apertures, which are closely connected with their archivolts;
finally (E) the roof, including the real roof, and the minor roofs or
gables of pinnacles and arches. I think, under these divisions, all may
be arranged which is necessary to be generally stated; for tracery
decorations or aperture fillings are but smaller forms of application of
the arch, and the cusps are merely smaller spandrils, while buttresses
have, as far as I know, no specific ornament. The best are those which
have least; and the little they have resolves itself into pinnacles,
which are common to other portions of the building, or into small
shafts, arches, and niches, of still more general applicability. We
shall therefore have only five divisions to examine in succession, from
foundation to roof.

[Illustration: Fig. LI.]

Sec. III. But in the decoration of these several parts, certain minor
conditions of ornament occur which are of perfectly general application.
For instance, whether, in archivolts, jambs, or buttresses, or in square
piers, or at the extremity of the entire building, we necessarily have
the awkward (moral or architectural) feature, the _corner_. How to turn
a corner gracefully becomes, therefore, a perfectly general question; to
be examined without reference to any particular part of the edifice.

Sec. IV. Again, the furrows and ridges by which bars of parallel light and
shade are obtained, whether these are employed in arches, or jambs, or
bases, or cornices, must of necessity present one or more of six forms:
square projection, _a_ (Fig. LI.), or square recess, _b_, sharp
projection, _c_, or sharp recess, _d_, curved projection, _e_, or curved
recess, _f_. What odd curves the projection or recess may assume, or how
these different conditions may be mixed and run into one another, is
not our present business. We note only the six distinct kinds or types.

Now, when these ridges or furrows are on a small scale they often
themselves constitute all the ornament required for larger features, and
are left smooth cut; but on a very large scale they are apt to become
insipid, and they require a sub-ornament of their own, the consideration
of which is, of course, in great part, general, and irrespective of the
place held by the mouldings in the building itself: which consideration
I think we had better undertake first of all.

Sec. V. But before we come to particular examination of these minor forms,
let us see how far we can simpl