The Tomb of Mausolus (W. R. Lethaby, 1908)

p37 "My eyes have looked on the Wall of Babylon and on the Zeus by Alpheus [Olympia], and on the hanging Gardens, and the colossal Helios [Rhodes], and on the high Pyramids, and the gigantic monument of Mausolus, but when I saw the vast Temple of Artemis [Ephesus] soaring to the clouds, the others were all dimmed, for except in Heaven the Sun has never looked on like." — Antipater of Sidon (c. 100 B.C.) on the Seven Wonders.

Sources and the Site

In the British Museum Catalogue of Greek Sculpture, vol. II, pp66 and 67, are set out small prints of eight various restaurations which have been suggested for the world-famous Mausoleum at Halicarnassus, together with a good bibliography of the subject. Having been interested in the noble remnants of this puzzling monument preserved in the Museum, I was drawn on to read what had been written on the subject with the hope of discovering that which had been proved in regard to it and what is mere embroidery of conjecture. I propose to review the evidence of the stones themselves, and to bring out the points in what has been written which seem to me most in accordance with the facts.

The main sources of evidence are three — the marbles in the Museum, Newton and Pullan's account of their discovery,1 and a short description by Pliny.a As shown by the surveys, the city of Halicarnassus was built around a deep, almost circular natural harbour. It was rebuilt by Mausolus, and his great monument has such a prominent position in the centre of the scheme that Adler is probably right in arguing that it was begun by himself before his death, in 353 B.C. The city is so p38 accurately described by Vitruviusb that it almost seems that he might have visited the Mausoleum, which it is clear that he regarded as supreme among buildings. "The site of the city is like a theatre in form. On the lowest part by the harbour is the Forum, on the curve, higher up, about the middle, which was a broad street or precinct, in the centre of which stood the Mausoleum, a work so marvellous that it was counted among the seven wonders of the world. In the centre of the highest part of the city was the Temple of Mars, with a colossal akrolithic statue. At the right hand point of the curve was the Temple of Venus and Mercury and the Fountain of Salmacis. On the left horn stood the Royal Palace, planned by Mausolus himself."2

The explorations of Sir C. Newton laid bare the platform of the Mausoleum, which was "cut like a step in the side of the rising hill . . . It is probable that it was connected with the Agora, which was on the shore of the harbour below, by a series of terraces." The monument stood in a temenos, apparently •over three hundred feet square, and within this space, on the south and east at least, there were also terraces. Towards the northern peribolus wall the actual foundations were discovered, in an area which had been excavated out of the rocky platform to various depths, but always with level surfaces. The depth averaged •about 10 feet. This space was rectangular, •108 feet north and south by 127 feet east and west. There were some extensions p39 beyond the general form, and to the west a rock-cut flight of steps descended to the bottom. "The whole of this area had been filled up with the courses of the foundation, consisting of slabs of a coarse green stone, strongly bound together with iron cramps, and generally •about 4 feet square by 1 foot thick. In some places as many as three courses of those foundation slabs remained" (Fig. 30). A still better account of the great foundation is given by Lieutenant Smith in a Government paper describing the excavations. "The cavity appeared like a bed cut out to receive the foundations. The whole of this quadrangle is cut out of the rock to depths varying •from 2 to 16 feet below the surface. Where the rock has failed at the sides the line of cutting is continued as a wall. Throughout the area the rock is cut in beds or levels at different depths, caused by the irregularity of the ground." The site seems to have been originally a quarry. This accounts for the irregularities of depth and the extensions beyond the square.

Besides this immensely strong basis, fragments of marble columns and entablature were found sufficient to give data for a complete restoration of an "order," also many fine sculptures, including friezes, and a quantity of wide step-like stones suitable for roofing. The sculptures and architectural fragments are now exhibited in the Mausoleum Room of the British Museum.

The description of the Mausoleum given by Pliny also furnishes important data. It is to this effect:— It was •63 feet on the north and south sides, and shorter on the front. Its entire circuit was •411 (or 440) feet, and the height [of this part?] was •25 cubits (37½ feet), around the pteron (peristyle) were 36 columns. Four famous artists wrought the sculptures on the several fronts. A pyramid equal to the lower part in height surmounted the pteron, diminishing in 24 steps to a meta (generally a steep cone or pyramid). On the summit was a marble chariot the work of Pythis (Pythios, the architect), the total height being •140 feet.

Falkener, in an article in the "Museum of Classical Art," brought together one or two other notices. Lucian says: "Nothing is equal to it either in size or in beauty. It is enriched with the most perfect works of art, with statues of men and horses of costly marble, such as can hardly be found in temples, so that it p40 is the perfect model for all tombs." Pausanias also says that its size was great and its decorations magnificent.

It is easy from these accounts to get some general idea of the monument, and before the site had been explored many restorations had been suggested. The finds fully confirmed the general good faith and accuracy of the descriptions. One main difficulty of interpretation arose in that sides of •63 feet and shorter fronts will not make up a circuit of •411 or 440 feet, and the restorations fall into two groups — (A) the small plan p41 type, and (B) the large plan type. In the type A the 36 columns are arranged in two rows, in the type B in a single row (see Fig. 31).

