History of Ancient Art Part 8

(_Fig._ 121.) Its walls begin to curve from the floor, which is of stamped clay pise. Upon this the first circular course of masonry immediately reposes. The walls then rise, in parabolic outline, to a pointed apex. They are not constructed upon the principle of a vault--that is to say, with wedge-shaped stones, and with the direction of joints to a common centre--but are laid in horizontal beds, each course so projecting over the one beneath it that, by this diminution of the concentric circles, they finally unite at the summit. They were smoothly cut upon the jointing surfaces, while the face was not chiselled until after the completion of the masonry. The blocks were rectangular, and the joints, which consequently increased radially in plan, were filled with the same pise used for the floor; the interstice between the wall door and the rock-cut inner chamber upon one side being also cemented with this substance. An entrance-pa.s.sage, the dromos, led from the valley to the tholos in a gently inclined ascent. It was bordered by walls of cut stone, but nowhere ceiled. Its floor, 6.20 m.

broad and 36 m. long, was paved with pise. Thisentrance-pa.s.sage was terminated without by a terraced retaining-wall, and within by an elaborate portal facade. The recent investigations of Stamatakis and Thiersch have given sufficient information concerning the composition and details of this front to permit a restoration of its chief ma.s.ses.

(_Fig._ 122.) The lower part was constructed of long stones, carefully cut and jointed. The stepped jambs of the opening, peculiar to all antique doors, were probably cut after the blocks were in position. Upon either side were decorative engaged columns, which are so entirely similar to the one represented upon the Gate of the Lions at Mykenae that it is possible completely to understand their nature by that general guide; by the help of fragments which still exist, and others drawn in former publications, though now lost; by traces upon the wall, and especially by the sockets cut for the swallowtail clampings of the bases and capitals. The shaft, instead of being diminished, increases as it ascends, as does also the column upon the relief over the Gate of the Lions. Its base, from this a.n.a.logy, and from the narrow s.p.a.ce left for it by the clampings, seems to have consisted of a simple tore. The abacus and parts of the mouldings beneath it still exist; the coronation was formed by two roundlets, separated by a scotia, the lower being considerably smaller in height and diameter than the upper. (_Fig._ 123.) Without the lower member, there is a certain similarity of the capital to a Doric echinos, which is increased by the proportions of the boldly projecting abacus; but the whole is so similar to an Asiatic (Ionic) base that it was not natural to believe it a capital, and the fragment published by Donaldson has. .h.i.therto been believed to be the foot of the shaft. The columns were entirely covered with an ornamentation in relief of zigzag lines alternating with the well-known spiral wave; they stood upon rectangular pedestals, of which the triply stepped plinths have been preserved. The existence of bronze ornaments upon the lintel of the door is evident from the traces of nails; five lion-heads can be distinctly recognized. An epistyle extended from capital to capital across the entire front of the portal; it projected far beyond the lintel, upon which it partly reposed. Above this entablature was a surface, like an attica, which masked the triangle formed by the relieving blocks over the lintel. The upper walls were not originally visible, having been reveted by thin slabs of stone, secured in position by dowels. Fragments from Mykenae deposited in the British Museum, in the Munich Antiquarium, and in Athens appertained to this upper facade; they all show spiral ornaments between horizontal grooves, and are similar to many other decorations of the same age. The borders of the casing over the relieving triangle and its extreme upper corner were patterned in like manner, as is plain from the mitre-joint of some of the slabs, and from a small fragment exactly fitting the upper angle of the opening. The entire triangle was probably closed by some light stone carving, since it could have had no function as a pa.s.sage for light. The door, as may be seen from traces of pivots upon the sill and lintel, had two wings, which, from their bolt-holes, appear to have been so large that, when closed, they considerably overlapped. Upon the exterior jambs a broad strip of metal was affixed, still to be traced by two vertical rows of nail-holes, in which fragments of bronze occasionally remain. This work leads to the supposition that the wings of the door were themselves overlayed with metal, and, with the characteristic forms of the decoration upon the monument, points to the peculiarities of Asiatic art. It is natural to attribute this to the influence of Phnicia; indeed, the effect of the civilization of that country upon early Greece can hardly be overestimated. A broad, horizontal strip of metal sheathing existed also upon the exterior, and small fragments of it are repeatedly met with in the rubbish filling the tholos; similar vestiges are found in a second monument of the kind near by. This overlaying of walls with sheet copper was by no means uncommon in ancient Greece. The subterranean bronze chamber of Danae may be explained as a tomb sheathed with metal. In mythical ages, in the sanctuary at Delphi, as well as in later times, in the Chalkioicos of Athene at Sparta, this wall-treatment appears employed for temples, even as Homer described it in palaces at Sparta and the Island of the Phaeacians. The Tholos of Atreus was itself subterranean; the exterior of the conical ma.s.s of masonry was covered with a hill of earth. In consideration of the almost perfect preservation of the interior, it is evident that some remains of a strictly architectural exterior would have been recognizable, had it existed. A tumulus covered and protected the structure; though its earth is now, for the greater part, washed away, to it must still be ascribed the good condition in which the kernel has remained.

The recently discovered grave at Menidi, in Attica (Lolling), is a parallel construction. As regards beauty of execution and richness of ornament, it is far inferior to the Tholos of Atreus; it is also much smaller, having an average diameter of 8.35 m. and 9 m. original height.

