p1 "London has long possessed the finest collection of both the larger and smaller works of art from Greece and Asia Minor, but Lord Elgin could not carry off Homer's sun, rocks, and seas." — Baedeker's "Greece."
It is impossible to understand any architecture from books alone, and this may be even especially true of the great works of Greek art, for they are not so much seen immediately as through a veil of traditional explanations, commentaries, and theories which are probably in great part a formal grammar applied long after the time when the architecture flourished as a living language. The British Museum, which is the richest collection of representative fragments of great classical buildings in the world, furnishes us with an invaluable means for looking directly at, and measuring the very stones wrought by Greek artists. In it are stored large and significant fragments of the Parthenon, the Erechtheum, the Propylaea, the Temple of Niké Apteros, and of that which once stood by the Ilissus — all in Athens; of the great Temple of Diana at Ephesus, and of the Mausoleum of Halicarnassus, the two most famous buildings of Asia Minor; also of the Temples of Bassae and Priene, and of several important tombs, from the prehistoric work of Mycenae to the late Nereid monument brought from Xanthus.
Of these, the Temple of Diana and the Mausoleum can only be properly studied in the Museum, which contains practically all the wrought stones of them which have ever been discovered. I propose first to examine the Temple of Diana.
The several phases of architecture are usually classified as Antique and Classical, or Mediaeval and Gothic. We are apt p2to accept the idea that there is some opposition in these, but this is not necessarily the case, although, of course, there is diversity. In a large degree all architecture is one, in that it is the work of men shaping materials according to their powers and desires; Greek and Gothic architecture resemble one another in both being what I may call primary styles; while Roman art in much, and Renaissance art still more, were consciously derived, and are secondary.
Before the discovery by Mr Wood of the long-buried site, the Temple of Diana was chiefly known by its reputation as one of the Seven Wonders of the world, and from a few short notices by ancient writers. According to Vitruvius,a it was Ionic, dipteral, octastyle, and had a cedar ceiling. Plinyb says that it was of the enormous and improbable size of •425 feet by 220 feet, that it had 127 columns, the gifts of kings, and thirty-six which were sculptured.
Falkener, by the publication of his work on Ephesus in 1862, in which he brought together many of the references to the temple contained in the ancient books and offered a conjectural restoration, must have generated the interest which led to a search for the site being undertaken in the following year by J. T. Wood. It was not until 1870 that the site was identified, and as the plain on which the temple was built had been covered by some 15 feet of alluvial deposit, and not one stone remained above-ground, the discovery was a triumph for what long seemed a forlorn hope. He was helped by an inscription now in the British Museum, which shows that the temple was outside the Magnesian Gate.
In 1877 Mr Wood published his popular account, restoring the plan from indications found, as having eight columns at the ends, twenty on the flanks of the outer row, and one hundred in all. In the accounts collected by Falkener the temple is said to have been rebuilt many times, but in the main remnants of only an earlier and a later building were discovered, and I shall call them the Old and New Temples. Wood speaks of the "last temple," "the last temple but one," and "the last temple p3but two," but the only remains he assigned to the middle one were those of a pavement intermediate between the higher and the lower levels. As the site was very low, a raising of the floor was quite likely to have taken place, and we need not infer from it a rebuilding of the temple. The site was practically a marsh, as had been remarked by Pliny. The Austrian survey referred to below shows the floor of the Old Temple 2.70 metres above the sea-level, the New Temple 5.42 metres, and the modern surface 9 metres (Fig. 1).
The remains of the Old Temple were of mid-sixth century work, and, doubtless, belonged to the temple to which Croesus contributed. The later temple was of the fourth century, and probably was not completed until after the visit of Alexander to Ephesus in 334. We may call them the temples of Croesus and of Alexander. The site was outside the city to the east at the foot of a range of hills which rose near the back of the east front. The west portico faced the city and the harbour, and possibly for this reason became the chief entrance contrary to the usual custom. It must have been the nature of the site which led to the elevation of the New Temple upon a platform reached by many steps. In building the new structure the old work was only taken down to below the level of this platform; even the bases and stumps of the old columns were left. These were built about with new foundations, and the cella walls were increased at the sides. The new foundations thus contained a core of the old building, and both were discovered p4together. The two temples were identical in size and general disposition of parts.
In 1884 Ferguson worked over Wood's materials, which had been inadequately published, with the object of showing that places will be found for 127 columns as mentioned by Pliny; and, for this purpose, extended the plan by two or three bays. He had already suggested, from an examination of the marbles in the British Museum soon after they were received, that the square sculptured blocks which Wood had thought were parts of the frieze formed pedestals for the sculptured columns (Fig. 3). As there were more than four angle pieces among these blocks it was shown that they could not have belonged to the frieze. The best result of this paper was that it led to a reply from Mr Wood in which he gave additional and much more workmanlike data with a plan of what was actually found, and sections of the steps and platform (Figs. 2 and 14). This plan also contained facts obtained in a further examination of the site made in 1883‑4.