The Peristyle

I have felt that it might be possible, by a detailed examination of the order, to obtain data which should offer proofs of one or the other types of restoration proposed; for in the order of the Mausoleum, if anywhere, we may expect an example of systematic relation of parts. Vitruviusc tells us how Pythios, its architect, with other contemporaries, gave up the Doric order because of the incongruous arrangements which arose in its use, and how Pythios wrote commentaries on architecture and a special description of the Temple of Athene at Priene, of which he was also the architect, at a later time.d Moreover, "Satyrus and Phyteus,3 who were very fortunate," wrote on the Mausoleum itself.e The beautiful order of the Mausoleum, says Laloux, was considered to have "particularly happy proportions, and they were copied in the principal productions of the school of Pythios." Adler speaks with less than enthusiasm of the order, but he judged only from poor, dry, and inaccurate prints. The actual stones at the Museum are surely very beautiful.

It is most difficult to get any statement of accurate dimensions of the order. Pullan's measurements, figured on his plates, are acknowledged to be untrustworthy, and the Museum should publish an amended account of the remains. For instance, Pullan gives the small top diameter as •2, 9.65.4 The Museum Catalogue says that "the diameter at the top of the flutes is •361⁄16." The large diameter he gives as •3, 5.35.5 In the Museum Catalogue the bottom diameter is not given, but only the sizes of the two top drums.

I have had the advantage of consulting a set of careful, full-sized drawings of this order, made by students of the Architectural p42 School at the Royal College of Art, under the direction of Professor Pite, and, further, I have found in the British Museum Lieutenant Smith's original measurements of all the drums found on the site, and on which Penrose based his calculation of the height of the order.6 The former gives •3 feet bare for the top diameter, and the latter 3 feet. The lower diameter I find given •from 3, 5 to 4, 7⅜. I estimate it as •3, 6½. These diversities as to the sizes of parts of the column which are actually in the Museum, may be taken as an example going to show on how unsubstantial a basis rest most of the calculations as to the proportions and refinements of Greek architecture. Watkiss Lloyd's most able tract on the proportions of Priene, in the fourth volume of the "Antiquities of Athens," for instance, is worked out with the minuteness of a problem in pure mathematics, and the only inaccuracy seems to be in the data on which the calculations are based. The latest German researches show that estimates of the diameters at Priene vary from 1.25 to 1.29 metres, that the height of the column was probably greater than had been thought, and yet that the whole order was much less than was supposed, because it had no frieze. All this may seem a little confusing on one hand, but on the other it should, I think, relieve the mind of the practical student, who, being accustomed to the rough approximations of modern building, cannot understand these interrelations worked out to the thousandth of an inch.

We come now to the important question of the intercolumniation. At Ephesus, which I look on as the type of a series of works which included the Mausoleum, the intercolumnar space was 1⅚ diameters; at Priene it is given as p43 1¾ by Pullan, and as 1⅘ by Rayet and Thomas; at Miletus, the latest built, it was 1½. In all, the spaces between the square plinth blocks were approximately square, and this stands out as the most certain criterion. Further, the capitals spread about half a columniation, that is, they are about as much on the face as the dimension of the plinth block, and this seems also to approximate to a rule. At Ephesus the abacus is •about 7 inches longer on the face than on the returns. At Priene it is nearer square. At the Mausoleum the abacus is square. This fact shows that the intercolumniation in this case was probably less than in the other two. Twice the size that the plinth must have been gives us •9, 9, or 1¾ diameters. This dimension has already been arrived at the Museum as resulting from the length of some stones of the lacunaria. As to this, it might be contended that the bays were not necessarily square on plan, and this, indeed, was the case at Priene, where square lacunar panels were fitted to oblong bays by means of additional shallow strips of panels. Now, at the Mausoleum there were similar narrow sinkings on the marginal stones of the lacunars, but three or four fragments of mitres of these sinkings have been found which show that here they ran all round, and not on two sides only (Fig. 32). The compartments were thus certainly square, as in the Museum restoration. This being settled, we have the fact that the margin-stone of one of the lacunaria gives the size of the square panel, and adding to this its own width twice over we again get the same dimension for the columniation (exactly •9, 9¼). It is now, I suggest, proved that the columns stood about •9, 9 from centre to centre (Fig. 33).

In trying to determine the size of the lacunaria we again come up against mistake upon mistake. In Newton's text the lacunar stone is said to show that "the exact size" of the panel was •4, 0‑⅜. On the plate, however, it is figured as •4.285, which is about 4, 3¼. In the Museum Catalogue the dimension is correctly given from the stone, but it is said that Pullan's plate is wrong instead of the text. I speak of this as some revenge for the awe I used to feel for these elaborately figured dimensions. The bases of the columns, as set up in the Museum, from the indications of the setting lines, gives a projection quite different from Pullan's version, and the bottom diameter is •about 4 feet 10½ p44 inches, or, as has been said, the plinth would have been one-half the columniation.7 There is a final proof of the dimension obtained for the columniation. According to Pliny the ends of the structure were shorter than the sides. It is manifest that this means that they had one or two columniations less. Now, •the difference between 108 feet and 127 feet, the size of the foundation, almost exactly equals two bays of •9, 9. It may now further be stated that the flanks were two bays longer than the front.

