26.

ARCHITECTURE OF THE ANCIENT MAYA¹

ByW. H.n/aHOLMESn/an/an/an/a

The Maya Race. At the period of conquest the Maya tribes, occupying the peninsula of Yucatan and considerable portions of neighboring territory to the south and west, are said to have comprised in the neighborhood of 2,000,000 souls. Today they are distributed over nearly the same area, but are reduced in numbers, it is estimated, to less than 500,000, half at least of whom continue to speak the Maya tongue in its purity. At the north where there has been much infusion of Spanish blood the race has been largely modified and an interesting and very homogeneous half-blood people has sprung up; but in the interior many of the tribes are of nearly pure blood and retain a strong spirit of independence. It is said that some bands have never been fully conquered and they practically substantiate the claim by holding the temples of their fathers by force of arms, defying all comers, whether white or red. . . .

In the culture scale this people stood at the head of the American tribes. They were still, properly speaking, barbarians, but in several respects seemed to be on the very threshold of civilization. Their status may be compared to that of the Greeks and Egyptians immediately preceding the dawn of history, and we may assume that they were, as measured by Aryan rates of progress, perhaps not more than a few thousand years behind the foremost nations of the world in the great procession of races from savagery toward enlightenment. It is certain that they were already employing a rude system of historic records and were the only nation on the western continent that had made any considerable headway in the development of a phonetic system of writing. Their hieroglyphics occupy a place, not yet well defined, somewhere along the course of progress from pictograph to letter, and are consequently difficult of interpretation. There is no doubt, however, that an age of literature was actually though slowly dawning in America when the shock of conquest came. . . .

Today the chief reminder of the great past of the Mayas is the crumbling remains of their architecture, but remarkable advance had been made in several other arts not embodied in such durable materials. They made paper of maguey, and their books, several of which have been preserved and are now in the libraries of Europe, show advanced skill in pictographic and glyphic writing, and a fertility of imagination hardly paralleled among the known primitive races of the world. . . .

Monumental Remains. Maya architecture, with its associated sculpture and painting, constitutes the best remaining index of the achievements of the race. The 70,000 square miles of Maya territory are so dotted with the ruins of towns and cities that the traveler is seldom out of sight of some mound, pyramid, or other massive structure. The preservation of these remains is wonderful, considering the four hundred years of decay and destruction through which they have passed. There is hardly a modern village or town on the peninsula of Yucatan that is not built in some part of materials derived from the ancient structures. Yet the work of demolition still goes on, and presently, unless the Mexican government takes adequate measures to preserve them, the traces of a conquered race and its strange art will exist only in books. Nature has vied with man in the work of leveling the noble monuments with the ground. The luxuriant vegetation which envelops the ruins sends a multitude of strong roots deep into the masonry at every vulnerable point; growing rapidly, they act like wedges, separating masses and aiding gravitation and the elements in their never-ceasing efforts at destruction. . . .

Building Materials. The nature of the materials at the disposal of a people inclined to building exerts a profound influence upon the results achieved. Stone of somewhat decidedly favorable qualities would seem almost essential to greatness in the art of architecture. The Mayas were especially favored in this respect. The peninsula of Yucatan is composed of massive beds of limestone, homogeneous in texture and easily cut, even with primitive tools. Nature had not only supplied the stone, but it had in some measure prepared it for building. Although the land is approximately a plain, it is still in a small way broken up by low ridges and steps, and by sinkage into underground channels. The forests, growing densely everywhere, have broken up the surface beds, giving great quantities of loose stone immediately available to the builder and directing the way to the opening and working of quarries. The presence of unlimited supplies of limestone together with timber made the burning of lime an easy task and this product was extensively employed. The Yucatec stone mason had, therefore, every necessary building material at hand, although he still lacked, in a great measure, materials suited to the manufacture of quarrying and cutting tools. Cherty seams or masses of indurated limestone, occurring in many parts, served for the ruder tools, and picks and chisels of special hardness were probably brought in from a distance. Copper chisels are occasionally found as far east even as Cozumel, but if used at all in the dressing of stone they must have taken an unimportant place in the work on account of the rarity of the material. I had no time to seek the quarries from which stone was obtained in Yucatan, but had the good fortune to come upon excellent examples in Oaxaca. Careful descriptions of these will be given in part II of this paper.

