ters or pedestals richly ornamented. Three doors turned towards the rising sun admitted the entering crowd.

Ch. 38. Of the Apse, and the twelve columns with capitals.

Opposite to these doors was the apse, the head of the whole work, raised to the very roof of the Basilica. It was surrounded by twelve columns, the number of the Apostles; and they were ornamented with large silver capitals, which the Emperor dedicated to God as a beautiful gift.

Ch. 39. Of the Atrium, the Exedra, and the Portals.

Hence, going forward to the entrances which were before the temple, he interposed an open space, namely, between the Basilica and the portals: there were also recessed chambers (exedræ) on each side, the first or entrance-court, which had cloisters attached to it, and lastly the gates of the court. Beyond them, in the

· ῶν αἱ μὲν ἐπὶ προσώπου τοῦ οίκου, κίοσι παμμεγέθεσιν ἐπηρεί δοντο· αἱ δ' εἴσω τῶν ἔμπροσθεν ὑπὸ πεσσοῖς ἀνηγείροντο, πολὺν τὸν ἔξωθεν περιβεβλημένοις κόσμου.

5 ἡμισφαίριον.

6 Έκφρασις μεσαυλείου καὶ ἐξεἐρῶν καὶ προπύλων. Ενθεν δὲ προϊ ὄντων ἐπὶ τὰς πρὸ τοῦ νεω κειμένας εἰσόδους, αἴθριον διελάμβανεν. ἦσαν δ ἐνταυθοί παρ' ἑκάτερα, καὶ αὐλὴ πρώτης στοαί τ' ἐπὶ ταύτῃ, καὶ επί πᾶσιν αἱ αὐλειοι πύλαι. This Chapter is the most obscure of the whole. Taken literally, as it stands in the Greek, it would place on each side of the Basilica an atrium with its cloisters and vestibules, which is not likely; and is, besides, contradicted by the title of the Chapter, which gives us the atrium in the singular number. Valesius conjectures that the παρ' ἑκάτερα should be transposed to the cloisters; στοαί τ' ἐπὶ ταύτῃ παρ' ἑκάτερα. It appears to me not impossible that we

should read ἦσαν δ ̓ ἔξεδραι παρ' ἑκάτερα for ἦσαν δ ̓ ἐνταυθοῖ παρ' ἑκάτερα; for the exedræ are mentioned in the title, but not in the Chapter itself; and the words Εξεδραι and ἐνταυθοῖ resemble each other sufficiently, especially when written in capitals, to be mistaken for each other. In the basilica at Tyre there were similarly exedra and chambers on either side of the basilica, and connected with the front door,......ἐξέδρας καὶ οἴκους τοὺς παρ' ἑκάτερα μεγίστους ἐπισκευάζων εὐτέχνως, ἐπὶ ταυτὸν εἰς πλευρὰ τῷ βασιλείῳ συνεζευγμένους, καὶ ταῖς ἐπὶ τὸν μέσον οἶκον ἐισβολαῖς ἡνωμένους. (Eus. Eccl. Hist. lib. x. c. 4.) The exedra of the ancients appears to have been a recess or chamber, partly open, and provided with seats, often appended to a porticus; like the apses at the west end of Fig. 2. I have not attempted to delineate the exedræ of the entrance-court.

very middle of the wide market-place, stood the propylæa or vestibules of the whole work, which being decorated in the most imposing manner, afforded to those who were passing a promise of the wonders within.........This temple did the Emperor construct as a Martyrium of the Saving Resurrection, &c.

In the above description, after the Holy Sepulchre itself, we are introduced to a paved court, surrounded with porticoes, or cloisters on three of its sides, and having the Basilica on its fourth or eastern side. We are told that this side was opposite to the cave, by which, of course, is meant the entrance to the cave; for the history of the different states of the Holy Sepulchre in Section III. above, has shewn that it was an isolated edicula having its entrance to the East, and hence it must be inferred that the court here described surrounded the cave of the Sepulchre, and that the cloisters were opposite the sides and back of the monument, but that the Basilica occupied that side of the court which faced the entrance. I think it most probable that the cloisters were semicircular towards the West, following the present outline of the outer walls; for the excavation and levelling at this end seems to indicate such a form, and the outer wall of Constantine's cloister would be so far protected by the rock behind it, that it would probably escape obliteration. The rock shews at least that the court could not have extended farther West than the present building. In my restoration of the plan of the Basilica, (Fig. 2, Plate 1), I have delineated the cloistered court in this manner; and the positions of the North and South apses, which lie wholly to the west of the centre of the Rotunda, and opposite to the Sepulchre, seem to indicate that they were framed with reference to the semicircular form, and not to the circular form,

