediately to the south-west of the Fayoum Province, but separated from it by a range of low hills, 2 kilometres in width and with heights of about 40 metres above sea-level. Two passes, however, leading from the Gharak basin, with level of + 26 metres, have been found in this dividing range and, except for these two passes or entrances, the Wadi is everywhere bounded by hills of at least + 36 metres.

“The soil of the Wadi is for the most part composed of desert sand and pebbles overlying in places a yellow clay, but this desert sand is for about one-sixth of the area hidden by drifting sand-hills or ridges rising some 5 to 10 metres above the general plain.

“Towards the south of the Wadi there are two fresh-water springs; and near these a few date-trees and some brushwood grow.

“The deepest level of the Wadi Raian reaches 40 metres below sea-level.

“To the east of the Wadi, and connected at a level of + 55, is the Wadi Muellah, a valley about 1½ kilometres wide and 7 long. Its lowest depression is at + 25.

“In the Wadi Muellah there are ruins of ancient buildings, and a fair amount of coarse vegetation near them.

“Another small depression, also connected with the Wadi Raian, has been found lying to the south of the Gharak basin of the Fayoum, and only separated by a ridge at level + 35, and 1 kilometre in width. This depression is some 10 kilometres in length by 4 mean width, and has a bottom at about + 15 metres.”

Now, there is no evidence whatever that the Wadi Raiân had ever any possible communication with the Nile except by way of the Fayûm depression through the two gaps in the encircling walls of the Wadi, the sills of which are stated by Colonel Western to be at R.L. 26·00, but which later surveys, not yet published, show to be one at R.L. 26·00, and the other at R.L. 27·00. The Wadi Muellah, on first inspection of the map, appears to offer the most likely line of communication with the Nile Valley, but an examination of this Wadi at its upper end towards the Nile Valley gives no evidence of any such communication having ever existed.

What seems a conclusive proof that the Wadi Raiân was never in direct communication with the Nile Valley, is the total absence of all trace of Nile deposit within the limits of the depression.

If the muddy waters of the Nile in flood entered a lake 60 to 70 metres deep, the silt would be deposited and remain, for the return flow from the uppermost stratum back into the Nile would disturb none of the Nile mud brought in. After a long succession of such annual deposits, the depth of deposit would be considerable. In the Fayûm entrance we find such a deposit up to R.L. 25·00, and it is to be noted that the Wadi Raiân was supposed by Mr. Cope Whitehouse to have been in working order, as Lake Mœris, after the Fayûm ceased to be so, and therefore there would have been less time for the disappearance of the Nile deposit of the Wadi Raiân than of that of the Fayûm.

In the Wadi Raiân, Nile deposit has not been found, though eagerly looked for. I think this fact is fatal to Mr. Cope Whitehouse’s theory of a direct communication between the Wadi Raiân and the Nile or Bahr Yûsuf.

Dr. Schweinfurth thus expresses his views as regards fresh-water deposit in the Wadi Raiân:—“The basin (Wadi Raiân) exists, but it comes from geological time, does not belong to the Nile, and offers nowhere in its tracts at some distance from the Fayûm any trace of a fresh-water formation. . . .

“The traces of a settlement of water and layers of Nile earth which are said to exist in some parts of the depression are certainly absent. The grey clay-layers of the old sea with shells of fresh water, innumerable fish vertebras, bones of tortoises, &c., are not to be overlooked where they exist. I could prove such fresh-water formations on the road from Talît over Raiân and Medinet-el-Bahrl (27 kilometres to the west of the actual lake) only at a distance of 8 kilometres from the lake (Birket-el-Qurûn). The yellow Eocene marls with stripes of erosion, results of the wind, moving sand, and of periodical rains, are not to be confounded with these lake formations. A man who does that will find traces of old water and Nile earth everywhere in the deserts of Egypt.”

Later on in the same letter he says that the question, whether fresh-water formations exist in the basin of Raiân or not, is to be answered in the negative.

But supposing an indirect communication with the Nile by way of the Fayûm Lake, it is easy to understand that no Nile deposit would be found in the Wadi Raiân, even if it had been thus repeatedly filled, because the top water only would begin to spill over into it after the Fayûm Lake level had risen above R.L. 26·00, and after the water had travelled at an extremely low velocity to a long distance from the point, at which it first spread itself out in the Fayûm Basin.