Mortar, made of lime and sand, and cement-like mixtures composed of mortar tempered with gravel, pounded stone, etc., were extensively used, and their durability is remarkable. Numerous floors and roofs are still preserved, and many fine examples of stucco modeling have withstood the destructive effects of the weather for four hundred years or more.

The builders made very considerable use of wood, which, considering the inferior grade of tools available, was cut, hewn and carved with much skill. Wood must have been extensively used in connection with the great stone buildings, as in doorways, in closing spaces between structures now disconnected and in various enclosures and barriers. There is no doubt that pliable vegetable growths, such as poles, bark, vines, twigs, etc., used in textile or semi-textile combinations, were very fully employed in ordinary domestic structures as well as in less pretentious buildings of other classes pretty much as they are today.

Transportation. The gathering of stones and the cutting out of masses from the living rock were followed by transportation, a most tedious and laborious task for a people without beasts of burden and probably without many of the effective transporting devices known to more advanced peoples. The work of carrying the earth, mortar and stones used in hearting the pyramid of the Castillo at Chichen or the triple-terraced pyramid of the Palace at Uxmal was of itself a great undertaking, but the transportation of the countless stones for the facing of both pyramid and superstructure and the lifting of the larger masses employed in columns, jambs, pillars and the like to heights reaching in cases nearly one hundred feet, required strong hearts and hands and a controlling power of exceptional vigor and permanence. The Yucatec Mayas did not, however, undertake to employ stones of enormous size, as did the ancient builders of Mexico and Peru. No block or mass observed was estimated to weigh more than six or eight tons. . . .

Masonry, Stucco Work and Painting. The masonry comprises, in general, hearting and facing. The former consists of earth, mortar and stones variously combined and usually forming strong, well-compacted bodies. The latter consists of stone cut or uncut and laid up, with few exceptions, in excellent mortar. Where the stones were accurately cut, little mortar appears in the face of the wall, but it was freely used in the hearting, and when the facing stones were deep they were dressed somewhat smaller behind, and set back in the mortar as a tooth in its socket. In the facing of many walls, however, the stones were very shallow—often mere tile-like slabs—and had but slight hold upon the body of the hearting.

In those centers of building operations where the limestone was readily worked and of fine, even texture, the facing is well cut, and the wall surfaces are in general so even and true as to stand the test of the square and plumb line; but in localities where the stone is uneven in texture and quite hard, or in provincial sections where building was not carried to a high degree of perfection, the facing is rarely well dressed, save about the doorways, arches, corners and especially exposed parts. Rough surfaces were very generally evened up with plaster.

A remarkable feature of these structures is the great thickness of the walls, and especially the extraordinary massiveness of the masonry above the spring of the arches. This is clearly shown in several of the sections inserted in the following pages. Where, for example, the outer wall is three feet thick and the arch within is ten feet wide, the mass of masonry thickens upward from three feet at the base of the arch to eight feet at the ceiling level, and in an inner wall, widening both ways, to thirteen feet, so that two-thirds or more of the space included in the upper half of the structure is solid masonry. The roof is often very thick, thus greatly increasing the bulk, and it seems a marvel that collapse from mere weight has not been more frequent than seems to have been the case. To all this bulk were added, in many instances, massive false fronts or colossal roof-combs laden with ornament. So strongly knit is the masonry, however, that but for the decay of wooden lintels, most of the great façades now in ruins would have been very fully preserved. I have computed that a single-chamber structure, with walls of usual thickness and with average arch space and roof mass, would have two-thirds of its bulk solid masonry, which looks like a lavish waste of space, material and labor. If we take the measurements of the Governor’s Palace at Uxmal, given by Bancroft, we find by a rough computation that the structure occupies some 325,000 cubic feet of space, upwards of 200,000 of which is solid masonry, while only about 110,000 feet is chamber space. If the sub-structure be taken into account, the mass of masonry is to the chamber space approximately as 40 to 1.

Notwithstanding the success of these Maya masons in erecting buildings capable of standing for hundreds of years, they were yet ignorant of some of the most essential principles of stone construction, and are thus to be regarded as hardly more than novices in the art. They made use of various minor expedients, as any clever nation of builders would, but depended largely on mortar and inertia to hold their buildings together.