Its only peculiarity is that the relieving blocks over the lintel, instead of projecting one over the other so as to form a triangle, are so placed as to leave four voids between as many horizontal beams, in a manner similar to the arrangement for relieving the ceiling of the princ.i.p.al chamber of the great pyramid of Gizeh.

The Tholos of Atreus offers a welcome commentary upon the thesauros of the royal palace at Ithaca, but only in respect to its construction. The purpose of the circular buildings still existing in Greece seems to have been entirely different from that of the treasure-house described in the Odyssey. It is true that eminent authorities deny this difference--and the a.n.a.logies of the round Homeric building, of the treasure-vaults at Mykenae mentioned by Pausanias, and of the treasury of Minyas in Orchomenos, lend their arguments some weight, and, at least, a greater probability than the suppositions that the structures of tholos form were intended for spring-houses (Forchhammer) or places of worship (Pyl). But there are reasons against all these a.s.sumptions. The treasure-houses of the Pelopidae must have been upon the acropolis, inside the fortification walls, not at various distances outside their limits, as is the case with those of Mykenae. Still less could such vaults for h.o.a.rded valuables have been as distant from the city as was the Tholos of Baphio from the ancient Amyclae, which stood entirely isolated in the midst of an open plain, without the possibility of communication with any royal residence. The tumuli of earth above the crypts would have but ill suited them to form a part of the palace building; while for a cell which was only to receive precious goods--for a magazine of deposit--the rich overlaying of the interior walls with sheet metal, and especially the elaborate carving of the portal front, seem out of place. These peculiarities, not to mention some of less importance, point to another purpose, for which they are, one and all, fitted--namely, the destination of the structures as tombs. Their position, before the acropolis and without the city walls; the covering of the chamber with earth in a tumulus form; the impossibility of their having had any communication with other buildings; the elaborate decoration of the entrance, and the princely wealth of metals in the interior--all support, with the striking a.n.a.logies beyond the aegean, this conception of the tholos buildings advocated by Welcker and Mure.

It is possible that it is to these structures that Pausanias refers as the treasure-houses of the Atridae; but Pausanias, like us, knew Mykenae only by its ruins. That patron of all _ciceroni_ upon cla.s.sic ground was not exacting for proofs of their legends. The hypothesis of Pyl may in so far be correct that the tholos itself did not serve as the place of sepulchre, which was provided by the small side chamber, but was a chapel for the funeral worship naturally to be a.s.sumed in connection with an heroic dynasty.

It is not possible to a.s.sign these tombs to individuals, like those of the early Persian monarchs, or even to dynasties: the questionable identification of the graves discovered in the agora of the acropolis, ventured by Schliemann, would here be inadmissible. It is reasonably certain, however, that the best-preserved tholos, that known by the name of Atreus, is about contemporaneous with the Gate of the Lions, and dates from the most flourishing period of the heroic age--before the downfall of the Atridae upon the return of Agamemnon.

A small chamber, only of sufficient size to receive the cinerary urn, in the centre of an upheaval of earth, was sufficient for the graves of the heroes who fell before Troy. Several of these tumuli exist. The larger of them, those of Hector and of Achilles, had a considerable elevation, and, standing upon a low promontory, were visible far at sea. They were without architectural features or decoration, mere cones of earth and stones; terminated, as Homer relates concerning those of Ilos, Sarpedon, and Elpenor, by a monument like a column, which must have resembled the piers upon Lydian tumuli. It is questionable whether the trees which grew in later times upon the mounds of Protesilaos before Troy, and of Alcmaeon in Arcadia, were originally and intentionally there placed, and are to be deemed characteristic of such works. Those planted upon the tumulus of Augustus in Rome may certainly be referred to his individual desire. From the account given by Pausanias of the tumulus of aepytos at Pheneos, in Arcadia; from foundations remaining upon the island of Syme, and from later ruins at Kyrene--not to mention a well-preserved tumulus of very considerable dimensions, reveted with stone, which, from its situation in Algerian territory, might perhaps be ascribed to the Carthaginians, or even to the Romans--from all these examples, it is evident that such mounds, like the tumuli of Lydia and Etruria, were, for the greater part, elevated upon cylindrical foundations. But whether the interior were chambered or solid, whether the cone of earth rose directly from the earth or from a drum substructure, the tumulus appears to have been, in primitive times, the most customary form of monumental tomb for persons of high rank.

The common man was probably buried in pits, as at the present day, the grave being marked by an upright stone, with or without some slight ornament. Schliemann's discoveries in the agora of Mykenae show that, under certain circ.u.mstances, this procedure was adopted even for princes. The kingly importance of these sepulchres is a.s.sured by their position, and by the immense quant.i.ty of gold and valuables found within them. The decorative style of these objects dates them conclusively to the heroic age; but the a.s.signment of the different graves to Agamemnon and his a.s.sociates is a mere hypothesis.

[Ill.u.s.tration: Fig. 124.--Pyramid of Kenchreae.]

A pyramidal form was only in isolated instances subst.i.tuted for the tumulus. Of a pyramid, described by Pausanias as existing between Argos and Epidauros, there now remains a ma.s.s of masonry measuring 12 m. in the line of the diagonal. A second, near Kenchreae, between Argos and Tegea, is better preserved. (_Fig._ 124.) Its plan is oblong, 14.5 m.

long and nearly 12 m. broad; the two chambers of the interior are at present unroofed. The structure appears to have served as a common place of sepulchre for the fallen, and, at the same time, as a memorial of victory. This destination is also evident in two further pyramidal remains, in Laconia and near Lessa, which are described by Curtius and by Ross. The Greeks adopted both Asiatic and Egyptian forms for their funeral monuments; but in the construction of both tumulus and pyramid they introduced comparatively large chambers, early striving for ends foreign to those despotic lands:--a wise economy of material and labor and a gain of s.p.a.ce.