But I regret, for the sake of Mr. Cope Whitehouse’s feelings, that even this cannot be admitted to have taken place, for in every situation where Nile water has been, fresh-water shells of distinct species are always found, and their total absence in the Wadi Raiân is sufficient proof to geologists that Nile water has never been there.

To the conclusion that the Wadi Raiân was never in direct communication with the Nile must therefore be added this further conclusion, that the Nile water never entered the Wadi Raiân at all, even by the only possible entrances over the sills on the side of the Fayûm Lake.

Mr. Cope Whitehouse has not distinctly stated how he supposes the Wadi Raiân was put into direct communication with the Nile, but I believe there are only three possible theories, each one without a particle of evidence to support it. One theory supposes a connection along the bed of the Wadi Muellah, another a tunnel through the hills dividing the depression from the Nile Valley, and the third a hill-side canal fed from the Nile waters entering at Lahûn and carried along the south slopes of the Fayûm.

In the absence of any evidence witnessing to the previous existence of such connections, and in the face of the fact that the Wadi Raiân contains no Nile deposit, I do not think that Cope Whitehouse’s Raiân-Mœris or Ptolemaic-Mœridis-Lacus theory can stand.

Failing better support to his theory, Mr. Whitehouse has called the Ptolemaic maps to his aid, and in his pamphlet on the subject he has reproduced the map of Egypt from the Atlas of Cl. Ptolemy, of which I here repeat the copy, with an outline map of the Fayûm, Wadi Raiân, and part of the Nile Valley, taken from ‘Egyptian Irrigation,’ by Willcocks, and which was compiled from the latest surveys in 1888 (Plates XII. and XIII.)

Mr. Whitehouse considers that the Ptolemaic map has been most accurate in giving the exact shape of a lake in the desert, whereas the representation of the features of the much better known Nile and Nile Valley is evidently most incorrect, and much distorted in longitudinal and transverse dimensions.

If, however, any argument can be based on the shape of the “Mœridis Lacus” of Ptolemy, as compared with existing depressions, it seems to me that its shape resembles much more closely the outline of the Fayûm Province, with the Bahr Yûsuf indicated, than it does that of the much indented Wadi Raiân.

Mr. Petrie has furnished me with the following observations on the Ptolemaic maps.

The Ptolemaic maps are built up from itineraries and ship routes, checked by a few latitudes. Now we know this much from Ptolemy, that Skiathis, Bakkhis, Dionysias, the Small Oasis and the Great Oasis were on one route, and that on this route Lake Mœris was passed. This was the desert itinerary from Alexandria to the Great Oasis.

Using another distinct itinerary from the Nile Valley, the route passes to Arsinoë (the modern Medinet-el-Fayûm) and Ptolemais (the modern Talît), and then on to Behnesa, without any connection being made with the Bakkhis-Dionysias route. Hence it is presumed that these two routes did not cross each other. It is therefore concluded that Dionysias can be identified neither with the ruins on the Wadi Muellah (as Cope Whitehouse identifies it), nor with Lahûn, and that it was probably on the west of Lake Mœris. Mr. Petrie (to whom I am indebted for the whole of this reasoning) supposes Bakkhis to have been at Dimeh (Dimay), and Dionysias somewhere at the extreme south-west of the Raiân valley.

If this conclusion is right, and if the Fayûm, or the Wadi Raiân, was the Lake Mœris of Ptolemy, the Lake has been placed too much to the west on the map, and should have been shown on the east of the line joining Bakkhis and Dionysias. In any case the Ptolemaic evidence, when sifted, does not support Cope Whitehouse’s theory, that the Wadi Raiân was the “Mœridis Lacus” of Ptolemy.

To show what little faith can be put in the identification of some of the ancient towns with modern remains, I may mention that Dr. Schweinfurth says of the monastery in the Wadi Raiân, that it is “evidently the Bakkhis of Ptolemy.” Thus we have this monastery identified as Dionysias by Cope Whitehouse, as Bakkhis by Dr. Schweinfurth, whereas Flinders Petrie places both Dionysias and Bakkhis on the far side of the Fayûm depression. Who shall decide when savants disagree?