The 36 columns mentioned by Pliny can only be arranged in the two simple ways described above — either with 9 columns on the fronts and 11 on the flanks, or, less obviously, in two rows, the outside ranks being 6 columns on the front and 7 on p45 the sides (Fig. 31). But only the former has the proper difference of two bays, and this disposition exactly suits the size of the foundation. In this we have a proof that Pliny's number, 36, is trustworthy.8

The Pyramid

The facts known in regard to the Pyramid give further proofs as to the dimension arrived at for the intercolumniation and for the size of the whole peristyle. Many of the "steps" which finished the roof or pyramid of the structure were found. These steps were of two dimensions on the "tread" — •17 inches and nearly 21½. From 40 to 50 steps were found. "In all cases but two the treads measured •1, 9 or 1, 5;" the two exceptions had treads of •9 and 10½ inches. It is evident from this that most of the steps were wide. Two angle stones in the Museum show the dimension of •21½ inches in one direction and 17 inches in the other. If there had been a continuous pyramid of such steps its base must have had the proportion of about 34 to 43 on plan.

It has been computed from the actual sculptures of the chariot group which stood on the summit that it required a platform of •about 20 by 25 feet. This proportion is again practically the same, and we may now state that the whole base of the pyramid had sides in the ratio of about 34 to 43. This agrees almost exactly with the ratio of eight bays to ten required by the peristyle, and, as will be seen on the diagram (Fig. 34), the columniation of eight and ten bays must have governed the proportion of the pyramid base.

p47 We can again prove this result and arrive at the actual size of the base of the pyramid in another way. The great foundation had a ratio of •108 to 127 feet. This, although still oblong, approximates more to the square form than does the base of the pyramid, just as it should, for the plinths and steps on the ground level would be the same at both the fronts and the sides. Indeed, if from the centre of the foundation rectangle of 108 by 127 feet we draw lines representing the diagonals of the pyramid p48 having a base of the proportion 34 to nearly 43 •(17 inches by nearly 21½ inches) and now intersect these with true 45 degree mitre lines drawn from the angles of the foundation rectangle, we obtain geometrically the proper base of the pyramid (Fig. 34).

As some of Pliny's dimensions are corrupt and different in the different MSS., and others are difficult to reconcile with each other, and are therefore doubtful, even the number of columns he gives is subject to the same doubt; but from the three facts made known to us by the actual remains (1. The size of the columniation; 2. The size of the foundation; 3. The proportion of the base of the pyramid) we may arrive independently at the number of bays, eight on the front and ten on the side, by spacing up bays about •9, 9 as close to the limit of the base of the pyramid as possible. When it is found that the line so found for the centres of the columns agrees exactly with the line of basis of the pyramid, the demonstration is surely complete.

In the diagram, Fig. 34, A B are the angles of a rectangle of 17 inches by nearly 21½ inches, which was drawn full size. C D are angles of the foundation to scale. At E and F the diagonals and the mitres intersect. This gives the base of the pyramid, and it is found that the peristyle divides up accurately on the same line, giving 36 columns.

Besides the wide steps which have been mentioned, others, but not so many, were found, having treads of 10½, 9, and less, as may be seen at the British Museum. A few such narrower steps are required to get in the full number of twenty-four around the pedestal of chariot group. About sixteen or eighteen wide steps and six or eight narrower ones would suit our dimensions. At the base the first step or attic was doubtless much deeper than the rest.

The Cella

The name pteron requires a cella to which it is added.9 Newton pointed out that a fragment of a cross beam which was discovered must have rested on the cella wall, for it was rough for 2 feet at the end. He also writes, "One stone of the cella p49 wall was discovered." It had an inclination from the perpendicular of 1 in 100. Dr Murray agreed that the existence of a cella was shown by the marble beam — "this beam has one of its ends roughened the extent of 2 feet, for the evident purpose of being let into a wall. Had that end rested on an inner architrave, as does the other end, then it certainly would not have been roughened as it is."

Retrospect

I must now quickly review what has been written on the question, and try to separate results from unverifiable conjectures. p50 In the first half of last century, Professor Donaldson had observed many fragments of shafts, capitals, and other ornaments of a "superb Ionic edifice" on the site of Halicarnassus, which he seems to have identified as having belonged to the Mausoleum. In publishing an engraving of a pilaster capital he said: "In general character it is similar to the pilaster capitals of the Temple of Priene and of the Temple of Apollo near Miletus, but the rosettes and torches are additional ornaments, in allusion probably to the rites of the edifice of which the fragment formed a part"10 (Fig. 40).

1. With these evidences in addition to the texts, Cockerell made a restoration, and he doubtless obtained from Donaldson some idea of the scale of the Ionic order he had noted. He arranged the columns in a double rank, showing 7 and 6 to the outside (A, Fig. 31), and surrounding a very small cella.11

2. After the excavations Newton and Pullan published in 1862 their results, together with a restoration. While they arrived, as I consider, at the general form of the monument, they did so in rather an inconclusive way. They presented their facts most inaccurately, and made little distinction between them and their own suggestions. The order is not properly worked out. No pedestal is set on the pyramid. The wide span between the cella and the peristyle at the ends is impossible as construction. The pteron, with its cornice •about 100 feetº above the ground, is far too elevated for so delicate a work (Fig. 35).