Mausoleums and sepulchres are always among the first traces of civilization, and the most ancient examples of architectural art. In Greece, however, there are contemporaneous remains significant of other purposes. Chief among these are the fortifications of towns, although in general these works enclosed only the acropolis, which contained the residences of the rulers and the sanctuaries of the people. The true age of these defences can by no means be surely determined. Not all Cyclopean masonry is to be attributed to the earliest ages of h.e.l.lenic antiquity, for this manner of polygonal jointing remained in use long after a time when cut and squared stones were generally employed. On the other hand, immense rectangular blocks, laid in horizontal courses, frequently occur in city walls which are known to be of the greatest antiquity and even to have been totally ruined in the historical period, such monoliths being regularly used upon corners, the jambs of gates, etc., where especial strength and independent firmness were called for.

When the surface of Cyclopean walls is perfectly smooth and exactly jointed, these may confidently be regarded as not of primitive antiquity; the erection of such masonry is a subtlety of greater difficulty than that of square blocks and horizontal beds. But walls built of enormous boulders, unhewn, and roughly piled up without calculation, the larger interstices being filled with smaller stones, are of extreme age. Such masonry appeared to later generations to be the work of giants, of Cyclops, and hence a name which might more fittingly be changed to Pelasgic than to Poseidonic, as suggested by Gladstone.

The walls of Tiryns (_Fig._ 125) are of such gigantic blocks--bulwarks mentioned by Homer and Hesiod, and admired in their ruins by Pausanias.

They are built upon a ridge of rock, which is over 190 m. long, only 70 m. broad, and elevated 10 m. above the surrounding plain. The masonry is from 7 to 15 m. thick; of its original height, estimated as 18 m., there remains from 10 to 12 m. The enormous stones vary from 2 to 3 m. in length and 0.9 to 12. m. in thickness. In its greatest breadth the wall is provided with galleries, roofed by projecting stones laid in horizontal beds and cut to the outline of a pointed arch. Such s.p.a.ces are provided with loopholes upon the exterior, and, without doubt, served as magazines and casemates. Within these fortifications must have stood the royal residence, famed in the legends of Heracles and Eurystheus; of it no recognizable traces remain.

[Ill.u.s.tration: Fig. 125.--Plan of the Acropolis of Tiryns.]

The walls of Mykenae are not of equally gigantic masonry, but are fully as old, and are especially interesting because of the city having been a complete ruin in the earliest historical times. Besides casemate galleries in the walls, there are in Mykenae a number of highly important gateways and portals; those of the fortifications at Tiryns were entirely destroyed, an inclined plane leading to the eastern side of the acropolis is there alone to be recognized as an approach.

[Ill.u.s.tration: Fig. 126.--Gate of the Lions at Mykenae.]

[Ill.u.s.tration: Fig. 127.--Smaller Gate of Mykenae.]

The doors were naturally of greater technical perfection than the long line of bulwarks; having been created for both admittance and defence, they required a certain constructive calculation, and permitted the employment of more exterior ornament. The simplest possible form of a gateway is the combination of three stones--the two jambs and the lintel--observable in two examples at Mykenae. (_Figs._ 126 and 127.) Such a construction had the disadvantage that the upright blocks could not be joined to the wall, and that the lintel, which necessarily lay clear for a considerable length, could not immediately receive the ma.s.sive continuation of the masonry above it. Notwithstanding the convergence of the jambs upon the great gate of Mykenae, the beam has a length of 4.6 m., with a span of 3.05 m.; the bottom of the door being 3.2 m. wide, and its height 3.25 m. A relieving gable was consequently constructed, similar to that common in Egypt during the age of the Pyramids, and to that described in the consideration of the Tholos of Atreus. A triangular opening remained above the lintel, by which the efficacy of the wall as a fortification was considerably impaired. The orifice was closed by one or two slabs, which did not press heavily upon the lintel; but they could not have been sufficient to escape fracture by heavy missiles, or to resist the blows of a battering-ram. The attack was therefore diverted from this vulnerable point by moral means. The panel received a certain consecration by some protecting sacred symbol being carved upon it--such, for instance, as a Gorgon's head--a recourse which was effective in times when the slightest desecration of a divine emblem was deemed more impious than the bloodiest deed of human violence. Such a carving has been preserved over the gateway of Mykenae, which has received its name from the lions represented upon it. As a work of sculpture, it will be considered below. The column between the animals has, however, a bearing upon the architectural forms of the epoch. It is the same shaft, diminishing from summit to base, which has been noticed upon the portal front of the Tholos of Atreus. A second gate of Mykenae resembled the Gate of the Lions, but was smaller and simpler. (_Fig._ 127.)

[Ill.u.s.tration: Fig. 128.--Portal upon Samos.]

[Ill.u.s.tration: Fig. 129.--Gate of Phigalia.]

The form of three blocks appears to have been soon changed, the wall itself serving in place of an especial jamb. The span of the lintel was decreased by two or four boldly projecting blocks as brackets. Examples of this development are offered by portals of Samos and Phigalia.