The walls of the cella were here completely traced, with the basis for the great image in the midst. The foundation of one of the antae and of one column on each side are accurately laid down. Further, large portions of great retaining walls which supported the platform are shown on both sides, with cross walls exactly opposite the columns dividing the platform into a network of walls (it was so also at Pergamos), and giving a columniation of 17 feet 1¼ inch along the flanks. At one end, however, two bays of 19 feet 4 inches were found, and at the other end, and in the right place in regard to the antae, the first of a similar pair of bays was found. It was evident that the two bays at each end were made wider in preparation for the very wide columniation of the fronts. Long portions of the bottom step of those which surrounded the platform were found in situ on one p5side and one end. The supplementary facts given in this paper were enough to prove that there were twenty columns on the flanks (outer row) and eight at each of the ends, and that a bottom step surrounded the outer row of columns at a distance of •about 40 feet from their centres, at least on the sides and one end, and presumably the other. Some valuable detailed sections were also given, but in these what was found and what was conjectural interpretation were not sufficiently distinguished (Fig. 2).
Working over the evidence has convinced me that the network of walls spoken of supported a raised platform reached by continuous flights of steps, surrounding the colonnade.
Wood says that he ascertained that the inner row of columns had no square plinths. p6"I found," he says, "a few inches of the front of the step which was here substituted for the plinth. The step under the base was also in two pieces."
I here add a few points derived from the volume of 1877. The marble pavement of the Old Temple was about 7 feet 6 inches below the level of the peristyle, which itself was about 9 feet 6 inches above the court. The plinths of the first temple were 7 feet 8½ inches square, those of the New Temple were 8 feet 8 inches. "The masonry which supported the steps, with the piers which united the masonry with the foundation piers of the columns, was of courses of limestone about 8 inches thick, which was about the height of each step" (p191). The square sculptured blocks were found at the west front (Fig. 3). Fragments of marble roof-tiles were discovered, which showed that they must have been of large size, as the rounded covering pieces were about 10 inches wide. A fragment of an acroterion was also found at one end. The temple rose in the middle of a paved court, surrounded by a stoa or colonnade. In one place a portion of the pavement remained against the outer step. In November 1873, a plinth of the stoa was found on the south side •nearly 31 feet beyond the lowest step; it was •25 feet wide. •Seventy feet away from the steps on the south side lay a Doric building parallel to the stoa, parts of four of its columns were found and of the wall beyond. The columns were •2 feet 6 inches in diameter, and 20 feet 6 inches apart, opposite the alternate piers of the stoa, which were marble and square. Still beyond was a high and strong parabolus wall enclosing the whole precinct.
Some time before 1895, when the late Dr Murray read a paper on the subject before the Institute of Architects, an essay at piecing the fragments together had been made by him at the British Museum. His main points were the following:—
He finally demonstrated that the sculptured blocks certainly made pedestals, each composed of four stones.
p8 Pointing out that a circular line on their upper surface would just take one of the sculptured drums with its roll moulding at the bottom, he endeavoured to prove that this arrangement, which had been suggested by Fergusson and adopted at the Museum, was the only possible one. "So far as I can see, there is no escape, much as it may be desired, from bringing the sculptured drums directly down on the sculptured pedestals as we have done in the Museum."
He maintained that a low square block made up of two stones found beneath one of the ordinary columns was cut so as to show that it "joggled" into other stones, and must have been part of a stylobate, instead of, as Wood had said, a plinth block under the base. "The [circular] base itself" (Dr Murray says), "directly above the joggles, has been in a careful manner cut into as if to receive a metal railing." According to Wood, he adds, this was one of the inner bases, but as a railing here would serve no purpose, Dr Murray thought that the base must have belonged to the outer row (Figs. 4, 12, 13).
Dr Murray took over Wood's general plan, which gave a hundred columns, eighteen at each end, making up the thirty-six sculptured columns of Pliny. He proposed, however, a great modification of the steps and platform, in p9order to find a place for the sculptured pedestals at a lower level, beneath the sculptured drums. For this purpose he pushed the steps into the porticoes behind the pedestals at the two ends, and, doing away with steps altogether on the p10flanks, he here substituted a continuous stylobate (Figs. 5, 6, 7, 11).
One of the drums is in bolder relief, and is, he says, •3 inches more in diameter than the rest. He placed it at the angle, and said that the increase of diameter "implies a proportionate increase of height." That is in the drum itself, I suppose.
In the spring of 1905, Mr D. G. Hogarth made a further examination of the site, which is the property of the British Museum, and gave some account of the results obtained in the Times for 8th August of that year. He agrees that the Croesus temple was exactly like its successor in size and arrangement. Below its level were found remains of a still earlier temple, small, in three divisions, and without any peristyle; in the centre was the basis of the cult statue. The site and the fragments found were carefully measured by Mr Henderson. It is most desirable that the publication of these should not be delayed.
