In addition to the arched West Gate of the city there are still standing in Bosra two monumental arches, like Roman triumphal arches, which are of more than usual interest. One of these is at the east end of the principal colonnaded street, and has been designated as the East Arch (111. 214), the other faces north upon the south side of the same street near, the centre of the city, and may be called the Central Arch (111. 215). Of the two, the East Arch is the smaller, and appears to be the elder. Both are among the better preserved of the monuments of Bosra.
East Arch. This arch is one having a single, broad, high opening, in form of a tunnel vault, through its minor axis, and two narrow, low, tunnel vaults piercing the masonry on either side of the main opening, through the major axis of the structure, at  right   angles  to  the   main arch.    The plan (111.  216) gives the arrangement of the

 which have been built about the arch and within it. The accumulation of soil is so deep in this quarter of the town that I have assumed that the portions of the arch still visible, as shown by the shaded masonry in 111. 217, are elevated upon a sort of podium, at least 2 m. high, as I have indicated in the resto­ration. Indeed a piece of the cap moulding of such a podium is visible on the east face of the arch on both sides of the opening and under the vault of the opening. Above this level, a part of the lower storey of the east face is visible in a courtyard, while the rest of this face can be seen in houses on either side. Across the west face a crude wall has been built, not 60 cm. from the building, all but concealing the lower storey.

This   crude   wall   was   built to   prevent the filling up of the lower part of the arch   and   the   minor  vaults   on either side of it, since they are inhabited; for the soil outside of this wall has risen to the level of the upper storey of the arch, as may be seen by referring to 111.  214.    The upper storey has been completely destroyed on the east face and at both ends, the west face preserves a considerable part of this storey, enough in fact to make the restoration quite certain,  as may be seen by comparing the restoration (111. 217) with the photograph (111. 

  214).    Assuming that the proportions of the   podium   are   approximately   correct,   we   find the first storey, on both faces of the arch and on either side of the opening, ornamented with pilasters at the angles, quarter columns adjoining the pilasters, and half columns flanking niches in the middle of each side.    The architrave breaks out en ressaut over each pilaster and half column, but the other members of the entablature above the architrave are omitted.    In the upper storey another  composition   is   repeated four times, twice on either face of the arch.    This is made up of four pilasters which ascend from the ressauts of the lower order, and three niches, one large semi-circular niche in the middle space and a small rectangular niche in   either   of the   side   spaces.     The pilasters have no bases, their caps are formed by a simple moulding which is carried across the face of the arch, breaking out over each, and curving upward in an arch above the middle niche.    The main arch is the central feature of this storey.     Its archivolt has a good set of architrave mouldings which spring from the ressauts above the lower pilasters.    It will be noted that two of the pilasters of the  upper   storey   terminate   below rather clumsily upon the extrados of these arch mouldings.    There   is   no   attempt   to  produce   the   effect  of an entablature above the order of the upper storey; the wall rises in two plain courses to a height a little above the crown of the main arch where a moulded string course is carried across the entire face   of the   arch. 

   There   are remains which show that the building was at least one tall   course   higher   than   the   string  moulding; but it is impossible to know how much higher  the structure was,  or how it was completed at the top.     I have added a deep overhanging   cymatium   of the   ordinary   Hauranian   type to finish the restoration; but there  may have been a complete Attic storey in the style of Roman arches.     Section A-B  gives   the   treatment   of the   two   interior faces of the arch, with the opening of the  low   vaulted   side passages.     Nothing in  this drawing is restored save  the podium and  the  profile  of the   crowning mouldings of the arch.    The ends of the edifice are almost wholly conjectural,  the lower storey being almost completely hidden by modern buildings at both ends and the upper part being in ruins.

The more important details of this building are given on a larger scale in Plate X. The order is most unusual, though, as we have seen, it was not unique in Bosra, and it is not difficult to detect the resemblance it bears to examples known in Petra, Hegra and Si'. For this reason I do not hesitate to call it a Nabataean order. The bases of the half and quarter columns may not be given with complete accuracy; for I found them badly injured; but the capitals and the architrave and other mouldings are in good preservation. The capitals of the half columns are small copies of the huge specimen described a few pages above (111. 211); they have a set of circular mouldings beneath a high hollow-sided abacus with far projecting corners and with bosses in the middle of the face. The caps of the grouped pilasters and quarter columns at the exterior and interior angles on the east face are compound designs in which the circular mouldings  of. the   half  columns  are replaced by a row of stiff,  erect, acanthus- leaves.