3. Fergusson, in the same year, published a study of the monument, in the main following Newton. He put a basis for the chariot group as Cockerell had done, and after the analogy of the Lion Tomb at Cnidus, but introduced many errors and fancies of his own.

In 1888 Mr J. E. Goodchild, who had been an assistant to Cockerell, issued a pamphlet, which attempted to show that the discoveries were compatible with Cockerell's "small plan" p51 scheme. He pointed out that some of the reasons postulated by Pullan "would not bear their own weight, much less the superstructure." He noticed that besides the wide steps there were others having treads of 9 inches, and of those he proposed to form the pyramid. He worked out the intercolumniation from the margin stone of the lacunar, and showed that was 1¾ diameters. He had Newton and Pullan's wrong dimensions for both the margin-stone and for the diameter, but those happened to compensate one another in giving a columniation of 9, 6½, a result right in principle although wrong in fact.

5. About 1893 the order was pieced together in the Museum, the arrangement following Goodchild rather than Pullan, with some further variations. From the lacunar stones the right columniation, stated to be •9, 9¼, was arrived at. Dr Murray used certain square sculptured slabs as panels for the lacunaria. An objection may be raised to this on the ground that the panels have not equal margins all round, the margins above the heads of the figures are twice as wide as the others, and the
p52 sculptures are so delicate that they can hardly be seen. The sculptured frieze was inserted in the order. On this see below and Figs. 51 and 52.

6. A German restoration by Petersen and a French one by Bernier have few special points. Both are of the large plan type. The former puts pilasters to the lower storey (Fig. 36), and it is just possible that some such arrangement did obtain, see 11 below. In the French scheme a possible disposition of the lions is suggested, as some method of pairing them seems called for, as pointed out by Collignon, who published this restoration (published later and fully by D'Espouy). Petersen's scheme is illustrated in Mr Walter's "Greek Art," 1906.

7. In 1894 Mr Oldfield rediscussed the problem, following Newton's big plan type in the main, but he broke up the peristyle into porticoes, and was thus able to obtain short fronts while retaining the total dimension of •411 feet. If all Pliny's measurements are to be reconciled, this seems the best suggestion of dealing with them, although the attempt in this case was mixed up with many visionary features (Fig. 37). Mr Oldfield's scheme was adopted by Professor Percy Gardner in his "Sculptured Tombs of Hellas."

8. In 1896 Mr Stevenson also published a restoration (Builder, August and September 1899), in which he reverted to Cockerell's small type plan. He adopted both sizes of steps for the pyramid, and proposed for it a spire-like arrangement, consisting mainly of the steep steps. He swept away the cella, and supported the pyramid wholly on the columns.

9. In the same year Mr Arnold pointed out that some foundations of isolated piers which appear in Pullan's survey offer some confirmation of the large type of plan.

10. In 1900 Mr Arthur Smith published drawings of the order, and surveyed the whole problem in the "Catalogue of Greek Sculpture," vol. II.

11. In the same year Prof. Adler of Berlin published a monograph on the subject. He follows Newton and Pullan, not only in the general form they arrived at, but in their method of obtaining the size of the plan, by setting twenty-four wide steps around the area where the chariot stood. He adopts Cockerell's and Fergusson's idea of the "meta," and thus lowers the peristyle. p53 Indeed there is practically a consensus on this point. He re-engraves Pullan's faulty plates of the order, and indeed does not seem to have seen the original stones. His best points seem to be an exhaustive history (in the main again following Newton), suggestions for setting the sculptures, and the presentation of what I must, from what has gone before, consider the best general view of the monument which has been produced (Fig. 39). He shows that the doubtful 440 of Pliny is to be preferred to 411 and better agrees with the foundations. His most original contributions are in the disposition of the sculpture, and in recalling that Donaldson had illustrated the pilaster capital as coming from the Mausoleum (Fig. 40). He sets such capitals on pilasters around the cella. Cockerell also had used this capital, but he placed it on square angle piers, which, now we have an angle capital, we know did not exist. Adler suggests that this may be the earliest Ionic pilaster capital of this type, but Inwood engraved an example from Athens on Plate 28 of his "Erechtheum," which is clearly of still earlier form.

12. In 1905 Dr J. Six published an article in the Hellenic Journal, suggesting that many of the sculptures were grouped in pediments over the two "fronts." He followed Adler generally, but increased the attics to receive these features. The existence of two actual angle pieces of the gutter precludes the possibility of pediments filling the fronts as he suggested, and the great horse and rider seem altogether too big to have rested on the thin shelf of such a delicate entablature. Moreover the square support under the middle of the horse shows that it was isolated.12 This learned article goes on to discuss the proportions of the structure, but as Pullan's faulty measurements are used the results cannot be accurate. Dr Six also suggests that Pliny's perimeter of •440 feet should be broken up into sides of 120 and 100 feet. I had come independently to this last probability or possibility as likely dimensions for the top step of the platform. It is also suggested that the columniation may have been set out at 10 Greek feet from centre to centre, but the true dimension is 3 or 4 inches short of this. The whole flank, p54 however, would appear to have been exactly 100 Greek feet, for 10 × 9, 9¼ = 97, 8½, and adding two half columns, 3, 6½, we get 101, 3 English.

