Page:Ruskin - The Seven Lamps of Architecture.djvu/216

168 in arithmetical, i. e. two in the second arcade, and three in the third, to one in the first. Lest, however, this arrangement should be too formal, of the fourteen arches in the lowest series, that which contains the door is made larger than the rest, and is not in the middle, but the sixth from the west, leaving five on one side and eight on the other. Farther: this lowest arcade is terminated by broad flat pilasters, about half the width of its arches; but the arcade above is continuous; only the two extreme arches at the west end are made larger than all the rest, and instead of coming, as they should, into the space of the lower extreme arch, take in both it and its broad pilaster. Even this, however, was not out of order enough to satisfy the architect's eye; for there were still two arches above to each single one below: so, at the east end, where there were more arches, and the eye might be more easily cheated, what does he do but narrow the two extreme lower arches by half a braccio; while he at the same time slightly enlarged the upper ones, so as to get only seventeen upper to nine lower, instead of eighteen to nine. The eye is thus thoroughly confused, and the whole building thrown into one mass, by the curious variations in the adjustments of the superimposed shafts, not one of which is either exactly in, or positively out of, its place; and, to get this managed the more cunningly, there is from an inch to an inch and a half of gradual gain in the space of the four eastern arches, besides the confessed half braccio. Their measures, counting from the east, I found as follows:

The upper arcade is managed on the same principle: it looks at first as if there were three arches to each