(_Figs._ 128 and 129.) But in the same measure as the danger from the great span of the lintel was diminished, that of the brackets being pressed downward and disjointed was increased. A third manner of covering the opening, by stones leaned against each other at an angle, was a still further advance. (_Fig._ 130.) When the side thrust could be well borne--and for this the walls were always sufficient--such a gable could support any pressure that could possibly be imposed, while allowing a great breadth of pa.s.sage. Finally, a triangular construction could be obtained by a gradual projection of horizontal stones, laid as they had been in so many instances for the relief of a lintel beneath them. This construction occurs in two varieties, differing in appearance, though not in principle: the projection of the horizontal courses of stone either began directly from the ground (_Fig._ 130), as has been noticed in the Tholos of Atreus (_Fig._ 122), or commenced at some height, the jambs being carried up vertically. (_Fig._ 132.) In both these varieties the line of the gable frequently appears concavely curved, as in the parabolic walls of the tholos, and the outline of a pointed arch was thus obtained. (_Figs._ 133 and 134.) In spite of their early familiarity with the abstract principle of the arch, as shown in _Fig._ 130, the Greeks refused to adopt the true arch, with its wedge-shaped stones, even in late historical ages, when they a.s.suredly were acquainted with its construction. An ill.u.s.tration of their feeling in this respect is given by the aqueduct adjoining the Tower of the Winds in Athens, where the semicircles are cut from monoliths.

[Ill.u.s.tration: Fig. 130.--Portal upon Delos.]

The influence of the gateways upon the masonry is evident from the more frequent adoption of the rectangular blocks, which had at first only been employed to give the portals an independent strength, both for the ramparts and for the out-works and protecting towers which these openings necessitated. Such a fortification, erected for the defence of a gate, still stands in Tiryns--the city to which succeeding ages ascribed the invention of tower-building (Pliny, _Hist. Nat._ vii. 56); it reaches a height of 13 m. Thetower which defended the gate of Mykenae was even larger. Homer mentions such structures at Troy, Thebes, and Calydon, and is also familiar with casemates and battlements. The latter are shown by paintings upon archaic vases to have been of the normal rectangular shape.

Schliemann's excavations in Mykenae have proved that in this city the agora was situated just within the princ.i.p.al gate. Some of the stone benches encircling the agora were found in almost perfect preservation; they were constructed of slabs standing erect in concentric rows to receive the horizontal seats. They lend a new confirmation of Homer's truthful characterization of locality, ill.u.s.trating a pa.s.sage which occurs in the description of the shield of Achilles, which describes the judgment scene upon the marketplace:

"On polish'd chairs, in solemn circle, sat The rev'rend elders."

[Ill.u.s.tration: Fig. 131.--Gate of Missolonghi.]

[Ill.u.s.tration: Fig. 132.--Gate of Messene.]

Though the remains of these prehistoric ages show in some degree the form of an ancient Greek acropolis, with its royal dwelling of courts and halls, and the sepulchral monuments before its gates, they are yet insufficient to complete even the main outlines of the picture by giving any understanding of the temple--that structure destined to become the ideal of h.e.l.lenic architecture. While the life and customs contemporary with the Homeric poems are, in other respects, represented with incomparable truth and distinctness, the epics are entirely silent upon this subject. It appears that the temples were neither of great size nor of artistic importance; among the ruins of Tiryns and Mykenae there are no vestiges of columns or entablatures. The symbolical images of the deities were placed upon cliffs, in caverns, among the branches of sacred trees, or in the hollows of their trunks, and simple altars were erected before them. Frequently the worship of a deity was merely connected with a grove, or with some other locality fitted by nature for this purpose, and was there performed without an image or other dead symbol. It was thus with the most primitive G.o.d of Greek mythology, Zeus of Dodona. When a building was provided at all, it was, in the heroic ages, restricted to the cella, a ceiled chapel of oblong plan, which stood in the centre of a consecrated area, the temenos. This original form--the whole of the primitive shrine--is recognizable even in the developed peripteros, as the kernel within the outstanding columns. It does not appear strange that we should be acquainted with so few of these chapels when it is considered that hardly greater traces remain of the entire architecture of the Teutonic races during the first seven Christian centuries. It is natural, in the development of civilization, that sanctuaries exemplifying different phases of advancement should seldom stand next to each other; after the destruction of the old, the new arises in its place, upon its consecrated site. Examples of such original cellas are not, however, entirely wanting. Several remains published by Dodwell and Stackelberg are to be explained as chapels. A structure upon Delos, designated by Thiersch as a tomb, is quite comparable to a columnless temple cella. There is less probability that the ruins upon Mount Ocha and near the village Stoura, upon Euba, were temples. They are chambers sheltered from above by slabs of stone, inclined like a gable. (_Fig._ 135.)

[Ill.u.s.tration: Fig. 133.--Gate of Thoricos.]

[Ill.u.s.tration: Fig. 134.--Gate of Ephesos.]

This method of roofing could not have been generally practised in early times, when simple and natural constructions utilized the materials at hand best adapted to the purpose. The builders, among the bald mountains of Euba, were forced to such a manner of covering their chamber by lack of wood. The south of the island produces no trees which could provide the timber for roof-beams; while, on the other hand, open quarries in the neighborhood furnished a kind of slate-stone which is easily split into large slabs like joists and boards. So clumsy a ceiling construction as that upon Mount Ocha was not natural in countries of dense forests, such as was the original home of the Dorians. In other parts of h.e.l.las than the rocky and sterile islands of the aegean, the chapels must have been roofed with wood. The most obvious considerations make it evident that ceiling and roof of the primitive cella were originally of wood. In the later marble architecture of Greece this a.s.sumption is confirmed by numerous reminiscences of timbered construction, sufficient even to explain the methods and form of the original carpentry.