As said above, attempts have been made to reconcile Pliny's dimensions in all kinds of ways. Mr Pullan supposed that the 63 feet applied to the cella, and the •411 feet to the peristyle. Mr Oldfield's plan was broken into the form of a cross with very short arms, which gave narrow fronts. This variety we may call the "broken type." Mr Stevenson suggested an inner building small and high, and an outer enclosing building low and large. If I must put forward a possible reading, I would say that as the measure about the exterior was •440 feet, •63 feet was the interior length of the cella, obviously! Professor Gardner suggests that cxiii should be read for lxiii, and this would do very well for the top step of the platform. But none of these reconciliations are so simple in the supposition that easiest thing a mistake. As Furtwängler says on another subject, "It seems somewhat better to set aside Pliny's information than to try to combine it with the known facts." In the parallel case of Ephesus, Pliny's dimensions have had to be abandoned. For the Mausoleum he gives the measure of the long sides as •63 feet, while his estimate of the total measure round about is •either 411 or 440 feet, according to different texts. Which of these last shall be accepted as right? It would not matter much which it was, except that another writer says that it was 1340! However, this last could be made to apply to an outer court, although it is rather like measuring a field for the size of a haystack, and other reconcilers think it is in mistake for 440. p55 But, again, there are three estimates of height by as many authors — •180, 140, and 80 feet. Shall we reconcile all these by supposing that the first is to the top of a flag-staff, the second to the chariot, and the third to the cornice? Nothing final can be done with dimensions like these, all sorts of conjectural restorations are possible, and unless we proceed independently the problem is insoluble. For instance, Fergusson grouped his columns at the angles into threes. Another might suppose that the angle columns, as is so frequently the case, were counted twice over, or another, again, that the •63 feet should be cubits like the dimension of height. On these lines I put on record a rejected hypothesis (Figs. 41, 42, 43) which combines many advantages. The long front has seven bays which at •9 feet each make up the desired dimensions (inter-columniations of 1½ diameters were quite possible, and the lacunaria need not p56 have been square except for our proofs). We get pediments at the "fronts" (without sacrificing the gutter returning at the angles, as we know it did). We get the pyramid firmly based on the cella walls. ("Hanging in void air" is rhetoric for high. The dome of Sta. Sophia, although it seemed to be suspended by a golden chain from the heavens rested on very substantial piers and arches.) The measure of •440 feet comes at the edge of a reasonable terrace. We get the proportion of two bays longer than wide to suit the foundations. A central pillar on the front is avoided, and there is room enough for a pyramid made up mostly of the wider steps. Yet the sufficiency of the proof for the simple solution following the large type of plan B (Fig. 31) drives me from such a desirable scheme, and I cannot doubt that the main facts as to the exterior have been established from the data made known by the excavations as above set out.

A great point has been made of Pliny's description of the apex of the monument as like a meta, but I have not seen the parallel case of the Tower of the Winds at Athens cited in this relation. This building had a very flat pyramidal roof of marble, and at the centre a finial like a capital, yet Vitruvius describes it as a tower finished with a meta of marble ("metam marmoream perfecti").

Again, those who would sweep away the cella point to the tomb at Mylasa; but this is very small, of late Roman work, and the form of the columns suggest, that there was a screen between them forming the whole into a cella. The columns were of the shape shown by Fig. 25 in our Ephesus section, and Pars' original drawings show that there were dowel holes in the vertical strips by which it is probable slabs were attached.

The small type of plan seems to me to be an impossible solution for the following reasons:—

1. The pyramid has to be designed as made up mainly of steep steps. The discoveries showed that it was mainly of wide steps.

2. The monument would have occupied such a small part of the immense foundation.

3. It would not have been the right proportion of plan to suit a rectangle •108 by 127 feet and to give two bays longer on the flanks than on the front.

5. The small scheme had its origin before the site had been explored. It is against the views of Lieutenant. Smith, Newton, and Pullan, whose conclusions were based on their knowledge of the site, of the positions of the stones found, and, in a word, their total impression derived from months of study of the excavations.

6. Such a building would, I think, be historically impossible. The basis of the true design seems to be the tumulus developed, consisting of a basement, a pyramid, and a trophy. It may best be compared with the Cnidus monument; but in later times the great tomb at Adamklissi, and the "mausolea" of Augustus and Hadrian in Rome followed the same tradition. The marvel must have consisted in setting over a temple-like structure a pyramid hanging high in the air.

The one point that may be claimed for the small scheme — that it satisfies Pliny's dimension of •63 feet — is neutralised by the fact that the large scheme satisfies the •440‑feet dimension. And if we are told that this was the size of something exterior to the monument proper, we can say that the •63 feet is just as likely to be of something interior like the cella.