In 1906 was published a most careful Austrian survey of the site by the Vienna Institute. This confirms Wood's plans as to the position of the fixed points laid down on them, but revises them by sweeping away the cross walls between the antae and the door, and in carrying the columns right back in the deep pronaos as at Miletus. This plan shows the retaining wall with cross walls to carry the platform, and, as described by Wood, the foundations of two more columns at the south-west angle; also indications of the foundation of one of the columns in antis (Fig. 7). It confirms Wood as to the position of the base now in the British Museum. The great foundation, •about 20 feet square, for the statue is shown exactly in the middle of the cella. A restoration is given of the west door, the jambs of which were about •3 feet 6 inches wide, and inclined inwards. An excavation was made directly to the west of the English site on the axis of the temple, and several inscriptions and tombs were found.
p11 The Austrian Survey concerned itself principally with the archaic temple of the sixth century. Of the Hellenistic temple of the fourth, the authors report that only one stone still remains in situ; "its form, however, could not have deviated in any important way from that of the Old Temple. One will not be able to proceed thoroughly to reconstruct it, following the suggestions of Murray, and the few of its architectural members which are to be found on the spot, until all the materials gained in Wood's excavations have been properly drawn and published." I wish I could make good this want, but for us, fortunately, the actual stones are published in the Ephesus Gallery of the British Museum.
Chersiphron, the architect mentioned by Pliny, was one of the architects of the sixth century temple, and when the writer says that so many columns were the gift of kings, it is probable that this also applied to the Old Temple, as we know that columns were given to it by Croesus, and the middle of the sixth century p12was more the age of kings than with the fourth. Its reputation as a world's wonder must go back to the earlier date when it was more without rival than at any other time. Herodotus, speaking of the labyrinth of Egypt, says that all the works of the Greeks cost less than it, "though the temple in Ephesus is deserving of mention, and also that in Samos." It is probable, therefore, that Pliny's whole account was a tradition as to the Old Temple. As we have seen, it was exactly the same size as the New Temple; like it, it had architraves of great span and columns with sculptured drums and others with square blocks.
The old cella walls, Wood says, were 6.4 thick, of marble, and remarkable for their finish; one block is shown at the Museum (the masonry added at the sides for the New Temple increased the foundations to •10 feet wide). The columns were •about 5 feet greatest diameter; the shafts above the sculptured drums were probably monolithic, and one of the purposes of these drums was to shorten the length of the single stones. There are restored capitals and bases at the Museum; the abacus is very long and narrow, not square as in later capitals; it is thus a corbel-capital. Amongst the débris of the temple some ox heads were found. Dr Murray, writing to the authors of the volume on Miletus, said, "they seem to me to belong in some way to capital of the Old Temple." The square attachments suggest rather to me that they were engaged in square piers, such as the antae. See Fig. 8, which, however, is doubtful, p13as I saw the fragment only for two minutes in the Museum cellars. Another fragment is a ram's head, catalogued as belonging to the New Temple, but of similar work (Fig. 9). The bases had various profiles in detail while agreeing in the larger parts. Dr Murray, giving no plinth to the new order, has withheld this feature from the old order too, but Wood found one in place under a base, and the Austrian account gives its depth as 0.33 metre — about a foot. The two stones, side by side, of which it was formed, were joined by dovetailed metal cramps (Fig. 7). The most remarkable feature is the gutter front so ingenious restored by Dr Murray. It is so big as to be a parapet, slightly inclined to the vertical plane, rather than a cymatium, and the surface is sculptured like a frieze between the lion's-head spouts. There cannot have been another frieze together with such a feature, and the cornice probably approximated to the form of Fig. 10. As mentioned below, the archaic temple was fully decorated with painting on the architectural members and sculpture. This building is of extraordinary historical interest, as in many particulars it set the type for this fully developed Ionic order — or perhaps it would be better to say Ionian order when speaking of the works found on Asiatic soil.
The weight of Dr Murray's case, summarised above, rests on the assumption that the sculptured drums must have rested on the sculptured pedestals, because on the top of the latter are traces of a circular setting-line suitable in size. When we realise that the ordinary fluted columns are of exactly the same diameter, and have an exactly similar torus moulding at the bottom, and ask why they may not have stood on the sculptured pedestals, we open up an alternative solution which is simpler, and, I think, more likely on its merits, apart from the fact that p14the former seems to be impossible having regard to the evidence. Dr Murray discussed and dismissed the possibility of fluted columns with deep bases standing on the pedestals, but did not consider the obvious possibility of the same columns with the tours only, being placed exactly like the drums.
Choisy, in adopting Dr Murray's scheme, allows that the disposition was unusual to the Greeks. I must think that such unnecessary complication was foreign to their ideals, and that the irregularity on the flanks would have been most marked and awkward in having three varieties of bases in succession (Figs. 6, 11). It is very awkward also to have the steps sloping down against the sculptured pedestals without any flat space between. The primary idea of the platform seems, as we have seen, to have been to lift the temple above the marshy site; allowing the principal columns to start from the lower level is opposed to this. As only a part of the columns were sculptured, it seems unlikely that some of them would be doubly decorated, making, indeed, a total of fifty-two sculptured portions, which would have sufficed exactly to surround the whole temple. There seem to have been both square sculptured stones and drums in the ancient p15temple, and Dr Murray would bring those also together — "the old sculptured columns," he says, "may have rested on square sculptured pedestals." But the Old Temple was not raised on a high platform, and the different columns must have stood on the same level.