One of the most elemental essentials of good work is the systematic breaking of joints in laying one course of stones over another. This idea had hardly been grasped, as it not infrequently happens that a seam or succession of joints is connected almost directly from base to summit of a wall, and at corners, within and without, and about doorways the stones are not bonded at all and are free to fall out as soon as the mortar gives way. The only possible explanation of this condition of the work that occurs to me is that the habit of treating the stones of a wall as so many elements of a mosaic pattern has tended to retard progress in the direction of what is sometimes called scientific construction. It will readily be seen that in carving and laying the stones of a geometric design, as a line of fret-work or of snouted masks, it would be extremely inconvenient to adapt the shapes to any system of jointage, and indeed such a thing would be out of the question.

Another considerable element of weakness in many of these structures was the employment of veneered facing over large surfaces without sufficient headers or long transverse bonding stones. The tendency to break away, even with very thin stones, is in a measure counteracted by giving the back a bevel almost from the face, thus allowing the mortar to come well forward in strong tongues nearly to the surface. In some cases the facing has fallen in a body from considerable areas, exposing the hearting, which presents a remarkably even surface as if built first as a rough wall to be faced up afterward at the convenience of quarrymen and stonecutters. . . .

It was the practice to finish plain walls in plaster, often rather roughly applied, and nearly all surfaces, exterior and interior, where effect was of consequence, were finished in color. Very often plain surfaces in corridors and important apartments were embellished with graphic subjects, ornamental designs, devices and glyphs in brilliant colors. Sculptures in the round and in all degrees of relief were also colored with great care and elaboration. The range of colors is wide, including black, white and various shades of green, blue, red and yellow. Their composition has not been made a matter of study, but they probably include both mineral and vegetal substances.

As to the methods of manipulating stone, mortar and color, little is definitely known, save through a study of the actual remains. Unlike the Egyptians, who pictured almost everything relating to their own arts and avocations, the Mayas give us but a few hints of these things, both graphic and plastic art dealing almost exclusively with sacerdotal subjects which furnish, incidentally only, hints of practical things. A notable exception is found in one of the Bodleian codexes, where various domestic episodes and illustrations of the practice of ordinary arts are given. Stone, when required in large bodies, was cut out of the mass, probably with rude stone picks, and flaked and pecked into shape at great expense of labor. Very generally the dressed surfaces show the chisel and pick or hammer marks. . . .

I am not able to say with certainty to what extent the dressed surfaces of the stone in the walls of the buildings were ground or polished, but it seems natural that abrading processes should have been generally employed. Hammer stones, sledge-heads, picks and chisels of hard stone are found, but not in the great numbers that might be expected. They are not superior in make to like tools employed by the average American savage, and none of them seem capable of having made the marks illustrated. . . . We are thus led again and again to wonder whether it is not possible that metal tools were used and that traces of their existence, save in the sculptures produced, are wholly obliterated by time.

The lime-burner and the color-man were most important auxiliaries of the Maya builder. Mortar was used in enormous quantities and manipulated with great skill, and the same may be said of color; and the trowels and brushes employed were no doubt such as primitive people usually devise. It should be observed that it was a common practice all over the Mayan, Oaxacan and Nahuatl territories to finish architectural ornaments, statuary and glyphs, where the stone was not of the finest quality and susceptible of high polish, in thick enamel-like coatings of varied colors which adhered with wonderful tenacity to the stone surface and were polished down with the utmost care, not reducing the relief or distinctness of even the shallowest sculpturings, but being made by skillful manipulation to emphasize and refine these features.

During the great days of temple building the scenes about and within one of these Maya cities must have been exceedingly animated and novel. The hosts of people planning and directing the work; quarrying, cutting, transporting and lifting the stone; burning lime, carrying water, mixing mortar, hewing wood, preparing paints, and engaged in the work of building and decorating, must have furnished scenes in striking contrast with the desolation of the dismantled and forest buried cities of today.