which the Rotunda of after ages assumed. Thus it is not impossible that these apses were also parts of Constantine's cloister, for such semicircular recesses (or exedra) are of frequent occurrence in Roman buildings'. But the restored plan which I have ventured to give must be considered as a mere diagram, shewing one out of many possible arrangements that may be conceived in coincidence with the description of Eusebius, which is far too loose, imperfect, and untechnical, to admit of certain interpretation into the accurate language of descriptive geometry. It may fairly be doubted, for example, whether the plural employed for the cloisters that surround the three sides of the court in question, is meant only for the three cloisters, one on each side, or is intended to convey the description of a double cloister on each side.

We now come to the Basilica; and to understand this it must be compared with those buildings of Constantine, the plan of which is better known to us. The whole of this Emperor's architectural works have been carefully collected and described by Ciampini 2. The plans of his churches are of two kinds; the larger ones appear to have been in the form of a parallelogram with side aisles, as the Lateran, Vatican, and St Paul at Rome. Others were of a circular or polygonal form, but were intended either for baptisteries or mausolea; as the Baptistery of Constantine, and the Mausolea of his daughter Constantia, and his mother Helena, all at Rome.

In the baths and temples at Rome, temples at Baiæ, Baalbec, Palmyra, palace of Diocletian at Spalatro, &c.

2 J. Ciampini, de sacris Edificiis a Constantino Magno constructis. (Romæ, 1693.)

At Constantinople he erected many which have disappeared; but it is remarkable that several of these are designated by the Byzantine historians as of a dromical form, a word singularly descriptive of a church with a rectangular body and an apse at the extremity; for the ancient dromos, or circus, was a parallelogram, square at one end, and circular at the other. St Sophia at Constantinople was, in its first state as Constantine built it, dromical, and so also were his churches of St Dynamis and St Agathonicus, in the same city. The great Church of the Apostles which he built for his burialplace was also dromical, and its sides were cruciform'. The church which he built at Antioch was octagonal.

There is nothing in the description of the Basilica, or House of Prayer, at the Holy Sepulchre, that would lead us to suppose its form to have been different from the parallelogram which I have just shewn to be the usual plan which the Emperor followed. It had double side-aisles, which we are told were partly above and partly below the ground. The survey of the original form of the ground, however, completely explains this phrase by shewing that to the present day the rock rises fifteen feet on the southern side of the site, and is exhibited on all sides, proving that the floor of the church must have been artificially sunk so much below the general surface, as to justify the expressions of our Author.

1 M. Couchaud, in his treatise on the Eglises Byzantines de la Grèce, has fallen into the singular mistake of asserting that Eusebius tells us all Constantine's churches were erected on an octagonal or circular plan, and covered with a dome, (p. 2.) It is true that Eusebius tells us the church of Antioch

was octagonal, (Lib. 111. c. 50,) but that is the only one so described by him. The church of Paulinus at Tyre was a basilica, of the ordinary dromical form, with its entrance at the east end, as appears from the description given by Eusebius in the tenth book of his Ecclesiastical History.

The words which he uses, in telling us that the colonnades in front had great columns, and those within had square pedestals, have led some to suppose that the first sort were placed in front of the building outside, and the others inside. But I believe his meaning to be, that the columns occupied the front ranks within, and that there were smaller pillars on pedestals behind, separating the two side-aisles from each other. This was exactly the case with the ancient Basilica of St Peter at Rome, and I have accordingly so represented our Basilica in the restored Plan. No allusion is made to a transept by Eusebius, who merely tells us that the doors were at the east end of the church, and opposite to them, the apse. In placing a transept in my Plan I have therefore taken a gratuitous liberty, but have nevertheless followed strictly the precedent afforded me in the plans of the Roman basilicas of the Emperor; and I have done so because the arrangement of the ground with reference to the form of Calvary appeared to indicate a transept, of which more below. To turn the doors of a church to the east, and the apse to the west, although contrary to the subsequent practice of Christendom, was the more usual in the time of Constantine; St Peter's itself being so turned, and most others of that age. The obscurest part of the whole description is in the last chapter, which contains a huddled list of the architectural members about the entrance-court, which, after all, was probably nothing more than the usual cloistered court which I have shewn in the Plan".

* Fortunately there is no ambiguity in the conclusion of the Chapter, which

tells us that the propylæum opened upon the market-place; a most valuable in