In his papers on Lake Mœris, Mr. Whitehouse makes reference to two lakes, and I believe his theory of two lakes is based on some ancient maps.[5] I have not seen the map or maps, but I should expect the lakes represented to be intended for Lake Qurûn in the Fayûm, and a corresponding lake in the Gharaq basin. The Gharaq basin is the Fayûm depression repeated on a small scale, and at some period of its development towards total reclamation from the waters that covered it, it must have had a lake at its south and lowest end, corresponding to the Birket-el-Qurûn, but of smaller dimensions.

The Gharaq basin is connected with the Fayûm depression by a gap in its surrounding higher lands with sill at R.L. 16·00. Consequently the basin would not have begun to dry up from evaporation till the Fayûm Lake had fallen below R.L. 16·00, and probably the fall was not continuous, but, through some accident at Hawârah or elsewhere, the Fayûm Lake, after falling below R.L. 16·00, may have risen again and re-drowned the reclaimed land in the Gharaq. This may have occurred more than once, and have given rise to the name “Gharaq,” or the “Flooded.”

Mr. Whitehouse, in his latest expression of views, supposes the Fayûm and the Wadi Raiân were filled to R.L. 30·00. I have given reasons for concluding there was never any Nile water in the Wadi Raiân. The evidence furnished also by Nile deposit and fresh-water shells on the Fayûm side of the entrance at Lahûn shows that the level of 30 was never reached.

The highest Nile deposit near and on the Fayûm side of Lahûn is at about R.L. 26·00. The highest in the Fayûm near the Hawârah pyramid, which is on the edge of the Fayûm basin, is at R.L. 24·50 or thereabouts. The highest level of the lake was probably never more than one metre above this level, and it is therefore almost certain that the water-level was never sufficiently high to flow into the Wadi Raiân; and if it ever did, it must have been but rarely, when extraordinarily high and prolonged Niles occurred; so that it must be concluded, if my views are correct, that the normal condition of the Wadi Raiân was then, as now, that of a dry waterless depression in the desert, and it cannot therefore be considered as having been Lake Mœris, or a part of it even, at any time.

LINANT’S OBJECTIONS TO THEORY FAVOURED.

Author’s Views of Lake Mœris generally stated.—I myself agree with those who are of opinion that the Fayûm Province, or depression, (including the Gharaq Basin and the neck from Lahûn to Hawârah), was by itself Lake Mœris, and that within its limits and along its borders was to be found the inhabited and cultivated region known as the Arsinoïte Nome, which possibly also extended into the Nile Valley along the course of the canal connecting the Nile with the lake.

The Fayûm then in its submerged state was, I believe, the Lake Mœris of Herodotus, Strabo, Diodorus, and Pliny, the modern Lake Qurûn being the persistent rudiment of this lake, and all that now remains of its formerly extensive sheet of water. (See Plate XX.)

Objections urged by Linant against these views.—This is no new theory. It is found passim before Linant Pasha in 1842 took great pains to point out its absurdity, but it was his own assumptions regarding the maximum height which the water surface of Lake Qurûn could have reached, that created the absurdity. Assuming without evidence that the villages on the second plateau were all in existence at the time of Lake Mœris, he limits the level of Lake Qurûn to the edge of the second plateau, which is the same thing as laying down that its water surface never rose above R.L. + 10·00.

Having come to this conclusion, he might have spared himself all his arguments against the theory, other than that which pointed out that a reservoir in the Fayûm at this level could have been of no utility in supplementing the low waters of the Nile.

It is, however, instructive to note how he deals with the arguments against this lake, which his imagination set bounds to, being Lake Mœris. After a separate review of each condition which Lake Mœris should fulfil and which the limited Lake Qurûn did not, he closes his reviews with the remark that “we may then conclude that Birket-el-Qurûn is not the Lake Mœris.” But he does not do so always. Should the condition be one with which his own theory is not in agreement, he explains it away or discredits it. The dimensions assigned to Lake Mœris by the ancient historians evidently trouble him, and he does his best to discredit their testimony on this point. After discussing this condition, he does not end his argument with the usual conclusion that “the present Lake Qurûn cannot be Lake Mœris,” but he says “an absolute importance must not be attached to all these measures in order to draw from them conclusions either positive or negative as to the identity of the position of Birket-el-Qurûn with that of the ancient Lake Mœris.”