[Ill.u.s.tration: Fig. 135.--Interior of a Structure upon Mount Ocha, Euba.]

A pitched and gabled roof seems to have been generally employed for these early structures. The horizontal ceiling might be sufficient for the changeless blue sky of Egypt, but could not suffice in Greece, where, in certain seasons, heavy rains were frequent, and even hail-storms not unknown. Still no land upon the Mediterranean was familiar with the great steepness of roof made necessary by the enduring snow and ice of the North. In colder climates the pitch of the covering was not only greatly increased, but all horizontal projections were avoided, and the upper surfaces of smaller members and mouldings inclined. The rafters required ceiling beams beneath them; because of the necessary support and jointing, they could not be placed directly upon the stone walls, and it was further desirable to support the summit of the triangle by a king-post. The ceiling thus provided stood in such relation to the roof that a beam tied together each pair of rafters, and was, consequently, so laid across the oblong enclosure that the ends reposed upon the side walls. Upon these horizontal timbers planks were placed which concealed the inclined roof. By this an independent ceiling was created; and, as the boarding was laid upon the beams and not fastened to their lower side, this gave rise to the formation of lacunae or long coffers. The ends both of the horizontal ceiling beams and of the roof rafters were visible upon the exterior: the latter, forming the eaves, projected beyond the wall, to further the shedding of water and to protect the sides of the building. As the upper surface of the roof had been so closed as to be water-tight, it is natural that this sheathing should have been carried around upon all sides of the projecting rafter ends. It was otherwise with the s.p.a.ces between the beams, which, being protected by the eaves, were not covered and masked by boards. The artistic instinct of the Greek would not permit him thus to conceal constructive forms when this was not rendered necessary by practical considerations. They received, on the contrary, an especial emphasis, that they might express their peculiar function with full force. Moreover, the closing of the aperture between the ends of the beams would have required the provision of other openings for light, as there were no windows in the walls of masonry.

This manner of roof and ceiling construction was generally employed in European Greece, being customary for palaces and dwellings as well as for the primitive temples. Open interstices between the horizontal beams existed in the hall of the royal dwelling at Ithaca. There can be no further doubt as to the development and original function of the metopes of the Doric entablature when it is considered that the Greeks, as late as the time of Euripides (_Iphig. in Taur._ 113), were familiar with the idea that it was possible to enter a primitive structure through these openings between the ends of the beams. The masking of the metopes would thus have been not only purposeless, but even detrimental; it was reasonable, however, to sheathe the ends of the beams themselves by small boards, which should at once protect and ornament them. The hewn extremities of such great timbers were rough and ugly; without covering, they would have been exposed to rapid decay. The simple decoration of three narrow strips of wood affixed to the ends of the beams was so customary in primitive carpentry that it became a typical motive in the later architecture of Greece. The chamfering of sharp edges of boards has been practised by the wood-workers of all nations. When two corners thus treated are placed together, there results a prismatic groove, which distinctly marks the edges of the separate pieces. Thus originated the primitive form of the triglyph, as the most natural and practical decoration of the rough-hewn ends of the ceiling beams by sheathing. The upper edges of the three strips were hidden against a plate beneath the rafters; the lower were covered by a continuous board, which united the various members of the frieze, and concealed any inexact jointing between the beams and the top of the wall. By placing the chamfered boards upright, an aesthetic advantage was obtained: a vertical line was repeated just before the conclusion of the entablature by the cornice, being thus emphasized in the midst of horizontal members. Other ornamental details were added, based, likewise, upon motives of the original wooden construction. The continuous strip affixed to the lower edges of the triglyphs was securely and visibly fastened. This was effected by several thick trunnels, so driven in from below that the heads were left protruding. Under the end of each beam the strip was doubled, to give additional strength where the wood was most weakened by perforation. The ends of the rafters were also sheathed, and brought into harmony with the frieze. The inclined eaves were covered with boards, and as these did not stand erect, like those before the ceiling beams, but hung from the lower sides of the rafters, there was particular need for an increased and distinctly secure attachment. The sheathing was consequently pinned by more numerous trunnels; and as every triglyph had been provided with a second strip, here a second board was placed under the end of each rafter. The projecting heads of these nails were called _guttae_ by the later Romans, but this cannot convince us that the peculiar form was intended as an ornamental petrifaction of hanging rain-drops: such a glorification of bad weather would have been foreign to the Greeks, accustomed to the clearest skies; and, for so primitive a construction, this explanation appears far-fetched. The imitation of rain-drops could nowhere have been more out of place than upon the inclined lower side of the eaves; drops might, perhaps, hang from the front edge of the cornice, but never upon its under slope, which rain could not even wet. The construction of an original work of carpentry thus provided the motives of the Doric entablature--navely expressing the advance from the roughest practical necessity to high architectural perfection. In the apertures between the beam-ends, or metopes, and in the open triangle of the gable, were placed votive offerings, which there found a secure and sheltered stand, heightening the exterior importance of the work. In small chapels this interference with the openings for light could have been of no disadvantage. The gable was closed by a boarding, which hid from view the rough inner construction of the roof. This veil, the tympanon, was placed behind the triangle formed by the outer cross-beam and rafters, as the ceiling had been laid above the other horizontal timbers. The low gable thus naturally developed upon the front; and in later times, when the votive offerings had been exchanged for sculptured figures, formed a most characteristic and imposing feature.

