[120] Chapter iv. pp. 396-400, Vol. I.

[Ill.u.s.tration: FIG. 108.--Part plan of the temple at Elephantine.]

[Ill.u.s.tration: FIG. 109.--Luxor, plan of the second court.]

In the cases where the portico is within the courts, it is sometimes confined to two sides, as at Luxor (Fig. 109); the columns shown at the top of our plan belong to the p.r.o.naos and not to the court. In the Temple of Khons it surrounds three sides (Fig. 110), while the fine court added to the temple of Luxor by Rameses II. has a double colonnade all round it (Fig. 111).

Both in the interior of the halls and in the external porticos we find an apparently capricious irregularity in s.p.a.cing the columns.

Sometimes intercolumniations vary at points where we should expect uniformity, as in the outer court of Luxor (Fig. 112). On two of the faces the columns are farther apart than on the other two. The difference is not easily seen on the ordinary small plans, but it is conspicuous in the large one of the _Description_.[121]

[121] _Description de l"egypte_, plates, vol. iii. pl. 5.

It is easy to understand why the s.p.a.cing should have been increased in front of a door, an arrangement which exists at Gournah (Fig. 113), and at Luxor (Figs. 109 and 111).

[Ill.u.s.tration: FIG. 110.--Portico in the Temple of Khons.]

[Ill.u.s.tration: FIG. 111.--Luxor, portico of the first court.]

In the hypostyle halls we find columns of different sizes and orders.

Six of the great columns which form the central avenue at Karnak cover as much ground, measuring from the first to the sixth, as nine of the smaller pillars. Between supports so arranged and proportioned no constant relation could be established (Fig. 114). The transverse lines pa.s.sing through the centres of each pair of great columns correspond to the centres neither of the smaller shafts nor of the s.p.a.ces which divide them. The central aisle and the two lateral groves of stone might have been the creations of separate architects, working without communication with one another and without any desire to make their proportions seem the result of one coherent idea.

In the inner hypostyle hall at Abydos the intercolumniations which lead respectively to the seven sanctuaries vary in width (Fig. 115).

This variation is not shown by Mariette, from whose work our plan of the temple as a whole was taken, but it is clearly seen in the plan given in the _Description_. These are not the only instances in which those early explorers of Egypt excelled their successors in minute accuracy.

[Ill.u.s.tration: FIG. 112.--Part of the portico of the first court, Luxor. From the _Description_, iii. 5.]

[Ill.u.s.tration: FIG. 113.--Portico in front of the facade of the temple of Gournah. From the _Description_, ii. 41.]

Here and there we find the s.p.a.ces in a single row of columns increasing progressively from the two ends to the centre (Fig. 105).

[Ill.u.s.tration: FIG. 114.--Part of the Hypostyle Hall in the Great Temple at Karnak.]

The combination of quadrangular with Osiride piers and of the latter with columns proper was also productive of great variety. In the speos of Gherf-Hossein six Osiride piers are inclosed by six of quadrangular section (Fig. 116). In the first court at Medinet-Abou a row of Osiride piers faces a row of columns (Fig. 117), while in the second court there is a much more complicated arrangement. The lateral walls of the court are prefaced each by a row of columns. The wall next the entrance has a row of Osiride piers before it; while that through which the p.r.o.naos is gained has a portico supported by, first, a row of Osiride piers, and, behind them, by a row of columns (Fig. 118).

[Ill.u.s.tration: FIG. 115.--Second Hypostyle Hall in the temple of Abydos. _Description_, iv. 36.]

[Ill.u.s.tration: FIG. 116.--Hall in the speos of Gherf-Hossein (from Prisse).]

[Ill.u.s.tration: FIG. 117.--Medinet-Abou; first court.]

[Ill.u.s.tration: FIG. 118.--Medinet-Abou; second court.]

In the temple of Khons the peristyle is continued past the doorway in the pylon (Fig. 119), and the inclosure is reached through one of the intercolumniations.[122] At Luxor, on the other hand, the portico was brought to an abrupt termination against the salient jambs of the doorway (Fig. 120).