The abacus of each pilaster cap is of the same "horned" variety as that of the half columns, and a single "horn" of the same type of abacus projects above the quarter column.    The architrave is low, with two bands and a very salient cymatium.
It is a pity that no inscription has as yet been found that might give a definite date to this monument. We may not even look to Petra for dated monuments that would assist in dating this arch. But the dated and dateable buildings erected in the Hauran under the Antonine emperors and under the emperors of the third century present none of the details which set this monument apart, and point to an earlier date, perhaps in the reigm of Trajan, probably earlier. The peculiar abacus of the capitals in this edifice is a common feature in the rock-hewn facades of Petra x which are believed to be earlier than the year 106 a. d. The combined pilaster and quarter columns under a composite cap of this strange order appear in rock-hewn tomb fronts discovered at Hegra (Medain-Saleh) and published by the Dominican Fathers Jaussen and Savignac". These particular details appear, without the circular mouldings below the abacus, beside the doorways of three tombs (E 18) (A 5) and (B 1), all dated by Nabataean inscriptions of the year 31 a. d. 3 and in the same part of another tomb (F 4) dated in the 24th year of Malichus IL, i. e. 63-64 a. d. Engaged columns, and pilasters with quarter columns attached, both having circular mouldings below the abacus, all very like those in the East Arch at Bosra, are found in a rock-hewn facade in Petra (No. 633) which bears a Nabataean inscription that is to be dated early in the second century of this era. It is therefore plain that these details were well known in the country of the Nabataeans in the first century a. d. and early in the second. The capital with a cluster of leaves below the "horned" abacus has not been found in Petra or Hegra. This fact however need not interfere with our assigning of the East Arch of Bosra to a date near the end of the first, or early in the second, century; for Greek influence was much stronger here than in the regions farther south. Central Arch. This monument has been published with great care and accuracy by Professor Briinnow5, and I am presenting it in this work, only in order to make the study of the buildings of Bosra as complete as possible within the compass of a single publication, and to offer a restoration based upon a study of certain somewhat similar monuments in Syria. The Central Arch is the most conspicuous of the Ancient buildings in Bosra after the West Gate. Illustrations of it have appeared in a large number of books, but the measured drawings of Professor Briinnow were the first of their kind to appear. The arch, which has triple openings through its minor axis, was also pierced with a vaulted passage through its longer axis, which divided the struc­ture longitudinally into two equal parts, or halves. The northern half stands in almost perfect condition (111. 215). The southern half has completely disappeared. One may observe by referring to the ground plan (111. 218) that the intersections of the longi­tudinal passage with the two minor openings of the arch involved the use of two cross vaults which so separated the two faces of the structure that one face could stand without the other. The plan and the photograph also show engaged columns at the east and west angles of the north face of the arch, which stand in line with the columns on the south side of the main colonnaded street. The blackened portions of the ground plan, with the one truncated column at the west, and such parts of the   superstructure. as are shown in the photograph

