APPENDIX, 9 449

stands perfectly clear all the way up, is exactly that which must best display its colour and quality. When there is much vacant space left behind a pillar, the shade against which it is relieved is comparatively indefinite, the eye passes by the shaft, and penetrates into the vacancy. But when a broad surface of wall is brought near the shaft, its own shadow is, in almost every effect of sunshine, so sharp and dark as to throw out its colours with the highest possible brilliancy; if there be no sunshine, the wall veil is subdued and varied by the most subtle gradations of delicate half shadow, hardly less advantageous to the shaft which it relieves. And, as far as regards pure effect in open air (all artifice of excessive darkness or mystery being excluded), I do not know anything whatsoever in the whole compass of the European architecture I have seen, which can for a moment be compared with the quaint shade and delicate colour, like that of Rembrandt and Paul Veronese united, which the sun brings out as his rays move from porch to porch along the St. Mark’s facade.

And, as if to prove that this was indeed the builder’s intention, and that he did not leave his shafts idle merely because he did not know how to set them to work safely, there are two pieces of masonry at the extremities of the facade, which are just as remarkable for their frank trust in the bearing power of the shafts as the rest are for their want to confidence in them. But, before we come to these, we must say a work or two respecting the second point named above, the superimposition¹ of the shafts.

It was assuredly not in the builder’s power, even had he been so inclined, to obtain shafts high enough to sustain the whole external gallery, as it is sustained in the nave, on one arcade. He had, as above noticed, [p. 95] a supply of shafts of every sort and size, from which he chose the largest for his nave shafts; the smallest were set aside for windows, jambs, balustrades, supports of pulpits, niches, and such other services, every conceivable size occurring in different portions of the building; and the middle-sized shafts were sorted into two classes, of which on the average one was about two-thirds the length of the other, and out of these the two stories of the facade and sides of the church are composed, the smaller shafts of course uppermost, and more numerous than the lower, according to the ordinary laws of superimposition adopted by all the Romanesque builders, and observed also in a kind of architecture quite as beautiful as any we are likely to invent, that of forest trees.

Nothing is more singular than the way in which this kind of superimposition (the only right one in the case of shafts) will shock a professed architect. He has been accustomed to see, in the Renaissance designs, shaft put on the top of shaft, three or four times over, and he thinks this quite right; but the moment he is shown a properly subdivided superimposition, in which the upper shafts diminish in size and multiply in number, so that the lower pillars would balance them safely even without cement, he exclaims that it is “against law,” as if he had never seen a tree in his life.

Not that the idea of the Byzantine superimposition was taken from trees, any more than that of Gothic arches. Both are simple compliances with laws of nature, and, therefore, approximations to the forms of nature.

There is, however, one very essential difference between tree structure and

¹ [This word has been misprinted “superior position” in every previous edition of the book.]

X. 2 F

[Version 0.04: March 2008]