line; and it is worked with what at first appears slovenliness, but is in fact sculptural sketching; exactly correspondent to a painter's light execution of a background: the lines appear and disappear again, are sometimes deep, sometimes shallow, sometimes quite broken off; and the recession of the cusp joins that of the external arch at n, in the most fearless defiance of all mathematical laws of curvilinear contact.
IX. There is something very delightful in this bold expression of the mind of the great master. I do not say that it is the 'perfect work' of patience, but I think that impatience is a glorious character in an advancing school: and I love the Romanesque and early Gothic especially, because they afford so much room for it; accidental carelessness of measurement or of execution being mingled undistinguishably with the purposed departures from symmetrical regularity, and the luxuriousness of perpetually variable fancy, which are eminently characteristic of both styles. How great, how frequent they are, and how brightly the severity of architectural law is relieved by their grace and suddenness, has not, I think, been enough observed; still less, the unequal measurements of even important features professing to be absolutely symmetrical. I am not so familiar with modern practice as to speak with confidence respecting its ordinary precision; but I imagine that the following measures of the western front of the cathedral of Pisa, would be looked upon by present architects as very blundering approximations. That front is divided into seven arched compartments, of which the second, fourth or central, and sixth contain doors; the seven are in a most subtle alternating proportion; the central being the largest, next to it the second and sixth, then the first and seventh, lastly the third and fifth. By this arrangement, of course, these three pairs should be equal; and they are so to the eye, but I found their actual measures to be the following, taken from pillar to