The depth assigned to Lake Mœris also gives rise to the following remarks, which will afford the means of judging of the value of M. Linant’s arguments. He states that Herodotus gives the depth of the lake at 92 metres, and remarks that if the whole Fayûm had been filled to form the lake, its dimensions would have surpassed by ten times the greatest given for it.

But as a matter of fact, they do not even come up to the greatest dimensions given, which are, for the depth 92 metres, and for the perimeter of the lake 720 kilometres (450 miles), or, assuming as he does that Herodotus made use of the small stadius, 360 kilometres. Now the perimeter of the Fayûm is 220 kilometres, and if that of the Wadi Raiân is added, namely 200 kilometres, the total perimeter becomes 420 and that figure is only obtained by measuring the indentations of the Wadi Raiân, which is of a peculiar shape.

The depth of the Fayûm Lake, if filled to say R.L. + 25·00, would be not less than (25·00 + 43·50 + 5·00 =) 73·50 metres, nor more than 88·00 at the highest estimate.

These dimensions agree approximately with those given by Herodotus, and are not, as rashly stated by Linant, ten times in excess.

To show with what unfairness Linant deals with statements made by Herodotus, his arguments about the bricks made for the pyramid built by Asychis may be noted. It was stated that the bricks were made from mud brought up from the bottom of the lake. Linant claims this statement as supporting his theory, as his lake was a shallow one, and as opposing the enlarged Lake Qurûn theory, as this latter would be a deep one. It does not seem to strike him that the workmen could have sought their mud along the shallow margins of the larger lake. He further argues that one could not reach down more than 4 metres with poles, and therefore the lake could not have been so deep as stated by Herodotus, and hence Herodotus contradicts himself! But Herodotus did not say that the lake was 92 metres deep all over, and that its shores were not shallow, but that its greatest depth was 92 metres.

Linant Pasha discussed the possibility of the submerged Fayûm being the Lake Mœris, but rejected the idea, because, to fulfil the condition of supplementing the low Nile, the water must have covered the second plateau, and risen to a level above the rock sill at Hawârah (R.L. 21·00). “Then,” he points out, “the whole Fayûm would have been only a vast lake and with a height of water impossible to reconcile with the existence of large towns, which formed the rich Crocodilopolite or Arsinoïte nomes. The great quantity of ruined towns, abandoned like Medinet-el-Mahdi, Medinet-el-Hêb, Medinet Nemroud, Kasr Keroun, indicate, as well as those which still exist, as Sanuris, Sanhur and all the others, that this part has never been under water, and they date from the time of Lake Mœris and of Crocodilopolis.” (Earlier in his book he states about Kasr Keroun, “Kasr Keroun is a little monument, quite modern as compared with the epoch of the Labyrinth.” Mr. Petrie and Dr. Schweinfurth both state that Qasr Qurûn is a Roman temple or town.) M. Linant continues, “If ever the Fayûm has been under water, as we have supposed it, it was long before it was habitable and before the Lake Mœris existed.”

Now as regards the modern villages of Sanûris, Sanhûr, and others, I am not aware on what evidence M. Linant states that they existed at the same time as Lake Mœris. As regards the old abandoned towns mentioned, some of them are on elevated spots, and probably were on the shores of Lake Mœris. When Lake Mœris declined and the water had receded to a distance from them, they were abandoned for more favourable sites, less remote from a water supply and water transport. Probably Sanûris and Sanhûr, and the other villages on the edge of the second plateau, are the successors in time of the ancient elevated towns mentioned as ruined and abandoned.

Thus, instead of considering the remains of the old high-level abandoned towns as evidence destructive of the theory that the whole Fayûm was filled with water, I consider their testimony distinctly favours such a theory.

Those towns especially, whose ruins are found on the north side of Lake Qurûn, would certainly have been built near the then borders of the lake, as they could have had no possible source of water supply other than the lake itself. It, therefore, is a matter of great interest to determine the levels of any ancient towns that may be found on the north of the lake; and the more ancient the town and the more remote from the present lake, the more suggestive will be the facts that may be ascertained with reference to its levels.