The effect was heightened by the partly protective, partly decorative, painting of all the wooden surfaces. Red and blue appear originally to have been the chief colors; the former, in a dark shade, being used for the sheathing of the tympanon, the latter for the triglyphs and other members. Upon the bands were figured ornaments, most of which had developed from Asiatic prototypes; they consisted of the meander, anthemions, and the woven ribbons, etc., observable upon a.s.syrian sculptures and upon the archaic bronzes and vases of Greece and Central Italy. The extended polychromatic treatment of the marble temple is doubtless a reminiscence of this painted wood. Without such traditions, it would have developed differently: upon a structure of stone it would have been less restricted to the frieze and cornice.

The entablature had thus far advanced without connection with that most n.o.ble work of architecture--the Doric column. The shaft and entablature of the style were not created in connection or simultaneously; the forms of triglyph and mutule are not a growth from the columnar root, but rather prove the Doric frieze and cornice to have been the primitive h.e.l.lenic expression of roof and ceiling, which preceded the column, even as the plainest constructive necessities precede ornament. The peculiar wooden character of the entablature could exercise no important influence upon the shaft. If the existence, in heroic times, of the peripteros, the temple with outstanding columns, be denied--and of such structures there is not a vestige--it cannot be supposed that columns existed at all. Interior supports of wood are, indeed, mentioned by Homer, and engaged shafts formed part of the facade of the Tholos of Atreus, and were represented upon the relief over the Gate of the Lions in Mykenae; but between these and the Doric column there is a distance only to be explained by the a.s.sumption that Asiatic influence was paramount, if not exclusive, in the architecture of the heroic ages of Greece. Though it is possible that rudiments of the Doric echinos may be recognized in the upper tore and scotia of the engaged columns of Mykenae, it is yet evident that the turned-work of these members resulted from a wooden prototype, and that the overladen decoration of the shaft, in its style, is due to familiarity with a sheeting of beaten metal--_i.e._, to Phnician artistic traditions. That the forms of the entablature were not created for the peripteros appears from the circ.u.mstance that the metopes lose their value as windows by the change of plan, and leave the cella without openings for light and air when surrounded by columns. With the appearance of the peripteral temple, the Doric entablature, which upon the oblong chapel had been the natural expression upon the exterior of roof and ceiling construction, became a functionless ornament, needing, as will be seen, many changes to bring it into harmony with the outstanding colonnade.

The development of the Doric column is not perfectly clear; it is more than probable that it was not wholly autochthonic and primitive Greek, like the entablature of the style. Its princ.i.p.al part, the shaft, was certainly imported. No prominent architectural feature can be deemed newly invented that has been in common usage in a neighboring and accessible country for centuries. The Doric shaft, with its characteristic diminution and channellings, was known in Egypt more than a thousand years before its introduction into Greece, as proved by the monuments of Beni-Ha.s.san. Commercial intercourse had existed between the two countries for centuries, and it cannot be a.s.sumed that the Greeks had not seen Egyptian works of architecture; they could not have arrived at precisely the same results by independent invention. It would rather be difficult to conceive how the receptive Greeks could have refused all instruction from the neighboring people, so far in advance of them for centuries after the Trojan war. Eight-sided drums have been found at Bolymnos, and an octangular shaft at Trzen; but these isolated instances offer no proof that the development of the channelled shaft from the square pier was effected in Greece in the same manner as had been done fifteen centuries or more previously in Egypt.

The genius of the Greeks, however, always showed its independence when the artistic perception of the neighboring nations had been at fault or defective. It was impossible for them to rest content with the termination of the so-called Proto-Doric columns of Beni-Ha.s.san. A simple plinth upon the upper end of the shaft was insufficient; it left without mediation the contrast between the forcible upright line of the channels and the long level of the epistyle. Some interposition was necessary between the vertical and the horizontal members, and a moulding of inclined outline was best fitted to fulfil this natural requirement, which almost appears to be an aesthetic law. The abacus plinth was retained as the transition from the circular drums of the shaft to the broader oblong of the lintel. The oblique and projecting member between the two, the echinos, was a link connecting the plans, as well as the directions, of column and entablature. The perfectly straight outline of an inverted cone was rarely employed in Greece for the echinos; a stele of Artemis Brauronia upon the Athenian acropolis, shown by inscriptions to be of great age, is an isolated instance. This rigid line was early exchanged for a curve, which, in its advancing stages of refinement, became one of the most characteristic features of Doric architecture. The moulding seems, at times, to have been ornamented with painted leaves, which, in the Ionic echinos beneath the roll, was changed, in the manner peculiar to that order, from the colored indication to carving. It is not certain whether this floral decoration was generally adopted, or existed only in the isolated instance by which it is known--the so-called Temple of Theseus. Upon the translation of the wooden construction to a stone entablature, which resulted in a narrow intercolumniation, the base was given up, and the upper step of the stylobate was regarded as a common plinth.

It appears that the employment of columns connected with temples commenced, in Greece, in the manner observed upon the rock-cut tomb facades of Egypt and Lycia, and the chapels of Mesopotamia and Phnicia: two columns were placed within the open front, between the projecting side walls; that is to say, the temple was _in antis_.

