of the
caterpillar that has bound them: and thus the last forms of the Venetian
leafage set forth the fate of Venetian pride; and, in their utmost
luxuriance and abandonment, perish as if eaten of worms.

And now, by glancing back to Plate X. Vol. II, the reader will see in a
moment the kind of evidence which is found of the date of capitals in
their profiles merely. Observe: we have seen that the treatment of the
leaves in the Madonnetta House seemed "indicative of a tendency to
transition." Note their profile, 1_a_, and its close correspondence with
1 _h_, which is actually of a transitional capital from the upper arcade
of second order windows in the Apostoli Palace; yet both shown to be
very close to the Byzantine period, if not belonging to it, by their
fellowship with the profile _i_, from the Fondaco de' Turchi. Then note
the close correspondence of all the other profiles in that line, which
belong to the concave capitals or plinths of the Byzantine palaces, and
note their composition, the abacus being, in idea, merely an echo or
reduplication of the capital itself; as seen in perfect simplicity in
the profile _f_, which is a roll under a _tall_ concave curve forming
the bell of the capital, with a roll and _short_ concave curve for its
abacus. This peculiar abacus is an unfailing test of early date; and our
finding this simple profile used for the Ducal Palace (_f_), is strongly
confirmatory of all our former conclusions.

Then the next row, 2, are the Byzantine and early Gothic semi-convex
curves, in their pure forms, having no roll below; but often with a roll
added, as at _f_, and in certain early Gothic conditions curiously fused
into it, with a cavetto between, as _b_, _c_, _d_. But the more archaic
form is as at _f_ and _k_; and as these two profiles are from the Ducal
Palace and Piazzetta shafts, they join again with the rest of the
evidence of their early date. The profiles _i_ and _k_ are both most
beautiful; _i_ is that of the great capitals of the Ducal Palace, and
the small profiles between it and _k_ are the varieties used on the
fillet at its base. The profile _i_ should have had leaves springing
from it, as 1 _h_ has, only more boldly, but there was no room for them.

The reader cannot fail to discern at a glance the fellowship of the
whole series of profiles, 2 _a_ to _k_, nor can he but with equal ease
observe a marked difference in 4 _d_ and 4 _e_ from any others in the
plate; the bulging ou