Now there are two monuments of antiquity known in such a situation, namely, the ruins of Dimay (Dimeh or Dimé) and an ancient temple (if it is a temple) discovered by Dr. Schweinfurth 7 or 8 kilometres north of Dimay.[6] Dimay itself is 3 kilometres from the nearest point of the present Lake Qurûn, and the surface of its causeway or quay at its upper end, near the old town, is 69 metres above the water surface of the lake (May 2nd, 1892), or at R.L. + 25·44. The south end of the quay is now about 2·85 metres lower, but it was doubtless originally somewhat higher than this, as its present is not apparently its original surface, some of the layers of stone having disappeared.

I had a trench dug against this quay or causeway, at about the middle of its length, to determine the depth to which the masonry was carried down.

If this had been merely a causeway, it is not easy to understand the necessity for so great a depth of masonry. It was therefore more probably a quay projecting into the water. This quay is 400 metres long, and its direction is due north and south. The level of the plateau sloping up to the end of the causeway on the south of Dimay is from R.L. 13·00 to 17·00; the plateau on the north side of the ruins is at R.L. 21·45.

The ruins of Dimay are Roman on the surface, but I do not know if it has been established that below the Roman remains there do not exist more ancient ones. Dr. Schweinfurth thus expresses his opinion about this old town: “Dimé seems to have filled the position of ‘tête-de-pont’ in relation to the Fayûm, as in consequence of its strong position, it afforded a secure outlet and final station for the caravan road opening out towards the Oasis. That the tribes of the Libyan Desert must even in the times of the Romans have been very restless and enterprising, is testified by the numerous similar fortifications, which in the day of the so-called good emperor, were erected on all the principal exits and entrances to the Oasis roads.”

Seven or eight kilometres north of Dimay (magnetic bearing from north entrance in Dimay enclosure wall 12° east of north) is found the ancient so-called temple, discovered by Dr. Schweinfurth in 1884. I give here photographs of the exterior and interior of the building, as well as its ground plan, that those who are capable of judging may have the means of estimating from them its probable age. (Plates XIV., XV., and XVI.) The important level, so far as the subject of this paper is concerned, of the old town, marked by mounds of ancient pottery on the south of the “temple,” was determined on the occasion of my visit. The level of the upper parts of these mounds was found to be R.L. + 24·58. The pottery was, of course, spread out to lower levels, but probably the ancient town was built between the levels of 23·00 and 26·00.

Dr. Schweinfurth remarks that the buildings dating from the XIIIth Dynasty are all distinguished by the same kind of four-cornered arrangement as this temple and generally scorn every kind of ornament; and he notes that the great size of the blocks and peculiar method of fitting the stones together give it a resemblance in style to other old buildings. Instead of giving his further description, I refer to the ground plan and photographs. It is worth noticing that the north-west room has no visible means of communication either with the exterior or with the other chambers of the building; also that the displacement of the stones, forming the upper half of the chambers on the raised floor, is suggestive of an earthquake, the upper stones having slid on the lower to a measurable extent in a north-easterly direction. Cracks in the roofing stones corresponding with the displacement seem to confirm the theory of earthquake action.

The object of the building is a riddle. Each of the raised cells has a recess for a door. “In the thickness of the south wall, on the east side of the principal entrance, runs a passage half a metre wide, leading to which, at the south-east corner of the temple, a door of equally narrow proportions is attached. This passage leads downwards to the chambers below” (Schweinfurth). Dr. Schweinfurth came to the conclusion “that the old temple, as well as the original settlement or formation, is one of the monuments belonging to the oldest times.”

Concerning the old town he writes:—“In the neighbourhood of the temple, from south-east to south-west, at a distance of about 500 paces, that is, on the edge of the rising ground, there are quantities of potsherds lying in heaps here and there. They are of the most weather-worn appearance, and have formed portions of coarse, thick vessels. No such things as fragments approaching the pottery work of the Greek or Roman period are found. The eye of the seeker sought in vain for remnants of that blue glazed pottery ordinarily so common, or the long amphoræ of the Greek shape.

“The amphoræ points or ends, which I picked up, were all stumps, and of an almost cylindrical shape. The corresponding pottery showed no sign of rings. They were almost entirely coarse, red clay fragments, with here and there a yellow or black bit, and all distinctly showed the work of the potter’s wheel.