Details of the Order

The shafts have twenty-four flutes •over 5 inches wide, which are nearly semicircular in form. The projecting hollow moulding at the top and bottom of the shaft is not a quadrant, but p58 a long, elliptical curve, and the flutes are set into this in a beautiful way. The entasis of the column seems to be quicker toward the top, and Penrose considered that it formed a part of a long hyperbola with its focus some distance above the top of the shaft. The top diameter seems to be one-seventh less than the bottom.

The capitals must be carefully examined at the Museum to be appreciated. They are almost exactly like those at Priene, of which an example is also exhibited. The height of the cap from the bottom of the volutes is half a diameter. One of the capitals is from an angle, and shows a diagonal volute. It should be noticed that this angle volute is not thrown quickly out from a general square form, but the whole of the two external sides curve outwards, both the cushion and the abacus. The eyes of the volute were sunk out about •2 inches, here and at Priene and Ephesus, and were filled again with marble studs, some of which remain.

Often the studs projected beyond the general face of the volute, and this doubtless furnished the reason why they were separately inserted. The eye space was made use of also, and might be slightly injured in tracing the curve of the volute. At Ephesus the fluted rolls or volutes were adorned with palmettes. An almost exactly similar ornament is found at the back of the angle volutes at the Mausoleum and at Priene (Fig. 44). Doubtless p59 the angle volutes at Ephesus were exactly similar. The inner angle of the Mausoleum capital was recessed, as a separate fragment shows, so to allow room for the bringing of two volutes together at right angles. At Priene the two volutes of the inner angle were complete. There was not room enough for this on the Mausoleum capitals where the outer revolution of the volute was lost at the mitre. Even to get this much the length of the side rolls had to be reduced on the angle capitals. See Fig. 45, which shows the relative sizes of the two capitals. Figs. 46 and 47 show the mitring of the volutes against the inner angle. Fig. 48 is a restoration of the abacus at this point. Fig. 50 shows the ordinary capital.

As to the height of the column at the Mausoleum, we are told that "Mr Penrose, after a most careful discussion of the dimensions of twenty-six drums measured by Mr Newton at Halicarnassus, and by those which are at the Museum, came to the conclusion, from the entasis, that the pillars were 8⅓ diameters in height, and consequently within a small fraction of •28 feet 6 inches in height."13 The diameter here accepted is Pullan's wrong value of 3, 5.35. With a more correct value, we get about diameters. The memorandum spoken of above is a sheet (in MS. 31980, by Lieutenant Smith?), giving the sizes of all the drums discovered. On it a series of six drums, which agreed most nearly with one another as they diminished p60 upwards, are superimposed, giving a total height for the shaft only of •26, 11¼. This furnishes a maximum dimension, for the drums of the several columns would have varied in their heights, and in the series of drums brought together the upper ones, for the most part, overlapped those next below by a quarter of an inch or so. Making allowance for these differences, I estimate the height of the shaft at about •24, 9. The capital and the base add •2, 6, and the plinth, following the analogy of Priene, would have been •9 inches. This gives a total of •28, 0, nearly 8 diameters, or, accepting Penrose's calculation of •28, 6, we get 8 diameters exactly. I am satisfied that the columns were, at most, not higher than this. A reason for the abnormally stout proportion is furnished by the great burden these pillars bore.

We have seen that Wood remarked that the columns of Ephesus diminished rapidly towards the top. Mr Penrose, in his discussion of the columns of Priene, says, "The principal amount of curvature is found toward the top of the shaft, whereas in the Attic examples it occurs near the base (at the Parthenon it is below) or the middle. If an hyperbola were chosen to represent the curvature its vertex would be above the capital." On plotting the dimensions of the memorandum to an exaggerated scale p61 (24 to 1), I find that the Mausoleum shaft followed the same rule, the lower half was very straight, but not upright, the upper part curves swiftly so that the vertex would be •10 or 12 feet above the capital (Fig. 49).

It has long been discussed whether the sculptured frieze belonged to the order, and Furtwängler and others have applied it to the podium. Following the analogy of Priene and Ephesus, I have shown that there could have been no frieze of the order. This is proved so far as the Mausoleum is concerned by the way in which the cross beams are notched down for the full depth into the epistle. If there had been a frieze the cross beams of course would have rested on the epistyle. Combining our entablature of architrave, egg and tongue course, dentil course, corona and cymatium we get a depth of •5, 11. Now at Priene the depth of entablature was three-fifths of a columniation, the arrangement proposed for p62 the Mausoleum gives the same. At Priene the intercolumniation was a little more open in proportion and much greater in absolute span, so that the advantage of strength is with our entablature.