Substructures. The ancient cities of Yucatan were built on plains or on comparatively level ground and were without the advantage of bold natural features, but art largely supplied this want, and no nation of builders, save possibly the Mound-builders of the Mississippi Valley, has ever equaled this people in the number, variety, and size of its terraces and pyramids; however, there appear to be no pyramids that are mere pyramids, no terraces that are mere terraces; all or nearly all were constructed to support buildings, altars, or idols, and their diversity of size, contour and position give striking and picturesque results. Usually the substructures are square or rectangular in plan. The largest reported in Yucatan is upward of 500 feet in length and width, and the height of the loftiest reaches nearly 100 feet. The sides slope at various angles and some are practically vertical in whole or in part; many rise in steps, the succeeding terraces or platforms being of equal or unequal height and of varying horizontal extent. As a rule they are or were faced with stone which was dressed smooth or plastered. In the finer structures the terrace faces were paneled in hewn stone or embellished with moldings or with sculptured or stucco reliefs. The corners were often rounded and formed of large and specially sculptured stones. They were ascended by substantial, generally steep and wide stairways on one or more of the sides. The interior mass was constructed of earth and stones or mortar and stones usually forming a solid or well compacted body. In cases, however, this pile was not depended upon as a sufficient support for the superstructure, and foundation walls were carried up from considerable depth or from the ground level. The upper surface was generally floored with cement, though paving with slabs is occasionally seen. No doubt these piles were in cases the result of a long period of growth, and it probably sometimes happened that when a loftier structure was desired ground floor apartments were filled in solid with rubble or masonry, giving firm foundations for a second story or superstructure. In some cases the exterior of vertical-walled substructures was enforced by abutting masonry entirely encasing the original nucleus and giving the effect of an ordinary sloped terrace or pyramid.

In Fig. 1 a few examples of terraces and pyramids are presented in outline. The variety of contour is very great and it may be said that no two specimens are alike. The most unique form is that of the Temple of the Magician at Uxmal, h, which is oblong in plan and rounded at the ends; the loftiest is that of El Castillo at Chichen, which is of the stepped type seen in f; while the grandest and most diversified in contour is that of the Governor’s House at Uxmal, shown approximately in g. In a, b, c, d and e we have what may be regarded as the most common forms. The substructure of the Temple of the Tigers at Chichen, i, is peculiar only in being associated with the great wall of the Tennis Court or Gymnasium.

Stairways. The stairways of the Maya pyramids (see Fig. 1), share in a large measure the boldness and magnitude of the constructions with which they are associated and of which they form an essential part. A single stairway would have afforded all necessary access to the lofty summits, but it is not unusual to find two flights, and three or even four flights are known leading to the same temple, and each built on an equally grand scale and finished with like elaboration. All are exterior and centrally placed, leading directly up the face of the pyramid. Usually they are wide and bordered with some kind of solid balustrade. The favorite design for the rail is a colossal serpent, the head with wide open mouth and protruding tongue extended upon the ground, the body, appropriately carved, extending to the summit. In Yucatan the steps are neither high nor wide, averaging perhaps a foot in rise and a little less in tread. The pitch is thus 45 degrees or more. The stones used are generally rather small and not very smoothly dressed or well fitted, and it is probable that all important flights were finished in cement and color. The stairway usually conforms to the slope of the pyramid or shows only a little relief therefrom, but occasionally the angle is reduced, throwing the base out from the base of the pyramid, suggesting the graded way of the Mound-builders. Where associated with a vertical or very steep ascent or a series of rises, it is built out solid or carried over arches, as in the Palace at Chichen. Interior stairways are not found in pyramids and are rare and unimportant in the superstructures; the winding stair in the round tower at Chichen and the several narrow flights in Palenque being perhaps the best known examples. The most interesting stairways met with on the voyage are in the courts of the Palace at Palenque. Here large stones were used, on the faces of which are glyphic sculptures. The evolution of the stairway in its various forms was probably simple and natural and seems to present no problems—no obscure passages—worthy of particular discussion.

Superstructures. I cannot undertake in this place to give more than a mere outline of the leading features and characteristics of the many buildings visited. A few only of the larger structures are built on the ground level of the site, though many are but slightly raised. In some cases the terraces and pyramids have developed in sections by the addition of parts needed to accommodate new buildings, and again, as already mentioned, the supporting pile has been built and completed at once to receive the superstructure upon its summit. The plan of the building in the one instance is often composite and irregular and in the other is simple and regular. A number of buildings may occupy a single large foundation mass, and buildings or chambers may occur independently of each other on different levels of the same substructure. In a few cases only, as at Tuloom and Palenque, do we find a second story built above a lower story which has not first been filled up. Where several buildings of different levels are associated, the lower tier stands against the base of the pyramid, the second, back of this, occupies the first terrace, and the third, back of this again, is on the second level.