The next step was the removal of these side walls, or parastadae, columns taking their place in the corners before them, and the _prostyle_ temple was thus obtained. These changes rendered several important alterations necessary. They caused a new wall to be erected before the interior of the cella, the naos, the colonnade of the front thus acquiring the nature of a portico, the p.r.o.naos. The jambs of the door in this wall were so inclined as to diminish the span of the lintel, the frame receiving upon its upper corners the stepped ears, or parotides, customary in Western Asia. A new member of the entablature was needed to replace the omitted wall and provide a bearing for the ceiling cross-beams--namely, the epistyle. It is possible that this member, distinctly separated, existed before the change, but it certainly was not necessary. The division of the cella into naos and p.r.o.naos finally altered the position of the front ceiling-beams; in the naos they lay, as before, resting upon the side walls, but in the p.r.o.naos they were placed lengthwise--from the columns to the newly erected division wall.

Besides improving the construction of the portico ceiling, this greatly added to the beauty of the front entablature: epistyle and ceiling-beams would otherwise have lain upon each other, in the same direction, but from this change resulted the frieze of triglyphs and metopes upon the front, as upon the sides. The gain was not effected without a difficulty arising in the frieze above the end of the side wall and the corner column, the outer ceiling-beam of the p.r.o.naos thus lying in its length upon the epistyle without the formation of a metope. And here the constructive truth was first sacrificed in favor of the exterior appearance: a cube, standing above the corner column, took the place of the outer beam, and the continuous alternation of triglyphs and metopes was carried out.

Having so far deviated from logical construction, the desire for an harmonious treatment of the exterior led to other and greater changes.

The dead-wall of the rear had had no part in the development of the frieze, and appeared intolerably bare. This deficiency could hardly be overcome otherwise than by a repet.i.tion of a portico upon the back, creating the epinaos, and carrying the entablature of triglyphs and metopes around the entire building, thus perfecting the _amphiprostyle_ temple.

The more these alterations were made in favor of the exterior appearance, the more was the original structure dismembered. The extreme boundary of possible concessions was attained, and, at the next step, the entablature, translated into stone, separated itself entirely from the construction and became an applied ornament. In one stride the ultimate type of the h.e.l.lenic temple was determined, by carrying outstanding columns entirely around the cella,--the building became a _peripteros_.

It is probable that these extensive alterations took place almost simultaneously, and were adopted at once for the most prominent shrines, while the preceding varieties--the temple in antis and the prostyle and amphiprostyle temples--though their entablatures were also executed in stone, were only employed in subordinate positions. With the heightened importance of the decorative exterior the monumental significance of the temple rose above the mere necessities of a chamber for the sacred image. The structure acquired equal solidity in every part exposed to view. It was built of a h.o.m.ogeneous material. The timbering of roof and ceiling was hidden by the stone symbols placed before the ends of the rafters and beams; the entablature was allowed an independent freedom of development and proportion. The heaviness of the material made it necessary to diminish the voids and increase the solids of the supports as much as was feasible. The stone shafts were allowed a greater diameter and placed more nearly together than when, as was the case in Etruria at a much later period, their burden had been of timber. The stone cornice, which was not as high as the epistyle, could not span the same clear width, and called for a second support over the intercolumniations,--a further triglyph. This was the more acceptable, as the appearance of the frieze was improved by its adoption; the breadth of triglyph and metope became nearly equal and better proportioned, their alternating rhythm more pleasing. The metopes, having upon the peripteros no importance as windows, were closed by thin slabs, which added to the unity and imposing force of the edifice.

It is surprising how faithfully the traditional forms were still retained, even to the smallest details, while they yet received a truly artistic conventionalization and those proportions which make the Doric temple the grandest and most perfect monument of architectural history.

It is probable that the completed peripteros existed as early as the seventh century B.C. The first steps of advance were rapidly made, and may, perhaps, be referred to the ages immediately preceding. It would indeed be interesting to know when, where, and by whom the incomparable design was perfected which gave to the world its proudest edifice; but it must suffice to understand the intentions of which the Doric temple was the final result.

[Ill.u.s.tration: Fig. 136.--Corner Elevation of the Middle Temple of the Acropolis of Selinous. Restoration.]