At the Museum there appear to be fragments of all these parts of the entablature. The dentils are blocks •10 by 7 inches and 5 inches apart. Following Ephesus and Priene I have put an egg and tongue moulding and a cavetto beneath the corona, and this takes the place of the shallow bed mould of the Museum restoration for which there is no evidence I believe (see Fig. 51). In the Museum restoration the projection which has been given to the cornice is much more than in Pullan's plates, the front of the corona being •1, 9 in advance of the bed mould below (Fig. 52). The under side of the marble was finely dressed to p63 this distance, "where there is a slight rise as if for a bed." As this gives a projection much greater than in other examples, I would suggest that it should be pushed back 4 inches or so, the dressing having been so wide in order to relieve the bed mould from the weight, and this, I see, is how Pullan has shown it. On the top of the gutter "a line is marked" •1, 10 back from its nosing. Newton says that this line marked the commencement of the pyramid. In the Museum Catalogue it is called a "weather line which is supposed to indicate the position of the lowest step of the pyramid." This entirely impossible arrangement would give no back to the gutter. It brings the weight of the pyramid not only on to a 4‑inch gutter stone, but about •3, 0 in front of the architrave, resting on the thin corona of the cornice. There are several yards of the gutter set up in the Museum (Fig. 53). It is divided into stones •42 inches long, with a lion's head spout in the centre of each. As before mentioned, remnants of two angles exist; both show that the palmette carving came up to the mitre, and one gives just enough evidence to show that the first head was p64 about half a stone back from the angle. The law of the distribution of the heads was to have as much carving between them, bending around the corner, as there was on the straight. The proportion of the cornice as above corrected was such as to throw the second lion's head about 6 inches behind the centre of the angle column as seen in elevation. We may find here some check on the dimensions already obtained, because the distance between the lion's heads should, on one side at least, be a multiple of •3, 6. Now as eight bays of the short side at 9, 9¼, equals 78, 4, and deducting 12 inches for the eccentricity of the heads nearest over the centres of the columns, we get a total of •77, 4, which is 4 inches more than 22 times 3, 6. On the other front we get •97, 8½, which is 3½ inches less than 28 times 3, 6. That is •3, 6 was the best possible mean for the dimensions of the two fronts. In other words the gutter stones would have averaged •about ⅛‑inch more, and ⅛ inch less than 3, 6 on the two fronts.

Fig. 54 is an enlarged detail of the base. Fig. 57 is part of an inscription, which is probably of the sixteenth century, cut into the shield borne by one of the figures of the frieze.14

p65 There appear to be several fragments of the Mausoleum built into the walls of the Turkish fortress of Budrum. I cannot find that they have ever been described.

Some interesting observations on the peristyle by Mr Marshall were published in the Builders' Journal, August 1899. He pointed out that the top bed of the angle capital was countersunk, and that half the top of another capital was similar, while another has no sinking, and suggested that corresponding settings must have been left on the architrave blocks, which thus at the angles set into the caps and resisted displacement. He also gives a good sketch of the ceiling over the peristyle.

It is probable that the columns all leaned inwards, at least one in the hundred following the inclination of the cella wall.

Sculpture and Colour

The principal sculptures could not have been in pediments, as has been shown, yet Dr Six is probably right in grouping them after the model of the Sidon Sarcophagus. The description calls for an equal distribution on all four sides. Adler's disposition of groups against the basement seems to best suit the evidence.

The great chariot group which stood on the summit is particularly noble; it might be nominated for a place amongst the most "universal" pieces of sculpture in the world, for that master of mediaevalism, Viollet le Duc, has rightly picked out the figure of Mausolus as having a character comparable with the best mediaeval sculptures. I only know one other figure which has the same "feeling," the slender draped girl's figure from Priene, which I should like to think of as by the same artist, Pythios. It might have come from Rheims p66 or Chartres. The lions are comparatively poor; from the position in which they were found it seems likely that they occupied a high situation. Lions appear as guardians of the tomb at Cnidus, Xanthus, Miletus,15 on the Lycian tomb in the British Museum, and the Sidon Sarcophagus at Constantinople, and many other places.

From the fact that they all turn their heads at right angles, alternately to the right and left, it is certain, I think, that they presented a side view, and could not have been seen end-on as by Adler. If they were above the cornice they may have formed a procession approaching the centre of the front along both sides.

The frieze is an extraordinary delicate work. It most probably surrounded the basement at no great height, like a similar frieze on the Nereid monument. It was decorated with many additions of bronze, doubtless gilt, and with colour. Newton says that "the bridles of the horses, as in the frieze p67 of the Parthenon, were of metal"; but there was much more. For example, I find by examining the positions of many drilled holes that the warrior on slab 1007 had a bronze crest in his helmet; an Amazon on slab 1006 carried a sword strongly pinned to her hand; another, mounted on horseback, held a bronze bridle; another, in 1013 thrust a spear into a man's side; another, in 1015, grasped a shaft of bronze, which was inserted into an axe-head which formed part of the marble relief; on 1016 was an Amazon riding spear in hand. The figures are all in most violent action falling into a series of diagonal lines, but the composition in its continuity and easy variety is very wonderful, and the execution most masterly. The tense and slender figures have "the accent of bronze." It is an extraordinary chef-d'oeuvre, but it does not satisfy like the Parthenon friezes. One or two of the fallen Amazons are charming in design and feeling (Figs. 55, 56).