The ground plan is usually rectangular, two or three examples only of round houses having been reported. Large buildings of independent position are mostly rather long and narrow, the width having been limited by the difficulty of widening the arch where one or two tiers of rooms are used, and of securing light in the inner chambers of multiple tiers, since the upper wall and roof are never perforated. In detail the plan of large buildings, even the most complex, shows little more than a mere multiplication of the simple rectangular cell unit. Exceptions are found in the Round Tower of Chichen and in the corridor-like galleries of Palenque, and, no doubt, also in several multi-columned structures now too much ruined to be fully analyzed.

The buildings usually classed as temples are not large and are generally squarish in plan. They have from one to four rooms. When the rooms are multiple they are so arranged as to indicate pretty clearly a specialization

Fig. 1. Examples of terraces and pyramids, superstructures omitted.

of use. The two essential features in such cases are an outer chamber or vestibule and an inner chamber or sanctuary. The vestibule is entered by a plain, single doorway in inferior structures, and by a wide doorway divided by columns or piers in those of the better class. Usually it extends entirely across the front of the building. The fully developed vestibule is a modified outer chamber, and is characterized by multiple exterior doorways separated by piers or by columns, giving the effect of a portico closed at the ends. The sanctuary is mostly entered by a central doorway, though lateral entrances are sometimes provided. Additional rooms are arranged about the sanctuary at the right or left or extend behind it, as in the case of El Castillo at Chichen. Most of the Palenque temples have an outer apartment of the vestibule type, entered between piers; and a back apartment enclosing a small roofed sanctuary, entered by a single door. Small rooms are placed at the sides. When there is a single chamber only, which is not uncommon, it exhibits frequently the characters of the vestibule. Altars are rarely found, the only example met with being in a small temple on the Island of Mugeres. In Fig. 2 a series of temple plans is given, illustrating the remarks just made. I take it that, if these varied structures are properly called temples, any apartment or any suite of apartments in any building may have served the purposes of a temple, though the term may not with propriety be applied to any structure not showing peculiarity of placement or style, in which there is not some variation from the mere grouping of simple chamber units.

Fig. 2. Specialization of the ground plan of Maya temples. a. Single-chamber building with plain door. b. Single-chamber temple with wide doorway and two square columns. c. Two-chamber temple, the vestibule with wide doorway and round columns, and the sanctuary with single plain doorway. d. Two-chamber temple, the vestibule with simple doorway, and the sanctuary with three doorways and a low altar. e. Four-chamber temple, Palenque type, the vestibule with three entrances and two squarish piers, the sanctuary with tablet chamber, and two small lateral chambers. f. Three-chamber temple, Chichen-Itza type, the vestibule entered by wide portal with two serpent columns, the sanctuary enlarged by introducing two square columns to support the triple vault, and a long gallery with three doorways extending behind.

Ordinary doorways are single and give entrance to a single room, or, at most, to a suite rarely having more than two or three rooms. Back rooms are entered by doorways closely resembling the outer ones, getting all their light through them. The various forms of doorways are described farther on.

Apartments of all classes and all vaulted spaces are, with a few exceptions, limited in width by the capacity of the native arch to twelve feet or less. The length has no necessary limit, reaching in cases sixty feet or more. Such long rooms may be entered by a number of doorways and thus approximate the corridor type. It is reasonable to suppose that some of the buildings, now represented by piles of debris from which protrude multiple rows of columns, as at Chichen and Aké, were much more expansive in their apartment spaces which were rendered coalescent by the use of columns instead of partition walls. A notable feature of the plan in quadrangular groups of buildings is the gateway or wide, arched passage which opens through one of the outer buildings into the court.

The greater Maya buildings, though at times appearing complex in plan, are really exceedingly simple. The unit is the single cell or chamber seen standing alone in a, Fig. 3. The building shown in b consists of several units combined in one; variety is given to the plan in unsymmetrical structures by adding other units in less uniform ways and of varying size. The building shown in d differs from the preceding in having a sloped instead of a vertical entablature, the interior arrangement being much the same as in b. A sketch, intended as a restoration of the Caracol or Round Tower at Chichen, is presented in c.² This edifice contains two circular, concentric chambers identical in construction principle with the rectangular forms. In e we have the Palenque type of temple, and f is the square tower of the Palace at Palenque, the plan and construction of which are peculiar in several respects. . . .