Other later Ionian temples stand on platforms reached by continuous steps. The Smintheium stood on a platform proportionately higher than that of Ephesus At Miletus was a temple which, in many respects, was a companion work to the Artemision. The full account of the latest excavations on this site, published in 1904, shows that it was almost the same size and had a continuous flight of steps entirely surrounding it. The columns at the two ends were different from the others in having ornamental bases (not drums).
p16 Coming now to the crucial objections to Dr Murray's scheme I would point out that it involves the implication that Wood, who explored the site for four or five years, misread plain evidence as to the steps and bases surrounding the temple. He himself appears to be mistaken in saying that Wood had stated that Museum base and plinth came from the inner row. In Wood's paper in the R.I.B.A Journal, this base is carefully drawn, and lettered "Foundation of Outer Column of Peristyle"; and in his book he gives an accurate lithograph of the base stones as found and set up in the British Museum, and describes them properly as from the south flank (Figs. 12, 14). This point is of little consequence, except that it relieves Wood of having made this mistake. Murray wanted the base to be on the outer row, and Wood had, in fact, stated that it came from that position. But the exigencies of the heights, to make it possible to get in the sculptured pedestals where Dr Murray suggested, required that he should get rid of any plinth blocks under these bases, and prove that what Wood had said was such a plinth should be part of a continuous stylobate (Fig. 13). Such a view can in no way be brought into harmony with Wood's detailed section, which shows a rough stone foundation pier under the bases projecting •2 feet beyond where Dr Murray's thesis requires a continuous fair marble wall face (Fig. 7). That this p17was a rough square pier projecting, as said, is confirmed by the accurate Austrian Survey, so that we may venture to say that there was no continuous marble stylobate such as suggested by Dr Murray in opposition to Wood's data. Secondly, the theory requires the sweeping away of the network of foundations which supported the steps and platform outside the columns, which are carefully shown on Wood's plan and sections (Figs. 2, 14), and confirmed by the Austrian Survey. Thirdly, Wood found the bottom step of those which surrounded the platform •about 40 feet out from the centres of the columns on the north and east sides. Dr Murray had no use for this, and it became a single step in the middle of the narrow space of court which surrounded the temple, and between it and the enclosing colonnade. Now, there are three portions of steps actually at the brums, and each one shows plainly that it was followed by at least another step. The only step that Wood speaks of finding is the outer one, and one of those at the Museum must be a part of it; probably the widest one, which, however, shows by a fixing line that only the normal size of •8½ inches by 22 inches was exposed. The others are •22 inches wide and deeper than 8½ inches, but the surplus was fitted into a notch in the step below each (Fig. 15).
p18 We are thus compelled by evidence of the building itself to look for another solution of the question of the distribution of the sculptured drums and pedestals. Measurement of the stones at the British Museum shows first, that the bottom of the fluted columns are exactly the same size as the sculptured drums (Wood says that both are •6 feet 1‑2 inch); secondly, that the pedestals and drums are of the same height •(6 feet); and thirdly, that the pedestals are about the same size across as the diameter of the drums (Fig. 16). I say about, because there is nothing which exactly fixes their size, and, as set up at the Museum, they vary •from 6 feet 1 inch to 6 feet 2 inches. I suggest, therefore, that outer row, at one end or both, had the square blocks ranging at the same level with the sculptured drums. I believe the blocks were all found at the west end.
On the supposition that we must provide for thirty-six sculptured columns we get rid of any which were doubly sculptured. As the returns would have square pedestals on the outside row as well as the fronts, we get only two varieties of bases on the flanks as shown by Fig. 11.
I have carefully examined the base at the British Museum, and can see no sign that the stones which Wood called a plinth made part of a continuous course. Except the one plain face in front all the rest is broken. Such internal data as can be gathered are all in favour of the blocks having formed a plinth. It was bolted to its foundation by four bolts, three of the holes for which remain, which Dr Murray read as a joggled joint. Such a bolt was, of course, put through the solid stone (Fig. 4).
These holes really conditioned the fracturing of the stone, or rather the bolts did, for it is clear that the plinth has been roughly yet purposely broken so that the metal might be extracted — a usual phenomenon at ancient ruins. The question of a railing as suggested by Dr Murray hardly touches the point, and I do not doubt that there may have been a railing, p19but the •8‑inch by 4‑inch cavity sunk about •10 inches into one side of the base seems to me rather a mending of the lower member than the fixing mark of such a railing (Fig. 13). Reasons for the stones having formed a plinth, besides these, and Wood's testimony, are the facts that the size, •about 8 feet 7 inches, works out to half a columniation as at Priene and Miletus, and the whole base resembles those of the companion works just named.
With the plinth the height of the base is two-thirds of a diameter, a fair proportion; but, without it, it is impossibly low. Of the columns the bottom diameter is •6 feet 1‑2 inch, and the top diameter is •4 feet 9½ inches, about four-fifths of the lower dimension. Wood noticed that the columns decreased rapidly towards the top, which is also a characteristic of the columns of the Mausoleum. The distance from centre to centre on the flanks was •17 feet 1½ inches — this is nearly the same intercolumniation of 1⅘ diameter, used at Priene; the Mausoleum seems to have been about 1¾, and Miletus was 1½. The exact spacing was settled by that of the old Temple of Diana, but the plinths p20to the bases and the spaces between them seem to have been equal squares as at Priene and Miletus.