"The whole frieze was coloured, the ground of the relief was ultramarine, the flesh a dun red, and the drapery and armour were picked out with colours." The ground of the chariot frieze was likewise blue, and the plain moulding below had a painted p68 leaf moulding, traces of which may still be seen in one place. A note on the memorandum in the Museum shows that the small square panels also had blue backgrounds. Indeed, this treatment of relief was customary. The principal sculptures and the lions also showed vestiges of colour. Newton reported that on its discovery the colossal seated figure plainly showed two colours. All the carved architectural members were painted. The colours were ultramarine and vermilion, "or pigments equal to them in intensity." Newton says that "the system adopted seems to have been to tone down the whole of the marble with a coat of varnish and wax, to paint all grounds of sculpture and ornament blue, and to pick out the mouldings with red." A lacunar margin-stone was found with bright blue upon it. This was doubtless in the recessed channel of it. At Priene, where exactly the same system of architectural coloration was followed, the recess on the soffit of the main architrave was blue, and the leaf moulding around it was picked out in red. The capital there had a red ground to its egg-and‑tongue, and blue to the leaf moulding of the abacus.

The eyes of the volutes were doubtless gilt. In the account for building the Erechtheum gold leaf for gilding the eyes of the columns is mentioned. Newton must be right, I think, in speaking of the ground of wax. The most perfect ancient painted work I have ever seen, the so‑called Sarcophagus of Alexander, is brightly coloured, yet the whole is harmonised and softened into waxy texture and hues. "The Greeks," says Choisy, "did not conceive of form without the association of colour. At all epochs colour was present, and even the statues were painted."

Construction and the Architect

The type of construction of which the Mausoleum is an example is remarkable from the way in which the marble is handled. As usual in Greek works it is put together without mortar, the joints being polished so as to sit very close. An abundance of bronze cramps was used to link stone to stone. p69 In the columns were fine bronze dowels, some of which are preserved at the Museum. The steps of the Pyramid have a raised fillet along the back and at the two ends. The latter, with those on the adjoining stones, make rolls which throw the water away from the actual joints. These fillets fit into cavities cut in the next course above. The scheme is derived from tiling, and I think it probable that the exposed joints were covered by a marble Λ piece. The entablature is, as we should think, carelessly constructed, being broken up into many pieces, and with fragile ornamental members inserted in rebates (Fig. 53). These points, and the use of mitre joints for the lacunar margins and carved mouldings, make one think of a sort of "marble joinery" rather than masonry. The workers must have been skilful in an extreme degree, and the whole outlook is very advanced and even doubtful. Adler suggests that the insertion of the delicate mouldings in rebated ledges was done with the object of hastening the works. The same custom is followed at Priene; and another reason, I think, is, that these parts are wrought in a much finer quality of marble.

The architrave is wrought in two beds, one 1, 6 deep with two facias, and the upper 1 foot deep with one. The heading joints of both come together over the columns, and at the back cavities are formed to receive the ends of the cross beams. The facias of the architrave are not in vertical planes, but incline outwards at the top (Fig. 51).

Pythios, the sculptor of the noble chariot group, may only have been a younger associate of Satyrus, with whom he is mentioned at the Mausoleum, as he was architect of the Priene temple, which was probably begun about 345 and was not completed until about 344.16 He belongs to a class of sculptor architects who shaped the course of Greek architecture. Pheidias, who himself was general master of works for Pericles, was followed by Kallimachos, Scopas, Polyclitos the young, and Pythios. The rôles of sculptor and architect seem to have been interchangeable as in the Middle Ages; the Greek sculptor was a "stone-cutter." After seeing the chariot group we can understand p70 Vitruvius when he says: "Those who are initiated in different branches of knowledge have facility in acquiring all, from their connection with each other. On this account Pythios, one of the ancients, architect of the noble Temple of Minerva in Priene, says in his Commentaries that an architect should have that perfect knowledge of each art and science, which is not even acquired by the professors of any one in particular." A large claim, but I am at least convinced that in all great times architects have been stone-cutters.

The Author's Notes:

2 In an account of the site by Covel (c. 1675) he mentions the walls and the fountain (B. M. MSS.)

6 The former will be referred to as "R. C. A. Survey," the latter as "the Memorandum."

8 Falkener dealt with objections to central columns. Where conditions were free, as in the flanks of temples, an odd number of columns seems to have been preferred. An interesting fact has recently been discovered in regard to the temple of Samos. It was dipteral, with 8 columns on the east front, 24 on the flanks, and with prostyles in front of each end of the cella, giving three rows of columns at the ends. At the west end there were 9 columns instead of 8 to lessen the spans. With the columns as described, and four more in the pronaos, we obtain a total of 127, the number Pliny said there were at Ephesus. Fergusson pointed out that at the latter the seven very wide spans of the principal front equalled eight of the medium spans, and it seems very probable that at Ephesus, too, the back portion had 9 columns like Samos.

11 Watkiss Lloyd made a slight amendment to the scheme, and this is embodied in the fine coloured drawing of Cockerell's restoration shown at the Museum. This drawing was made by Mr F. Pepys Cockerell.

12 There were no sculptures in the pediments at Ephesus or Priene, and Rayet and Thomas point out that they were generally plain in Asia Minor.

Thayer's Notes:

a The passage in Pliny's Natural History is 36.iv.30‑31; some slight additional information is found in 35.xlix.172 and 36.vi.47: the Mausoleum was constructed of a brick core clad in Proconnesian marble.


Fig. 47 — Sketch of the Cast.	Fig. 48 — Restored Abacus of the Cast.

W. R. Lethaby: The Tomb of Mausolus, London: B. T. Batsford, 1908.