The illustration given in Fig. 4 will serve to indicate sufficiently the construction and relations of the various features of an ordinary Maya building. The upper part of the substructure or pyramid is included and shows the stairway at the left, approaching the front doorway, and plain slope at the right. Details of the masonry of this mass are somewhat hypothetical, as I have not been able to determine whether or not it is the rule that a special foundation wall with vertical outer face was built from the ground up, but it is certain that this was often the case, and that the stairway and abutting masonry were afterwards added, as here shown, transforming the vertical-faced substructure into a sloping one. The floor is cemented as a rule, but occasionally is flagged, and the inner floor is in cases a step higher than the esplanade without. The superstructure here utilized, has two chambers, or two tiers of chambers, vaulted with the ordinary arch, and the walls are vertical without as is usual in Yucatan. The nature of the facing and hearting is shown in section in the back wall at the right, and the illy jointed and bonded masonry is correctly represented. The use of larger stones in the jambs of the doorways is indicated at the left. At a is the plain lower wall with doorway at b, and above is a sectional view of the wooden lintels, c. The front and back chambers are connected by a second doorway, d, identical with the outer one. The sloping sides of the corbellate or offset arch, dressed with the bevel, are seen at e and the capstone is at f. Special features seen within the rooms are the small, square wall perforation at the right, the poles or braces within the arch above, and two forms of cord fasteners—not large enough to be clearly made out—at the side of the inner door. One pair of these is made by drilling holes from adjoining faces of the stone until they meet, and the other by building a deep depression in the surface of the wall into which is fixed a vertical piece of round stone. The medial moldings, separating the two mural zones, typically developed, are shown at g. The upper zone with its sculpture-mosaic surface is seen at h, and the upper or frieze molding and coping course appears at the top, i. Continuous with the façade plane is the false or flying front, repeating the decorations of the façade proper more or less faithfully, and solid or perforated as the builder pleased or the nature of the ornament suggested. In some cases this feature is repeated in the same form over the medial wall of the building, but more frequently we have a more ambitious roof-comb, as indicated at k, and typicall y illustrated in the House of the Pigeons, Uxmal. It appears that the two forms arc not likely to occur on the same structure. Details are given in other connections. In the drawing the combs are disconnected from the building so that the ordinary roof may be seen in its level simplicity.

Doorways and Other Wall Openings. The wall perforations of Maya buildings may be arranged under six heads. They consist of (1) simple rectangular doorways with jambs, lintels and sometimes sills, (2) multiple or compound doorways in which the wide void is divided by one, two, three or more columns or pillars, (3) arched doorways which are of rare occurrence, (4) certain window-like openings or air holes of small size and varied shape, (5) the diversified openings in flying façades and roof-crests, some representing the interspaces of geometric ornaments, and others resembling doorways in their construction, but serving no function save that of embellishment, and (6) the so-called arched portals or gateways which are not wall perforations in the same sense as the others, but vaulted passageways opening entirely through the building from side to

Fig. 3. Examples of Maya buildings. a. Single-chamber building—a unit of construction. b. Multiple-chambered building—an assemblage of 12 or 14 units. c. Restoration of circular building, Chichen-Itza. d. Building with sloping entablature, Chichen-Itza. e. Temple with sloping entablature and roof-comb, Palenque. f. Square tower of four stories, Palenque—roof restored.

side, and not communicating with the apartments. The latter are described under the heading of the arch. . . .

Columns and Pillars. Developing pari passu with the doorways and arches we have a great variety of pillars and columns. The American column, in the nature of things, exhibits certain parallelisms with the columns of the eastern continent, but in all departures from the most elementary treatment and use it may be said to be characteristically American. Square columns, most numerous in Chichen, and pillars or piers, typically developed in Palenque, were usually simple in form though often embellished with elaborate sculptures or plastic designs in low relief, whilst the round column had advanced beyond the more elemental form with its shaft and simple cap, and was given, in whole or in part, varied and remarkable life forms, the feathered serpent being the favorite motive embodied. Among the most striking features of the great buildings of Chichen-Itza are the massive serpent columns, and on the Island of Cozumel, in a diminutive temple, the life-sized figure of a human being or man-like ape is sculptured in high relief against the face of the column.

