226 THE STONES OF VENICE CONSTRUCTION

§ 10. The circular window furnishes an exception to the common law, that the bars shall be vertical through the greater part of their length: for if they were so, they could neither have secure perpendicular footing, nor secure heading, their thrust being perpendicular to the curve of the voussoirs only in the centre of the window; therefore, a small circle, like the axle of a wheel, is put into the centre of the window, large enough to give footing to the necessary number of radiating bars: and the bars are arranged as spokes, being all of course properly capitalled and arch-headed. This is the best form of tracery for circular windows, naturally enough called wheel windows when so filled.

§ 11. Now, I wish the reader especially to observe that we have arrived at these forms of perfect Gothic tracery without the smallest reference to any practice of any school, or to any law of authority whatever. They are forms having essentially nothing whatever to do either with Goths or Greeks. They are eternal forms, based on laws of gravity and cohesion; and no better, nor any others so good, will ever be invented, so long as the present laws of gravity and cohesion subsist.

§ 12. It does not at all follow that this group of forms owes its origin to any such course of reasoning as that which has now led us to it. On the contrary, there is not the smallest doubt that tracery began, partly in the grouping of windows together (subsequently enclosed within a large arch*), and partly in the fantastic penetrations of a single slab of stone under the arch, as the circle in Plate 5 above.¹ The perfect form seems to have been accidentally struck in

* On the north side of the nave of the cathedral of Lyons² there is an early French window, presenting one of the usual groups of foliated arches and circles, left, as it were, loose, without any enclosing curve. The effect is very painful. This remarkable window is associated with others of the common form.

¹ [For a discussion of the origin of tracery, see Seven Lamps, ch. ii. § 21 (Vol. VIII. p. 87).]

² [Ruskin studied the cathedral of Lyons on his way back from Venice in April 1850; see below, p. 432.]

[Version 0.04: March 2008]