[122] This is a mistake. By a reference to Fig. 208, Vol. I., or to Fig. 126 in this volume, it will be seen that the peristyle was not continued along the inner face of the pylon.--ED.

[Ill.u.s.tration: FIG. 119.--Portico of the Temple of Khons, looking towards p.r.o.naos.]

[Ill.u.s.tration: FIG. 120.--Portico of first court at Luxor.]

The Egyptian architect, like his Greek successor, made frequent use of the _anta_, that is, he gave a salience to the extremities of his walls which strengthened his design and afforded structural members, akin to pilasters or quadrangular pillars, which were combined in various ways with columns and piers. Sometimes the anta is nothing but a slight prolongation of a wall beyond the point where it meets another (Fig. 121); sometimes it is the commencement of a returning wall which appears to have been broken off to give place to a row of columns (Fig. 122); a good instance of the latter arrangement is to be found on the facade of the temple at Gournah. Sometimes, as at Medinet-Abou, it is a reinforcement to the extremity of a wall, and serves to form a backing for colossal Osiride statues (Fig. 123), sometimes it gives accent and strength to an angle, as in the Great Hall at Karnak (Fig. 124). At the Temple of Khons the terminations of the two rows of columns which form the portico are marked by antae on the inner face of the pylon (Fig. 126), while the wall which incloses the p.r.o.naos is without any projection except the jambs of the door.

This arrangement has an obvious _raison d"etre_; if the columns were brought close up to the pylon their outlines would not combine happily with its inclined walls. At the other extremity of the court, the wall being perpendicular, there was no necessity for such an arrangement.[123] A glance at Fig. 126 will make this readily understood. At Medinet-Abou the portico is terminated laterally by two antae, one corresponding to the row of columns, the other to the row of caryatid piers. In another court of the same temple the antae on either side vary in depth, at one end of the portico there is a bold pilaster, at the other one which projects very slightly indeed (Fig.

128). This is another instance of the curious want of symmetry and regularity which is one of the most constant characteristics of Egyptian architecture.

[123] The arrangement in question is capable of another and, perhaps, more simple explanation. The two rows of columns of which the portico in question is composed, run in an unbroken line round the court with the exception of the side which is filled by the pylon. It was natural enough, therefore, that they should each be stopped against an anta, even if there had not been an additional reason in the inclination of the pylon. The ordonnance as a whole may be compared to a long portico, like that in the second court of the temple at Gournah, bent into two right angles.--ED.

[Ill.u.s.tration: FIG. 121.--Anta, Luxor; second court. _Description_, iii. 5.]

[Ill.u.s.tration: FIG. 122.--Anta, Gournah. From Gailhabaud.]

[Ill.u.s.tration: FIG. 123.--Anta, Medinet-Abou.]

[Ill.u.s.tration: FIG. 124.--Anta in the Great Hall of Karnak.]

[Ill.u.s.tration: FIG. 125.--Antae, Temple of Khons. _Description_, iii.

54.]

[Ill.u.s.tration: FIG. 126.--Anta and base of pylon, Temple of Khons.

_Description_, iii. 55.]

The anta is often without a capital, as, for instance, in the temple of Khons (Fig. 126). Elsewhere the architect seems to have wished to bring it into more complete harmony with the magnificence of its surroundings, and accordingly he gives it a capital, as at Medinet-Abou, but a capital totally unlike those proper to the column.[124] It was identical in form with that gorge or cornice which crowns nearly every Egyptian wall. Considering that the anta was really no more than a prolongation or momentary salience of the wall, such an arrangement was judicious in every way (Fig. 129).

[124] In this the Greek architects took the same course as those of Egypt.

[Ill.u.s.tration: FIG. 127.--Antae, Medinet-Abou.]

[Ill.u.s.tration: FIG. 128.--Antae, Medinet-Abou.]