Columns were usually assembled in pairs, where introduced into doorways to support the entablature, but appeared in groups and rows numbering scores or hundreds where extended façades or large roof areas were supported. Few specimens are monolithic, save in the east, as at Cozumel, where the size was reduced to a minimum and the available stone was perhaps more than usually massive. The proportions are considerably varied, but all are short and heavy. The diameter is to the height, approximately, as 1 to 3½. The square column is always built up of a number of heavy blocks.

The round column had become such a familiar feature of the building art that it was employed outside of its normal range of functions, appearing very frequently in the field of pure embellishment. In many of the Yucatec buildings it was used, on a reduced scale, to decorate the façades, where it was effectively introduced in moldings and friezes, forming long rows set in contact side by side. Generally the form was rounded only in front, while the back was flat or uneven and set in mortar. The form was varied in cases by formal moldings encircling the shaft, giving the effect, in a simple way, of our turned balusters.

The genesis of the stone column would seem to be easily made out, as prototypes are found in the wooden and stone roof supports employed in most primitive structures. The association with it of animal forms may perhaps be satisfactorily explained on the assumption that the figures or monsters embodied represent the divinities associated with the temple of which the column formed a conspicuous part. This association is in marked contrast with the more rational use of vegetal forms by the Egyptians and Greeks, though animal forms and figures of men and women

Fig. 4. Transverse section of an ordinary Yucatec building. The upper part of the pyramid is shown with the stairway at the left. a. Lower wall-zone pierced by a plain doorway. b. Doorway showing squared and dressed stones of jamb. c. Wooden lintels cut midway in length. d. Doorway connecting front with back chamber and showing position of cord holders. e. Inner face of arch dressed with the slope. f. Ceiling, or cap-stones of arch. g. Lower line of moldings, a survival of the archaic cornice. h. Decorated entablature zone. i. Upper moldings and coping. j, k. False front with decorations (occasionally added). l. Roof-crest with decorations (occasionally added).

were occasionally used in those countries. The association of animal forms with portals and columns was more common in the far East. It is notable that the round column was more generally and freely used in eastern Yucatan, as on Cozumel Island, at El Meco, Chichen, and Aké, than elsewhere. At Uxmal I saw no compound doorways and hence no portal columns, and in fact no columns of any kind, and at Palenque heavy flattish pillars, mere sections of the perforated wall, take their place. . . .

The Arch. The Maya arch presents a number of interesting forms and phases, all, probably, directly traceable to the more primitive forms of chamber spans or vaults in common use all over America. Among these earlier forms we have, first, the beam of wood or slab of stone connecting two lateral supports or walls and forming part of the roof or serving to support it; second, the single lean-to, in which the parts are placed against some fixed vertical surface or support; third, the double lean-to, where opposing parts are set against each other with or without a ridge pole; and, fourth, the circular lean-to, in which the parts form a cone with or without a central supporting pole. All arc equally elementary, and it will be impossible to determine just which varieties contributed most toward the development of the higher forms of vault in use among the Mayas. There are, however, but two principles of construction involved in all of these spans—the horizontal span and the lean-to. The latter is never used alone but occurs in combination with the former.

The prevailing form of Maya arch is based on the horizontal span, employing not single long slabs, but a series of short slabs so placed as to bridge the void by degrees. A course of stones is laid along the top of each of the opposing walls, projecting a little, a second course is laid in like manner, and others follow until by a series of offsets the sides have approached to within a foot or two, when a course of large well-squared slabs is laid across, completing the span.

In examples employing the lean-to principle, the construction is the same up to the point of connecting the closely approximate walls. Instead of laying a course of flat capstones across, two courses were employed, set on edge on the upper courses of the walls and inclined together at the top, continuing the pitch of the walls and forming the true cuneiform arch. The object of the off-setting is, of course, to reduce the span of the void, thus permitting the use of ceiling stones of small size instead of large and long stones which were hard to obtain and easily broken, or beams of wood which soon decayed. These arches really represent the emancipation of the Maya builder from the thraldom of the wooden beam. The prevailing variety was used in all forms of chambers and also in certain large vaulted passages, as in the Palace at Uxmal, and occasionally in smaller openings, as at Palenque, but the fiat span or lintel remained in nearly universal use for ordinary doorways. A unique appearing arch is found at Palenque, the sides being curved in such a way as to give a somewhat trefoil effect to the opening. The principle of construction is, however, the same as in the prevailing form of the cuneiform arch, the profile being curved instead of straight.