   diameters   in height, and are provided with suitable bases,   which   give  a   fair  proportion   to  the   order.    On   either hand the ends of the street colonnades have been restored from the remains published in 111. 210, and statues of life size have been placed upon the consoles, or brackets, which were made to receive them.    Above the Corinthian pilaster only one member is shaded, this is the architrave which   is   in   place.     In   its   present  condition the structure above this is wholly out of keeping  with   the  lower storey,  as the photograph (111.  215) will show, — a perfectly plain  wall   containing  the  main  arch,   with   capless and baseless pilasters at the ends, and a salient cymatium across the top.    The masonry of this storey is not so smooth or so well laid as that of the storey below it, and it must have puzzled many beholders to  reconcile  this   crude   work  with the very excellent work on the pilasters and lower arches.    It  will  be   observed that the mouldings of the architrave are returned at the arch,  and  were intended to be carried over the semicircle; but the face of the present arch above its springing is without mouldings.    It may be urged that the Arch, having been begun in a costly manner, was completed by change of plans in this simple and inexpensive way even in the Roman period; but I believe that the Arch was originally completed  in  quite   a   different manner, and was rebuilt in its present form at a com­paratively recent date.    This arch is called by the natives " il-Kandil"; it is known from an Arabic inscription  that there was once a mosque in Bosra of that name, and I am convinced  that  this arch once formed the front of that mosque, that the ruins behind it are the ruins of the mosque and that the present upper storey of the Arch belongs to  the   period   of the   building of the Djami' il-Kandil.    But this assumption does not restore the arch to its- original design.    Suggestions for the restoration of the Arch to the original plan are to be found, first, in the building itself, and secondly, in buildings of the  same general type and period in the neighbourhood.    The returned moulding's of the  architrave   demand   at  least   one completely arched member of an entablature, and  suggest  that  the  two   others   were  employed. 

 The   other  two  Arches  of Bosra have only the lower member, it is true; but neither of them preserves a true Classical order. It would be impossible to restore the Central Arch after the manner of the others, for the reason that the upper storey would be too low in proportion to its width. The Corinthian pilasters and other ornamental details of the Arch compare favourably with similar features in other monuments of the Hauran and of the cities of the Deca­polis that belong to the period of the Antonine emperors, and in all such monuments in which the arcuated architrave occurs, it is accompanied by the arcuated frieze and cornice. It is not important. to consider the third-century inscription that appears below the statue console at the west end of the Arch at Bosra; for this is simply an honor­ary inscription that might have been engraved at any time after the completion of the building. In proceeding with our restoration we can do no better, I believe, than to take the Propylaea of Djerashas a model; for the pilasters of that building are almost precisely similar to those of our Arch. These Propylaea belong unquestionably to the Antonine period, as is attested by an inscription. The Djerash Propylaea, like this Arch at Bosra, were set upon the line of one of the street colonnades; they consisted also of two main parts longitudinally divided, a portico of four columns with a broad middle intercolumination, and a wall with one large and two small openings in it, faced with pilasters. The middle intercolumniation of the portico was surmounted by a complete entablature in arch form, and a raking cornice covered the whole façade. The colon­nade of the street joined the portico on either hand, and the covered walk passed behind the columns and in front of the wall. The plan of the Arch at Bosra is very-similar ; but the portico of columns is replaced by pilasters embracing arches, and the side walk passes under cross vaults behind the lower arches. Since we have the evi­dence for the arcuated architrave in the Bosra monument, which seems to correspond with the monument at Djerash in size and style, it will not be assuming too much to complete the entablature, and cover the whole with a raking cornice •, that is, to take the entablature and gable from the columnar portico of Djerash and place it upon the pilasters of the Bosra Arch, and this is what I have done in the restoration of the-north face (111. 218). As to the restoration of the two ends I am not at all confident; but I have placed narrow gables directly above the roofs of the side walks of the colonnade which abut upon the ends of the Arch. The roof of the Arch might quite as well follow the simple pitch of the raking cornice. It is quite evident that the entablature was stopped and returned upon the end walls immediately after rounding the angles; for the architrave mouldings are terminated in this manner.
The restoration of the Section is conclusive, having been taken from Brunnow's drawings and photographs which were made while the interior arches and vault of the east half of the monument were still standing. The eastern half of the south face also shows considerable portions of the Arch which may now be studied only from photo­graphs. Since the entablature of the north face was stopped at the ends of that face, I have restored the south face of the principal arch and the walls on either side of it without ornament of any kind. Indeed the window opening shown in Brunnow's photograph, Fig. 900, occupies a space upon the level of the entablature of the front of the Arch. Here again the end gable might be omitted, allowing the high roof to follow the straight line of the raking cornice direction of the Theatre, are not to be questioned ; the only features open to debate are the solid back walls of the colonnades which are shown extending out at right angles from the corners of the Arch, and the manner of roofing the side walks. The former seem necessary to the completion of the plan of these colonnades; the latter might equally well have been a simple shed roof of single pitch.