ed, the true profile of the cornice, but beneath
which, more or less, the simpler profile is seen or suggested, which
terminates all the incisions of the chisel. This under profile will
often be found to be some condition of the type _a_ or _b_, Fig. LXIV.;
and the leaf profile to be another ogee with its fullest curve up
instead of down, lapping over the cornice edge above, so that the entire
profile might be considered as made up of two ogee curves laid, like
packed herrings, head to tail. Figures 8 and 9 of Plate XV. exemplify
this arrangement. Fig. 7 is a heavier contour, doubtless composed in the
same manner, but of which I had not marked the innermost profile, and
which I have given here only to complete the series which, from 7 to 12
inclusive, exemplifies the gradual restriction of the leaf outline, from
its boldest projection in the cornice to its most modest service in the
capital. This change, however, is not one which indicates difference of
age, but merely of office and position: the cornice 7 is from the tomb
of the Doge Andrea Dandolo (1350) in St. Mark's, 8 from a canopy over a
door of about the same period, 9 from the tomb of the Dogaressa Agnese
Venier (1411), 10 from that of Pietro Cornaro (1361),[88] and 11 from
that of Andrea Morosini (1347), all in the church of San Giov. and
Paola, all these being cornice profiles; and, finally, 12 from a capital
of the Ducal Palace, of fourteen century work.

Sec. XXIX. Now the reader will doubtless notice that in the three
examples, 10 to 12, the leaf has a different contour from that of 7, 8,
or 9. This difference is peculiarly significant. I have always desired
that the reader should theoretically consider the capital as a
concentration of the cornice; but in practice it often happens that the
cornice is, on the contrary, an unrolled capital; and one of the richest
early forms of the Byzantine cornice (not given in Plate XV., because its
separate character and importance require examination apart) is nothing
more than an unrolled continuation of the lower range of acanthus leaves
on the Corinthian capital. From this cornice others appear to have been
derived, like _e_ in Plate XVI., in which the acanthus outline has
become confused with that of the honeysuckle, and the rosette of the
centre of the Corinthian capital introduced between them; and thus their
forms approach more and more to those derived from the cornice itself.
Now if the leaf has the contour of 1