The noble capitals seem to be the parent examples of the whole later Ionian group (Fig. 17). The extreme dimension is about •8 feet 7 inches — the same as the plinth block; its height is about half a diameter over the volutes as at the Mausoleum and Priene. The faces of the volutes are not vertical, but lean outward toward the top. The rolls of the volutes at the sides are divided into four large flutes, the bottoms of which are filled by palmettes (Figs. 18, 19). Cockerell drew a capital almost exactly similar at Sardis which he said was the most beautiful he had ever seen. The abacus is not square, but •about 9 inches longer in the direction of the epistyle. In front of the cap are three enormous "eggs" of an egg and tongue moulding which at Priene and the Mausoleum became continuous, the eggs being the same in number as the flutes of the column. This is clearly a development.
The columns of the inner row seem to have been a little less in diameter than those outside. It is true that Wood says that the inner columns were of the same diameter as the outer ones, but p21on his section he figures it as •4.2 from the centre of one of these columns to the face of the step or plinth which carried it, that is •8.4 instead of •8.7 across. He also says that these inner columns had thirty flutes instead of twenty-four, according to a custom which Vitruvius mentions. Dr Murray says: "We possess two capitals which differ in size, and have assumed that the smaller belonged to the inner row." On the smaller capitals there is also a variation in the design, the palmettes come between the ends of the flutes of the roll of the volute, instead of at the ends of the flutes as in the larger capitals (Fig. 18).
Wood's account of the greater number of flutes on the smaller columns is not borne out by the portion of the smaller column erected in the Museum. The flutes are only •about 1‑4 inch less in width and were twenty-four in number.
As to Dr Murray's fifth point, the drum with the higher relief, which he suggested formed part of the angle column, Wood measured it as •3 inches less in diameter than the rest. Murray, in calling it larger, must, I think, have included the projection of the sculpture. It is much inferior in style, and is bedded in its height. I suggest that it formed part of the inner row between the antae.
Of entablature, Wood found several fragments. The architrave was •3.10 deep. What looks like a joggle-joint in the Museum fragment is probably a saw cut made by Wood. There is some difficulty as to this architrave. Its soffit is panelled, and supposing, as we must, that this panel was central on the surface, the total width would not have been less than •6.3. As restored, the margins are narrowed so as to give a soffit of about •5.3. On the evidence we must suppose that the architrave overhung the abacus by •2 or 3 inches. The abacus was so very narrow at the Old Temple that we may easily suppose that the architrave projected over the capitals, and this may have been carried over into the New Temple. There exist also in the Museum portions of two separate rows of egg and tongue mouldings and a length of the sculptured gutter, but no part of a frieze was found. As the Catalogue states: "Of p22moulding above the architrave and of the frieze nothing remains." I believe that they never existed, and that the entablature was of the traditional Ionian form which Choisy calls "the Architrave Order."
We must turn to the order of the Temple of Priene, which obviously much resembled Ephesus, and where the carved gutter, as will be shown, is a copy of that, to elucidate this point. When it was explored by the Dilettanti Society we are told that here again no moulding was found which would have occupied a place above the architrave, and although the order had been restored with a frieze until quite lately, the latest German researches show that it, the great altar to the east of the temple, and the second temple were all examples of the architrave order. The authors of the work cited say that "the temple, in fact, did not have a frieze, but, instead, directly above the egg and tongue moulding, which crowned the architrave, lay the powerfully projecting dentil course." These Priene entablatures are now set up in the Pergamos Museum. The dentil course resting on the architrave is the essential part of the earlier Ionian entablatures. A fragment of a cornice from Xanthus in the British Museum consists of a representation of a row of round poles, set close and far projecting, carrying a band. The two Lycian tombs have cornices made up of an architrave, a row of projecting blocks, representing ends of timbers, and a plate above of two facias. This type of cornice, consisting of architrave, dentils, and corona, is general in the Lycian rock tombs (Fig. 20). Choisy has pointed out that this is the true Asiatic (Ionic) type. At Athens, he says, dentils are only known at the Caryatide porch, and there, again, there was no frieze. At Miletus a frieze of sculptured p23scroll decoration appears, but it is of the Roman epoch. At an earlier time the tradition of the architrave order was so firm in Ionia that the Nereid monument in the British Museum, which in much seems to derive from Athens, has figure sculptures on the architrave but no frieze. The late colonnade of the Zeus altar at Pergamos still has no frieze, nor have the Sidon sarcophagi.
In Fig. 21 I have laid down the entablature of Priene and the fragments from Ephesus to the same scale, following the proportion of the former where we have no other guide. The result gives an entablature of •about 11 feet deep, and of practically the same proportion to the whole order as at Priene. At p24that temple the columns were at least 8⅔ diameters high, the dimension calculated by Cockerell, or nearly 9 diameters, the height arrived at by the German explorers from the masonry of the antae. The lesser proportion for Ephesus would give a column of •52 feet 4 inches, the greater of •about 54 feet, and a total to the top of the cornice of •about 65 feet. The greater dimension may, I think, be accepted. Wood proposed •55 feet for the column, and at Miletus the proportional height is still greater.