Semper has suggested that a canopy-like roof, supported by columns, was placed above and around the small temple cella to increase its extent, and, at the same time, to express its power and sacredness by that oldest symbol of terrestrial and celestial authority. This attractive a.s.sumption does not interfere with the theory of the previous development of the temple in antis and the prostylos, or with the historical considerations based upon the appearance of an imperfect peripteros centuries before in Egypt. The cella and outstanding columns rose from a stepped foundation, the crepidoma, the kernel of which, the stereobate, was formed of ma.s.sive walls, or, when possible, of the native rock. The blocks were too high for human steps, and are not to be conceived as stairs. Such an ascent entirely surrounding the temple would have been purposeless, and contrary to the isolating character of the crepidoma. They formed a base, such as is displayed in an exaggerated manner by the Mesopotamian sanctuaries, where, however, the chapels elevated upon the gigantic terraces were small in proportion to the substructure. In buildings of greater dimensions, the few and ma.s.sive steps serving as the base of the Greek temple were increased, not in number, but in size. They were thus always proportional and fitted to their function as a foundation. Accessible stairs from all sides would have given a pyramidal effect to the lower part of the composition; while, at the foot of the upright supports, the horizontal line should rather be emphatically p.r.o.nounced. Smaller intermediate blocks were provided for the ascent to the temple, thus made possible only upon the front. The upper step, the stylobate, was, as has been said, the common plinth, the columns being without base-moulding, and, consequently, without individual functions or isolated independence. The comparatively narrow intercolumniations were the better pa.s.sages from this absence of projections at the foot of the columns. The powerful shafts were doubly modified by the diminution and by the entasis. The first refinement found its model in the natural contraction of all ascending bodies; a greater strength is needed below because of the increasing weight. To this must be added an optical motive: every diminution modifies the perspective effect, increasing the apparent height or distance of bodies thus bordered by lines slightly converging, though apparently parallel. The entasis was entirely decided by such optical considerations. It overcame a deception, resulting from the diminution, which makes a straight-lined cone of very steep sides appear of slightly concave outline. The shafts usually had twenty, in a few instances sixteen, channels, of nearly elliptical profile, separated by sharp arrises. As may be seen in unfinished temples, these grooves were not executed until the last stone of the building was in place, that the chipping of the delicate edges by the imposition of the drums or blocks next to them, and by other accidents during the process of building, might be avoided. It was only upon the capital that the channels were cut in advance, as a guide. To avoid the chipping of this stone, it was necessary to prevent its sharp lower edges from resting directly upon the top of the drum beneath it. To this end a diminutive step, a scamillus of smaller diameter, was turned upon the bottom of the capital block, or the same effect was attained by slightly slanting off and increasing the right angle of its lower edge. It was contrary to the artistic feeling of the Greek architect for constructive truth to mask even this slight necessity by priming and painting. It was, rather, made more distinct by increased size and a characteristic profile, in some instances even by a repet.i.tion of the incision. The upper end of the shaft was thus distinctly separated, notwithstanding the continuous channellings, and was related to the capital as the mediating neck of the column, the hypotrachelion. The echinos began its projection with several annulets, which still more definitely marked the junction of the capital with the shaft. It would be difficult to decide whether these mouldings were reminiscences of the binding-ribbons upon the necking of Egyptian floral columns. They were not placed beneath the echinos, but upon it, and consequently follow the curved profile, enlarging concentrically with its projection. The Doric capital, among all capitals that we know, attains the highest aesthetic perfection by its fulfilment of the requirements of a transitional member: by the proportion of its projection, and especially by its expressive and characteristic curve, which rises from a firm and almost straight line to the decided turn beneath the abacus. The outline is more elastic than a simple oblique angle, more vigorous and capable of resistance than the concave curve. The echinos provides the requisite projection; the abacus upon it forms the second transition from the circular plan of the shaft to the rectangle of the entablature. In the Doric style this upper half is about the same height as the echinos beneath it, while in the capitals of other orders the curved members of circular plan have been developed at the expense of this plinth, which is dwarfed to a thin plate.

It was first noticed by c.o.c.kerell in 1829 that the axes of the columns surrounding the cella are not vertical, but lean inward. This peculiarity was chiefly adopted to counteract an optical deception, resulting, like the deviation which led to the entasis, from the diminution of the shafts, making these, when perfectly upright, appear inclined away from the neighboring wall and from each other. The deception is particularly felt upon the corner shafts; these were corrected to lean in the direction of the diagonal, and decided the inclination of the columns of the front and side. The absolute deviation from the vertical is very slight, about 1-150th of the height, and by no means makes the inner sides of the diminished columns parallel to the wall. The inclination was effected by the irregular cutting of the first block, which was lower within than without, being so formed that the surface of its base was not circular, but slightly elliptical. All the succeeding drums had perfectly round beds, and consequently slanted in the manner decided by the first. The contact of these stones of the shaft was restricted to a narrow rim upon the exterior of their plan. In their centre they were steadied by an encased dowel of wood, the form of which is known from the remains of the Parthenon; this served as a pivot for the grinding of one block upon the other.

The stone beams of the epistyle lay from axis to axis of the columns. In buildings of great dimensions several slabs were laid side by side as lintels, each having the entire height of this member, which, as forming the conjunction of the columns, may be conceived as a representative of the wall. The outer surface of the epistyle block was carved upon its upper edge with the tainia and trunnels, described as securing the triglyphs of the original timbered entablature. The forms of these details show the great reverence with which the primitive wooden prototypes were imitated, while, at the same time, they were fitted to be cut in stone in a far more artistic manner than were the direct copies of carpentry observed in Lycia. The slits of the triglyph terminated at first in elliptical lines, which became, in the decline of the style, straight and horizontal. The triglyphs themselves were so distributed that one was placed over each column and one over the centre of each intercolumniation. An exception was made at the corner, where the triglyph could not be placed in the axis of the shaft, being needed for the support of the angle. It would be contrary to the open and non-sustaining character of the metope for this to be a.s.signed to a position so constructively important. Vitruvius, regardless of this consideration, recommends that the corner triglyph be placed in the axis of the column beneath it, like all the others; but only one debased instance is known where this occurs--the so-called Temple of Demeter at Pstum. The disturbance of symmetry which resulted to the frieze by the removal of the corner triglyph from the axis was counterbalanced by the metopes being made slightly larger, and especially by the outer intercolumniations being greatly diminished in width. This last step was also desirable from other considerations, notably because the dark background of the cella caused the openings between the inner shafts to appear narrower than the free and light s.p.a.ce between those of the exterior.

[Ill.u.s.tration: Fig. 137.--Entablature of the Parthenon.]