The width of the intercolumniations also varied between one court or hall and another, and, at least in the present state of the Egyptian remains, we are unable to discover any rule governing the matter, such as those by which Greek architects were guided. We may affirm generally that the Egyptian constructor, especially in the time of the New Empire and when using columns of large dimensions, preferred close s.p.a.cing to wide. His tendency to crowd his columns is to be explained, partly by the great weight of the superstructure which they had to support, partly by the national taste for a ma.s.sive and close architecture. The s.p.a.ces between the great columns in the hypostyle hall of Karnak, measured between the points of junction between the bases and the shafts, is slightly less than two diameters. The s.p.a.ces between the smaller columns on each side are hardly more than one diameter.

A better idea of the original character of these ordonnances may perhaps be gathered from the plate which faces the next page (Pl.

VIII) than to any plan to which we could refer the reader. It represents that part of the colonnade, in the second court of the temple at Medinet-Abou, which veils the wall of the p.r.o.naos, and it shows how little s.p.a.ce the Egyptian architects thought necessary for the purposes of circulation. The s.p.a.ces between the columns and the wall on the one hand and the osiride piers on the other, are not quite equal to the diameter of the bases of those columns, which have, however, been expressly kept smaller than was usual in Egypt. If they had been as large as some that we could point out, there would have been no room to pa.s.s between them and the wall.

Did the Egyptians ever employ isolated columns, not as structural units, but for decorative purposes, for the support of a group or a statue? Are there any examples of pillars like those which the Phnicians raised before their temples, or the triumphal columns of the Romans, or those reared for commemorative purposes in Paris and other cities of Modern Europe? It is impossible to give a confident answer to this question. The remains of the great colonnade which existed in the first court at Karnak, of which a single column with bell-shaped capital is still upright (Fig. 130), suggest, perhaps, that such monumental pillars were not unknown to the Egyptians. These columns display the ovals of Tahraka, of Psemethek, and of Ptolemy Philopator. The width of the avenue between them, measuring from centre to centre, is so great, about fifty-five feet, that it is difficult to believe that it could ever have been covered with a roof.

Even with wood it would have been no easy matter--for the Egyptians--to cover such a void. We have, moreover, good reason to believe that they never used wood and stone together in their temples.

A _velarium_ has been suggested, but there is nothing either in the Egyptian texts or in their wall paintings to hint at their use of such a covering.

It would have been quite possible to connect the summits of these columns together lengthwise. The architraves would have had less than twenty feet to bridge over. But not the slightest relic of such a structure has been found, and it is difficult to see what good purpose it could have served had it existed.

The authors of the _Description_ came to the conclusion that there had been no roof of any kind to the avenue formed by the columns, that they merely formed a kind of monumental approach to the hypostyle hall.[125] Mariette also discards the idea of architraves, which would have to be unusually long, but he cannot accept the notion that the columns were merely colossal venetian masts bordering the approach to the sanctuary. He supposes the centre of the courtyard to have contained a small hypaethral temple built by Tahraka. This temple figures upon his plan, but neither he himself, by his own confession, nor any one else has ever found the slightest trace of it in reality.[126] In the excavations made by him in 1859, he did not find a vestige even of the two columns which he inserts upon each of the two short sides of the rectangle. These columns were necessary in order to complete a peripteral arrangement, similar to that which exists in the hypaethral temples at Philae and in Nubia. The closest study of the site has brought to light nothing beyond the twelve columns shown in our plan (Fig. 214, E, Vol. I.).

[125] _Description, Antiquites_, vol. v. pp. 120, 121. In their _Description Generale de Thebes_ (ch. ix. section 8, -- 2), the same writers add: "We are confirmed in our opinion by the discovery on a bas-relief of four lotus stems with their flowers surmounted by hawks and statues, and placed exactly in the same fashion as the columns which we have just described. They are votive columns. We are also confirmed in this opinion by the fact that we find things like them among those amulets which reproduce the various objects in the temples in small." This bas-relief is figured in the third volume of plates of the _Description_, pl. 33, Fig. 1.

[126] MARIETTE, _Karnak_, p. 19, pl. 4. _Voyage dans la Haute-egypte_, pp. 13, 21, 22.

[Ill.u.s.tration: Ch. Chipiez del Hibon sc.

THEBES