The Mausoleum, a famous companion work to Ephesus and Priene, has had its order restored at the Museum with a sculptured frieze, although here again fragments were found of only the two usual egg and tongue mouldings, and the whole makes up the abnormal depth of nearly a third of the column, instead of about a fifth. I have no doubt that here again there should be no frieze.
The sculptured gutter stone mentioned above is much broken (Fig. 22). It is clear, however, that a lion's-head spout occupied the centres of •lengths of about 6 feet. The rest is filled by elegant scrolls of acanthus, of which the upper part is broken away. On comparing it with the Priene gutter, also in the Museum, I find that the latter is a smaller and comparatively poor copy of the Ephesus gutter, so that one can be completed from the other with certainty. It is all the more certain, therefore, that both entablatures were alike (Fig. 23).
p25 There are at the Museum two stones of the slabs, •8 inches thick, which cased the pediment. In Dr Murray's restoration he has given dentils to the raking cornice, but this is most unusual, if not unexampled. At Priene they were omitted, and the band which took their place was less in depth so as to lighten the raking cornice. The "acroterion" found by Wood, and now at the Museum (it is not set upright), is much too small in scale to have surmounted the pediment. A series of them may have furnished antefixae along the top of the gutter, but it seems to me to be later in style than the rest (Fig. 24).
Wood's published plans refer to the New Temple. From his text we find that he discovered the positions of at least two of the ancient columns at the west end which are not laid down on his plan. This is what he says: "At last we found part of p26the base of a column . . . and in position a large square block of marble which proved afterwards to be the plinth stone of the base belonging to the more ancient temple. . . . The size is •7 feet 8½ inches, while that of the plinths of the last temple is •8 feet 8 inches, but the position appears to have been identical." This was in December 1870. On the following February: "The fine base of one of the columns on the south flank was discovered in position . . . this base is now re-erected in the British Museum. I had now two certain points, viz., the plinth stone of the base near the western end, and another near the centre of the south flank." (Fig. 7) By reference to the recent Austrian p27Survey it is clear that the plinth first found was that of the second column back from the angle on the south side, an important point. Again, on p267, he says: "We found in the west front the plinth of a column of the Old Temple as well as part of the base of one of the inner columns, consisting of the plinth and lowest circular stone." The former of these is probably that first mentioned, the latter is shown by the Austrian Survey to be that directly to the south of the south-west anta. "The position of these points," he says, "corresponds as nearly as I could ascertain with the columns of the last temple, giving a satisfactory proof that the temples were built on the same plan and of the same size." The Austrian Survey, in addition, gives some height indication of the position of one of the columns between the antae. In another place Wood says: "In January 1873 I obtained particulars relating to the position of the columns at the wet end." We can now see how he was able to get the dimensions for the remarkable spacing of the columns of the west front, which he gives as follows: From the left to the centre, •19 feet 4 inches, •20 feet 6 inches, •23 feet 6½ inches, •28 feet 8½ inches. This may be compared with the Austrian Survey, which gives in metres, 6.16, 6.16, 7.20, 8.75. These spans are immense; the nearest parallel case was found in the Temple of Sardis, where the openings, following the same order, were •16 ft. 3 in., •17.8, •21.7, •25.4. Wood broke up the cella with an opisthodomos to the east, but from the exactly central position of the great basis for the statue, and from the uniform disposition of certain rough masonry piers along the foundations of the side walls, it seems more likely that the cella was undivided in this way. Such an arrangement would further allow of an eastern door, entering the cella, for even if we are driven to suppose that the great altar and the principal door of entrance were to the west, I can hardly think that there was no central eastern door. The Austrians endeavoured to find the basis of the great altar to the west, but without success, nevertheless, they too consider it certain that the temple faced west. Wood concluded that piers spoken of above were Byzantine, but as they contained fragments of the Old Temple, and were p28spaced pretty regularly opposite the columns outside, it seems more likely that they were internal buttresses to stay the walls against the raised peristyle. The cella was •about 70 feet wide; if it were roofed it must have been subdivided. Wood found a curious Corinthian capital "near the cella," and supposed that it represented internal rows of columns. It is described as elliptic on plan, but it is really made up of a pier and two attached half-rounds (Fig. 25). It is only •2.6 high, and is of quite late Roman work. Such a capital might belong to an external stoa, or possibly to some small erection about the great image, but it can hardly have formed part of the structure of the cella. I am no believer in much which has been written on hypaethral lighting, but the evidence in this case seems to suggest that the cella was open as at Miletus. The lack of internal foundations, and the immense size of the great basis in the midst (•20 feet square, which would well have supported a covered shrine as well as the statue), seem to support this view. Wood also found a drain in the foundations of this basis, which he called an altar — "probably for carrying away the water used in washing the surface after sacrifice." There would be plenty of room in the pronaos, treasury, and other parts for the cedar roof mentioned by Vitruvius. Save for the seeming need of having a treasury at one end or the other of the cella, I had come to think that columns would be disposed in the space between the antae walls as at Miletus, and I find that the Austrian plan has set forth this view. With the inscriptions at the Museum are several walling stones.
