_Baptistery and Loggia._--Of early buildings detached from the church we have the round building at the north-east, which we regard as having descended from the earlier church, and the south-west baptistery, with a loggia attached to its north side. The space between the church and the baptistery on plan looks like a covered way, leading from the church with a screen in the middle, but the part next the church is, and always must have been, open. The part next the baptistery is covered with a large semicylindrical vault, arched _transversely_ to the “screen,” and penetrated by a less cylinder in the direction of the length of the loggia. Rebates (on baptistery side) round the doorway which stands between the pair of columns show that there was a door, and strips down the sides of the pillars, which stand above the transom, show that pierced slabs or other closures filled the arched front of the vault. If we add breast-high closures in the lateral openings, as in the portico of St. John Studius, the whole becomes an inclosed loggia against the baptistery. Salzenberg states that there was a door in the north wall of baptistery, and Labarte places another in the western compartment of south aisle of church, but for the latter there does not appear to be a particle of evidence; and consequently the court and loggia cannot have formed a direct passage to the baptistery. 1. Salzenberg on his plan draws the transverse axis of the baptistery, and that of the western bay of the church; these do not agree by a foot or two, but the doorway of “screen” agrees with _neither_, nor is it a mean between them, but varies by excess. 2. In the section (Salzenberg, Plate xi.) it is seen that the present level of floor in this loggia is that of baptistery, and is below that of church; but the columns have no bases, therefore the loggia floor was beneath both church and baptistery. 3. A large arch is shown between the church and west pier of this loggia, from which it springs properly, while at the other end it is cut off incomplete by the wall of the church. These reasons together lead us to suggest that the loggia is possibly older than the church, and that it may be a part of an arcade retained when the present church was built. The style of the screen would readily allow of its being twenty or thirty years older than S. Sophia. The capitals are not found elsewhere in the church, while similar ones form the chief order at S. Sergius; and the door is inserted between the two columns, exactly as in the portico of S. John Studius. We do not however insist on its being earlier than the church so much as on the evidence pointing to its being part of a continuous arcade (see plan, Fig. 39). Doubtless it might be determined from a careful examination whether the loggia or the baptistery was built first. [Illustration: FIG. 39.--Restoration of Loggia by the Baptistery. Scale about eight feet to an inch.]