oarse stone. The architect has then no choice but to arrange his
work in successive stories; either frankly completing the arch work and
cornice of each, and beginning a new story above it, which is the
honester and nobler way, or else tying the stories together by
supplementary shafts from floor to roof,--the general practice of the
Northern Gothic, and one which, unless most gracefully managed, gives
the look of a scaffolding, with cross-poles tied to its uprights, to the
whole clerestory wall. The best method is that which avoids all chance
of the upright shafts being supposed continuous, by increasing their
number and changing their places in the upper stories, so that the whole
work branches from the ground like a tree. This is the superimposition
of the Byzantine and the Pisan Romanesque; the most beautiful examples
of it being, I think, the Southern portico of St. Mark's, the church of
S. Giovanni at Pistoja, and the apse of the cathedral of Pisa. In
Renaissance work the two principles are equally distinct, though the
shafts are (I think) always one above the other. The reader may see one
of the best examples of the separately superimposed story in Whitehall
(and another far inferior in St. Paul's), and by turning himself round
at Whitehall may compare with it the system of connecting shafts in the
Treasury; though this is a singularly bad example, the window cornices
of the first floor being like shelves in a cupboard, and cutting the
mass of the building in two, in spite of the pillars.

Sec. XI. But this superimposition of lightness on weight is still more
distinctly the system of many buildings of the kind which I have above
called Architecture of Position, that is to say, architecture of which
the greater part is intended merely to keep something in a peculiar
position; as in light-houses, and many towers and belfries. The subject
of spire and tower architecture, however, is so interesting and
extensive, that I have thoughts of writing a detached essay upon it,
and, at all events, cannot enter upon it here: but this much is enough
for the reader to note for our present purpose, that, although many
towers do in reality stand on piers or shafts, as the central towers of
cathedrals, yet the expression of all of them, and the real structure of
the best and strongest, are the elevation of gradually diminishing
weight on massy or even solid foundation. Nevertheless, since the tower
is in its origin a building for s