We have seen that the temple court was surrounded by a colonnade; the plan of this is shown by Wood. Of the Doric building which stood beyond it on the south side there is p29a remnant at the British Museum of the triglyph frieze (No. 2,562), •18 inches high. Outside all was the great precinct wall, •about 12 feet high, of which Wood gives a diagram. Amongst the fragments are a base of a column •5.6 in diameter (Fig. 26), and a piece of guilloche moulding (Fig. 27).
A skilful restoration of one of the sculptured drums of the archaic temple set up by Dr Murray is shown in the Ephesus Gallery. Of the later temple, portions of four pedestals have been pieced together, and on these are set the sculptured tambours of the columns.
The merits of the sculptures are hardly sufficiently recognised. Reliefs so noble in style are indeed difficult to find. The standing figures of one of the columns are obviously inspired by the "Magistrates" of the Parthenon. The seated figures, on the drum of which I have attempted a restoration (Fig. 28), recall the goddesses of the same frieze, and the style of all has affinity with the finest Attic sepulchral reliefs. The gracious forms, severe while sweet, and at once broad and detailed in their treatment, follow the high traditions of the fifth century. Their scale is that of nature, amongst the few most perfect architectural sculptures of Greek art. The Hermes has been found copied, as Dr Waldstein has shown, on the silver patera of Bernay. The sculptures of the square pedestals, in higher relief and more violent action, seem to belong to a more advanced school. I suppose that they held the most important place in the front rank of the portico, and that here we find the work or direct influence of Scopas (Fig. 3).
p30 The most perfect group has been explained in several ways but I do not know if its striking resemblance to the group of Hermes, Eurydice, and Orpheus (Collignon, Fig. 69) has been pointed out. The latter is said to be a late fifth century work. The Hermes may be compared with that on the grave relief in the British Museum (No. 627), which itself seems to be a copy of the Hermes of Andros (Collignon, Fig. 201). The noble woman's figure, indicated on my sketch (Fig. 29), slightly restored, appears to me to be so close an adaptation of the famous Eirene of Kephisodotos that we may speak of it as a p31copy of that work. The pose is similar, and the draping is almost identical. The Eirene was wrought about 370; in turn it derived from the Maidens of the Erechtheum (c. 420). Further to the right is the lower part of a seated male figure in the attitude of one of the gods of the Parthenon frieze, and of a figure of Aesculapius from a coin of Epidaurus. This last is a small point, were it not that the coin represents the famous cult statue; and a bas-relief of a similar figure (see Collignon) from the same place has open-work shoes, which are identical with those on our column. On the left of the Hermes are two superb figures almost complete, and a small fragment of a third which was like one of the magistrates of the Parthenon who lean on p32their staves. The parallels to the Hermes mentioned above have been likened to the Hermes of Praxiteles, but our figure has quite sufficient earlier sources (see Nos. 22, 23, 25, and 30 of the west frieze of the Parthenon). Our sculptures, then, are of the Attic school of the transitional period of the first half of the fourth century. Are the seated women three of the Muses? The perfection of the modelling and finish may best be judged from an exquisite fragment, being the cheek and side of the head of a female figure (No. 1,239). This lies neglected in a case, and has never, I believe, been photographed. It deserves setting against a background in a glass case.
The remains in the Museum show that the archaic temple was fully painted. The throat of one of the volutes is still coloured bright red, and a leaf pattern around the top of the flutes of the column is painted. In one of the volutes a part of a gilt lead fillet was found. The sculptures of the gutter front and of the bottom drums of the columns were decorated. The hair, lips, and eyes of the figures were painted, and the dresses had bands of palmette and key patterns.
Sufficient was found to show that the New Temple was also painted, even to the bases, one of which was found coloured red in the scotia; and Wood illustrates a small fragment of egg and tongue moulding (Plate X, Fig. 6), which had the ground coloured blue. The use of these colours, bright blue and red, picking out the grounds of reliefs, was the traditional method of coloration in works of this school, and is also found at the Mausoleum and at Priene. Near the temple a fragment of Doric cornice was discovered, showing vestiges of colour, "blue, vermilion, and gold." This probably belonged to the building which stood to the south of the temple court.
Several of the volutes of the capitals in the Museum have never been completed. One of the capitals is nearly finished on one half, while the other half is only roughed out. It is evident that the capitals were fixed in an unfinished state, and a comparison with other examples shows that this was a general method of procedure. Choisy speaks of the "ravalement" after fixing as p33general. At the moment of setting, the faces were left rough (d'ébauche), only the reliefs were fully sculptured in advance, the rest was finished in place. For a column they marked in the shops the extremity of the flutes at the top and bottom, and the remainder was wrought after fixing. The most remarkable example of this procedure is the Temple of Miletus, where unfluted columns, unmoulded bases, and half-finished carvings were found. There is considerable freedom in the Ephesus work. The large capital from the outer row is wrought together with the top of the shaft, the bed being just below the rounded termination of the flutes. The smaller shafts rose to the underside of the eggs and tongues of the capitals, including the astragalos moulding in the more usual way. Neither of these carved mouldings space equally with the flutes, and the latter varies in spacing from 5½ to 6½ centres under the same capital.
