§ VIII. Evidently the first and roughest idea which would occur to any one who found a corner troublesome, would be to cut it off. This is a very summary and tyrannical proceeding, somewhat barbarous, yet advisable if nothing else can be done: an amputated corner is said to be chamfered. It can, however, evidently be cut off in three ways: 1. with a concave cut, _a_; 2. with a straight cut, _b_; 3. with a convex cut, _c_, Fig. LII.
The first two methods, the most violent and summary, have the apparent disadvantage that we get by them,--two corners instead of one; much milder corners, however, and with a different light and shade between them; so that both methods are often very expedient. You may see the straight chamfer (_b_) on most lamp posts, and pillars at railway stations, it being the easiest to cut: the concave chamfer requires more care, and occurs generally in well-finished but simple architecture--very beautifully in the small arches of the Broletto of Como, Plate V.; and the straight chamfer in architecture of every kind, very constantly in Norman cornices and arches, as in Fig. 2, Plate IV., at Sens.
§ IX. The third, or convex chamfer, as it is the gentlest mode of treatment, so (as in medicine and morals) it is very generally the best. For while the two other methods produce two corners instead of one, this gentle chamfer does verily get rid of the corner altogether, and substitutes a soft curve in its place.
But it has, in the form above given, this grave disadvantage, that it looks as if the corner had been rubbed or worn off, blunted by time and weather, and in want of sharpening again. A great deal often depends, and in such a case as this, everything depends, on the _Voluntariness_ of the ornament. The work of time is beautiful on surfaces, but not on edges intended to be sharp. Even if we needed them blunt, we should not like them blunt on compulsion; so, to show that the bluntness is our own ordaining, we will put a slight incised line to mark off the rounding, and show that it goes no farther than we choose. We shall thus have the section _a_, Fig. LIII.; and this mode of turning an angle is one of the very best ever invented. By enlarging and deepening the incision, we get in succession the forms _b_, _c_, _d_; and by describing a small equal arc on each of the sloping lines of these figures, we get _e_, _f_, _g_, _h_.