The remarkable arrangement of the triforium is characteristic of all the four main divisions of the church. From the large pillars spring circular arches worked with heavy round mouldings. _Underneath_ these arches, not above them, is the triforium which is a blind arcade of two arches set in the tympanum beneath the main arch. The reason why there is this space under the main arch is because corbels in the form of half-capitals are set on the further side of the great pillars, a good way below the true capitals, to support the vaulting of the aisles. In this way, says Scott, "the pillars and arches have been divided, as it were, into two halves in their thickness, the half facing the aisle retaining its natural height and proportions, but that facing the central space being so raised as to embrace the triforium stage, the openings of which appear between the two ranges of arches; the clerestory ranging above." Of course, by this arrangement, the pillars avoid the low and stumpy proportions they would otherwise have, and the general effect of height in the nave (which is actually only 41 feet 6 inches) is considerably increased; for, were the triforium in the usual place above the main arches, the main pillars would not come any higher than the lower half-capitals. The arrangement is very unusual in England; though it is found in Italian Gothic, and even in Renaissance work in that country, as in St. Petronio, Bologna. It occurs in the transept of Romsey Abbey, in the choir at Jedburgh, in Dunstable Priory, and in Tewkesbury Abbey. That it existed in Saxon times is proved by a drawing in Cædmon's Paraphrase (c. 1000) in the Bodleian (_c.f._ "The Pre-Norman Date"). Dr. Ingram, who wrote in 1830, thought that this arrangement was made in order to raise the height of the building in the twelfth century, the triforium being the clerestory of the old Saxon church peeping out under the later work. And though his zeal was not according to knowledge (he thought the chapter-house doorway was Saxon), yet there is a possibility that this theory of his may have some truth in it. [Illustration: CHRISTCHURCH, OXFORD. PLAN.] Until lately, the church was thought to belong altogether to Prior Guimond's time. Sir Gilbert Scott fixed the date of the rather heavily carved capital over Bishop Berkeley's monument at 1170-80, owing to its close resemblance to certain capitals at Canterbury Cathedral of this period. The others seem to be of earlier date than this, and possibly of Ethelred's time. Strange as they are, however, they do not suggest a Saxon origin so strongly as do those of the choir. They are unique in design, and have neither the massiveness of Norman, nor the crisp severity of Early English work. The light, graceful, and rather fantastic foliage of the three eastern capitals on the south side--almost like iron-work--will be noticed. The third capital on the north side bears some resemblance to two of those in the choir.