The Lesbian leaf moulding on the pedestals varies in spacing •from 4½ to 5½ inches, and is quite sketchy in parts. In the Old Temple intentional variation of parts is as marked as in a Gothic building.
There are two records which bear on the date of the earlier temple. We are told that the famous artist Theodorus advised that a bed of charcoal should be laid over the site — a well-known and here very necessary precaution against damp — and that Croesus gave some columns to the structure. These facts would date the foundation as about 580, and the erection of the columns as about 560.
The building of the New Temple is usually dated by the story given by Plutarch and others that the old one was burnt on the night in which Alexander was born in 356. This story, on its face, seems to bear the character of a myth. Falkener pointed out that another date, at the beginning of the fourth century, is given in the Chronography of Eusebius, which contains valuable old material. He accepted both burnings, but it seems more likely that the true date was 395, not 356. The earlier year p34agrees much better with the architectural facts. The Mausoleum, which was begun about 353 (+ or -), seems to be considerably later in its style. At Priene, which it is held was begun about 345 and finished about 334, the order is copied even to the gutter, so that the whole height of the immense mass of Ephesus was finished before Priene. The carving of the gutter at the latter is poor, and seems much later than its fine prototype.
According to Vitruvius the temple was begun by Chersiphron of Cnossus and his son Metagenes, and was completed by Demetrius, a servant of Diana (Dianae Servus), and Paeonius, the Ephesian, who, he says, also built the Miletus temple in association with Daphnis. It is generally agreed that Chersiphron must have been a master at the erection of the early temple. Vitruvius speaks of him as contriving machinery for bringing the columns to the site, and of his writing a book, with his son Metagenes, on the order of the temple. Pliny tells a story of his setting the great epistyle. This last certainly could not have been reached in one generation, and it is much more probable that the temple was planned by the master who advised as to the foundations. It is somewhat astonishing to find that the archaic temple was so advanced in style, and so immense in scale. We have seen that Herodotus mentions it along with the temple of Hera in Samos, and the remains of the latter, which have been measured, show that, while in many respects it was similar, it was even larger by a few feet. Indeed, Herodotus himself says in another place: "The Samians have three works the greatest of all that have been wrought by the Greeks. The first the aqueduct, the second the harbour, the third a temple, the largest ever seen, and its architect was Rhoecus, son of Phileus, a native."
Dr Murray has remarked that the resemblance between the earlier Ephesus capitals and those of the Hera temple is "particularly striking and interesting," because of a recorded connection between the artists who worked at both — Rhoecus, the architect of the Samian temple, and the Theodorus who is associated with him by ancient writers.
According to Collignon, Rhoecus began the Hera temple about 600. Theodorus, who appears to have been younger and p35perhaps his pupil, was in full activity about the middle of the sixth century. His work at the Artemision was probably about 580. Rhoecus himself cast for Ephesus the most ancient bronze statue of which there is record; it was called "Night," and was placed at the altar of Artemis. Theodorus was an architect, sculptor, and writer who "possessed all the technical knowledge of his time." The signatures of both Rhoecus and of Theodorus have actually been found on votive works.
Recalling again the likeness of the two greatest temples of antiquity, the relation of the two cities, the fame of Samos as the chief art centre of Ionia, it seems more than probable that both temples were planned by those great early masters. Indeed, we may almost say that from no other centre and by no other artists could a hand in which sculpture is so integral have originated. We have seen that Vitruvius gives the names of Demetrius and Paeonius as having completed the temple, and this might mean, as has been supposed, that they were the chief architects for the later structure. Strabo, however, saysc that Deinocrates, the celebrated architect of Alexander — the master who planned the new city of Alexandria — was its builder. It might be held that the conqueror called the master from his great work at Ephesus; but Vitruvius tells another story of his introduction to Alexander, and says that he followed the latter from Macedonia to propose the scheme of cutting Mount Athos into the form of a statue. This itself is perhaps not a very certain story, but if we consider the dates, it becomes certain that Strabo's account is an instance of a big reputation spreading too far.
As to Paeonius, it is said by the last writers on the subject that the Miletus temple, which he is stated to have built, may have been begun as early as 332, but all which remains seems to me much later in style, and taking into account what Vitruvius also says as to the erection of an altar of Apollo (at Miletus) by Paeonius, it seems more likely that the architect of Ephesus is not the same as he of Miletus. Certainly if the New Temple of Diana was begun soon after 395 one master could not have planned the two structures. An inscription found at Miletus, p36with phrases parallel to the Servus Dianae quoted above, clears up one point, and shows by analogy that the Demetrius mentioned above was a slave attached by purchase to Diana's Temple, and proves that a master builder might be of servile condition.
c The passage in Strabo's Geography is XI.1.23; but all the manuscripts except one read Cheirocrates; we owe "Deinocrates" to the emendation of scholars (see notes there). Plutarch on the other hand — or his MSS. — calls him Stasicrates, and adds several interesting details (Life of Alexander, LXXII.5‑8).
Images with borders lead to more information.
With my warm thanks to
the author of a biography of W. R. Lethaby,
who introduced me to this study
and provided the copy from which this webpage was produced.
A page or image on this site is in the public domain ONLY
Page updated: 19 Sep 12