It is evident that considerable difficulty was experienced in carrying up the long slopes of the larger vaults, and the high angle adopted was one means of lessening the tendency to collapse. The projecting stones were largely held in place by the masonry of the body of the wall, which was carried up at the same time, but even this, especially in cases where the outer surface was also inclined, could not have prevented the frequent falling of the work when approaching the apex. In meeting this difficulty it was a common practice to use timbers—generally poles of medium or small size—which were placed across and built into the masonry as it rose, holding the walls apart. These beams are preserved in hundreds of cases and nearly every vault shows by its numerous beam sockets that this device was extensively relied upon. I believe the theory is advanced by some writer that a core of masonry was first built of the proper shape, and the vault constructed over it. I doubt if the numerous examples of masonry-filled apartments observed are satisfactory proof of this, but a careful examination of the surface finish in a room so filled might readily settle the question.

In Fig. 5 I present sketches of six examples of the Maya arch. These do not cover the entire ground, but others so far as I have seen are merely variations of the two prevailing types, shown in a and b, the first, terminating above in two rows of inclined slabs, forming the apex, and the second closed with a course of horizontal slabs. The former is seen in Chichen-Itza, but is rare elsewhere, and the latter was almost universally used in chamber vaults. The specimen shown in c differs from b only in having the corbellate or offset margins of the stones dressed with the slope, making a plane surface. That given in d is identical with the preceding, save that its inclined faces are slightly curved; it is the form sometimes used in the portal vaults which open through one or more of the buildings of a quadrangular group communicating with the court. It is seen also in chamber vaults, in bridges and aqueducts. The fifth example e is also a portal vault typically developed in the Governor’s Palace, Uxmal; indeed I cannot say that other illustrations are known. The slopes are long and it is probable that they were intended to be straight though now considerably warped, possibly by sagging. The sixth specimen, f, is the trefoil arch of the Palace in Palenque, which is the most ambitious attempt at arch elaboration in America, and shows, in connection with kindred wall perforations in the same building, an uphill struggle of the aesthetic in a field where construction was only blindly feeling its way.

Fig. 5. Examples of Maya arches. a. Section of cuneiform arch with acute apex, Chichen-Itza. b. Section of ordinary arch with fiat capstone. c. Section of ordinary arch with dressed surfaces. d. Section of ordinary arch with dressed surfaces and curved soffit slopes. e. Portal arch with long slopes, showing masonry of exterior facing. f. Section of trefoil, portal arch of Palenque.

The arch was rarely employed in ordinary doorways, exterior or interior, the few cases at Palenque being exceptional. The flat form of opening was preferred because the prolonged apex of the cuneiform arch led to troublesome complications with the interior vaults, as well as with the exterior medial moldings and the ornamented zone of the entablature.

It may be added that in numerous cases all four walls of the chamber are made to approach toward the apex of the vault, thus more thoroughly distributing the thrust of the superincumbent masonry.

The Maya builder did not often essay to construct his arch over a space more than twelve feet wide, though in the loftier buildings a much greater span was possible even with the ordinary pitch of the opposing walls. The average incline appears to be about 65 degrees, but occasional examples rise to 80 degrees, while others fall to 60 or even 55 degrees; the latter pitch would, however, give a weak construction, as the outward thrust would be increased to a dangerous degree. A building twenty-four feet high with roof three feet thick would accommodate a vault twenty-one feet high. If the vertical walls below are carried up to half this height, which is perhaps not far from the average relation of upper and lower spaces, an incline of 65 degrees in the opposing walls, allowing eighteen inches for the capstone span, will give a vault nine feet in width, or nearly ten feet, measured on the floor level, as there is usually an offset at the spring of the arch of from three to six inches on each side.

In the vaulted passageways through the Governor’s Palace at Uxmal the incline of the arch begins within a few inches of the ground, so that in the long rise of twenty feet or more, even with the high pitch of 70 degrees, the width spanned is not far from eighteen feet. The highest arch met with in my own investigations is in the outer annular chamber or gallery of the Round Tower at Chichen. The height is about twenty-four feet, while the width is only six feet; the pitch of the vault walls is therefore unusually great, and the apex correspondingly sharp.

¹

² It seems reasonably certain that the wails of both stories of this building were vertical as indicated, but the number and position of the openings of the upper turret, and the character of the platforms, or roofs, remain problematical.