shafts. If such was the case, the vaulting
must have been found too heavy for the walls, and a wooden roof have
been therefore adopted in its stead. The easternmost bay of the
triforium, on each side, is apparently later Norman like the rest, but
is really the work of masons of the Decorated period. It had been
demolished in connection with the rebuilding of the nave, in progress at
that time but abandoned when only two bays were finished. It was then
found that the best way to make the junction of the styles good would be
to restore the old work as accurately as possible. This was well done,
but differences of material and in methods of working save us from being
deceived.

The two bays of Early Decorated work, just alluded to, complete the nave
eastwards. The transition from the round-arched to the pointed style is
made still more conspicuous by an increase in the height of the arcades,
which involved the discontinuance of the triforium; and the banded
shafts of dark Purbeck marble clustered round the later piers also
emphasize the change. The two piers at the junctions of the styles do
not pair, but we cannot regret the difference on the south side, as we
owe thereto the beautiful foliated capital here illustrated (p. 68).

The clerestory of the nave is divided from the stories below by an
enriched string course. It is of the same style throughout and dates
from the Perpendicular period. The predilection of the architects of
that time to substitute work of their own for that of their predecessors
in clerestories and great west windows of ecclesiastical buildings, has
been noticed by many writers. At Rochester they could not in either case
resist temptation. Their clerestory contains plain and uninteresting
three-light windows, which are, moreover, unsymmetrically placed with
regard to the arches beneath them. The roof is apparently of the same
date; it is flat and of wood, carried by corbels carved and painted to
represent angels bearing shields.

[Illustration: THE NAVE, LOOKING WEST (FROM A PHOTOGRAPH BY H. DAN).]

The two tower piers at the end of the nave, where the latter joins the
main transept, have their Purbeck marble shafts stopped at some height
from the ground. The most likely explanation of this is, that there used
to be here a solid stone screen [1], or rood loft, against which the
parish altar of St. Nicholas stood before 1423. On the west side of the
northern of the two rises a mass of masonry, so