“Below the scarp of the lowest rising ground no more pottery was to be found, neither did the marl mounds display on examination any admixture of manufactured pieces. The heaps of pottery formerly existing appear to have been flattened down and spread out over a much wider space by the disintegration and sweeping down of the marl bed. A similar occurrence may be observed on the few stone walls yet remaining of the old temple settlement.”

For Schweinfurth’s further remarks see pp. 101 to 107 of his veloci-graphed letter to Paul Ascherson on his journey in the depression of the Fayûm, 1886.

The line of levels, which I had taken between this old building (temple) and Dimay, followed a direct line between the two, crossing the elevations and depressions given in the list below:—

These levels I am convinced are correct, as they were taken with the utmost care, as I myself saw, by Messrs. W. O. Joseph and A. Pini, who had been in constant practice at levelling. The levels between Lake Qurûn and Dimay were taken twice over; the first levels, taken by Monsieur Pini alone, giving a difference of level between the lake and causeway of 68·952 metres, while the difference found, when both read, was 68·978.

The levels for the old town near Schweinfurth’s “temple” having been found to be from R.L. 23 to 26, the theory that Lake Mœris was a little below the level of R.L. 23·00 is favoured by the determination of this level. The presence and peculiarities of the quay at Dimay, if it is such, and the existence of an old town on the heights where the Dimay ruins stand, if they can be used as evidence of what the lake level used to be, point to high levels rather than to low ones, and do not answer to Linant’s appeal to the old abandoned towns to bear witness in his favour.

In connection with the levels of Dimay and Schweinfurth’s “temple” the levels of the ruins of Biahmu should be studied. These are given on Plates XXII. and XXIII.

The top of the highest corner-stone of the enclosure wall, now in situ, is at R.L. 21·59, and, accepting Mr. Petrie’s restoration of these ruins, the top of this wall, when complete, would have been at R.L. 23·00, which would seem to indicate that the maximum water-level of the lake was below R.L. 23·00, but higher than R.L. 17·00, the level of the ground outside the enclosure. These ruins are referred to more fully on p. 83 et seq.

Mr. Flinders Petrie’s Views of Lake Mœris.—Having discussed the theories of Linant Pasha and Mr. Cope Whitehouse, the only two that I can find stated with any distinctness, and the only ones that have been put forward by travellers having a personal acquaintance with the Fayûm, I will, before setting forth my own reading of the past history of the province and my theory as to its connection or identity with Lake Mœris, first give Mr. Petrie’s views, who should be included with the two foregoing theorists, as a traveller having a personal acquaintance with the Fayûm, and, in a special line, a very intimate one. I do not think that he would claim that the expression of his views constitutes the enunciation of a new theory of Lake Mœris, but only his way of viewing an old theory with some side-lights of his own added by way of illumination.

The views, that I have adopted, are in general agreement with those favoured by Mr. Petrie, and as he, an Egyptologist and archæologist, has thrown light on the subject from his standpoint, I propose to make the same attempt from my point of view as the Public Works officer in charge of the irrigation of the Fayûm. The working out of the problem of Lake Mœris would seem to require an alliance between a palæontologist, an archæologist, an Egyptologist, a geologist, and a hydraulic engineer.

The following is copied from ‘Hawara, Biahmu, and Arsinoë,’ by W. M. Flinders Petrie, published in 1889:—

“Medinet el Fayûm (Plate XVII.) is the modern town which represents the ancient Arsinoë, so named by Ptolemy Philadelphos in honour of his sister-wife; it lies at the extreme south of the old site, which covers a space of over a mile long and half a mile wide, a vast wilderness of mounds strewn with pottery. At the opposite end of the ruins, toward the north, is the great temple enclosure of the old Egyptian town. Before its name of Arsinoë, the city had obtained the name of Crocodilopolis, from the worship of the sacred crocodiles maintained there; and still earlier it was known as Shed, meaning, apparently, that which is saved, cut out, delivered, or extracted, referring to the district being reclaimed from the great lake. The whole province was known as Ta-she, ‘the land of the lake’; and, whatever may have been the mistakes of historians about Lake Moiris, there is no doubt that the lake was the main feature of the district.