Repairing the Nave Clerestory at Rochester Cathedral
|The south nave clerestory prior to repairs: distorted lead rainwater pipes and the
decaying hood mouldings and mullions are obvious. (Photo: Colin Tolhurst)
||Rusting glazing bars were causing the stone of the jambs and mullions to spall as
the corroding iron slowly expanded. (Photo: Universal Stone)
Rochester Cathedral has
its origins in the 7th century. In 604
Justus, the first bishop of Rochester,
was consecrated by Augustine of Canterbury.
However, it was not until 1083 that the present
building was begun by the Norman bishop
Gundulf, who founded the Benedictine
monastic community which survived until
the dissolution in 1541. What can be seen
today, as so often in such ancient buildings,
is an amalgam of different architectural styles
surviving from different building and rebuilding
programmes over the centuries. These are
overlaid by the work of subsequent repairs
and restoration. This is particularly true at
Rochester, which is an architectural potpourri
with examples of excellent work from every
era. The overall effect is one of ramshackle
charm rather than imposing grandeur. The
nave is a typical example of this: the arcades
and triforium are fine mid-12th century work,
partly rebuilt in the 13th century, and the
clerestories and roof were raised in the 15th
century and extensively restored in the 19th.
||Wrongly bedded mullion stones were splitting along
the natural grain of the stone. (Photo: Universal Stone)
The 2003 quinquennial survey of the
condition of the cathedral’s fabric identified
a number of defects in the nave clerestories
that required urgent attention. The hood
mouldings of the windows were failing, with
many splitting through in the plane of the wall.
This was probably exacerbated by the stones
being wrongly bedded. Sedimentary rocks, such as sandstone and limestone, have a natural
horizontal grain. Experience has shown that
these stones are most durable when set in the
building in the same orientation, with their
beds horizontal. There are exceptions to this
rule: in the construction of arches the bedding
planes should be at right angles to the curve
of the arch, and cornices or copings should be
laid with the bedding planes oriented vertically,
running back to front, so that they continue to
shed water even when eroded.
The window mullions had been fixed in
very long lengths with the bedding planes
upright and this had resulted in splitting. Many
mullions and jambs were also being spalled
by the corrosion of the external iron window
bars. The masonry and ironwork had all been
renewed in the 1870s during Sir George Gilbert
Scott’s extensive campaign of restoration.
Repair works were specified as one of
three packages for a successful application
to English Heritage for grant aid under their
cathedrals scheme. The other two packages
related to repair and improvement of access
arrangements at high level and rainwater
disposal. Competitive tenders were obtained and a contract was let for all three packages to
Universal Stone of Great Dunmow in Essex.
Work commenced in July 2006 and was
completed by Christmas 2006.
||Above: renewed mullions with three stones on their natural
bed in place of one long vertically bedded stone. Below: a new cast lead hopper replaces one of the broken
cast iron hoppers on the south transept: a repaired
lead hopper can be seen in the bottom picture, below left.
(Photos: Universal Stone)
One issue that had to be resolved at the
outset was access. Very substantial savings
were made by using roped access techniques
for the repair of the rainwater goods. The nave
clerestories, however, had to be scaffolded.
Externally, raising scaffolding from the roofs
of the aisles was straightforward. Internal
scaffolding was more challenging as it had to
be arranged to minimise the impact on the
constant use of the cathedral nave not only for
services but also for secular functions such as
The defects noted above were all visible
externally. However, the condition internally
was unknown since there is no access at that
level. Scaffolding was therefore required not
only for inspection and possible repairs but
also for health and safety reasons, should the
external works dislodge any masonry or glazing
inside. The south side was dealt with first and
scaffolding was then transferred to the north
so that as much as possible of the nave was left
unobstructed at any given time.
The clerestory windows had been entirely
renewed by Scott in Chilmark stone, a fact
confirmed by Bernard Worssam, the consultant
geologist. This grey-green stone from the Vale
of Wardour in Wiltshire is very similar to
Tisbury stone, from which Salisbury Cathedral
is largely built. It provides a reasonable match
for Reigate stone, which is why Scott favoured it for renewing that material. It cannot quite
make up its mind whether it is a sandstone or a
limestone, being defined as a slightly glauconitic
calcareous sandstone, or a sandy limestone. It
was decided to use the same material for repairs
since the defects had largely arisen from the
workmanship rather than the material itself.
The new stone has certainly blended in amongst
the old remarkably inconspicuously.
changes of appearance were accepted, however,
on the grounds of practical necessity. First,
more joints were introduced in the replacement
mullions since the new stones were laid on their
natural bed, limiting their available height to
about 450mm. Second, lead weatherings were
added to the tops of the hood mouldings to
improve water shedding and to prolong the
life of the stones that were retained and which
might also be wrongly bedded, even if they
had not yet developed splits. Some of those
which had incipient splits were given additional
support by drilling and inserting stainless steel
dowels bedded in epoxy resin. These changes
are not visible from ground level.
It was fortunate that virtually all of the work
could be tackled from the outside, leaving the glazing in situ. Mullions and jambs were cut
back to the line of the glass and jointed to the
internal half which had remained sound. Only
one window required complete renewal of the
mullions with the consequent resetting of the
panels of leaded lights.
It was also possible, with great care, to
remove all the iron bars for repair on the bench.
The corroded built-in ends were cut off and
stainless steel replacement sections welded on.
When the bars had been reset in the repaired
stonework the glazing then had to be reattached.
This also required access from inside. It was not
possible to solder on new copper wire ties with
the glazing in situ, so holes were very carefully
drilled through the hearts of the lead cames
on either side of the iron bars to allow copper
wire to be threaded through from inside to out
and then twisted up to secure the glazing.
||Hood moulding and stone repairs, before and after:
(above) the SPAB-style tile repairs were added in
the 1920s; (below) the lead weathering was added to
extend the life of new and retained hood mouldings.
New Caen stone indents can also be seen here.
(Photos: Universal Stone)
Repairs to the general walling of the
clerestories were limited to localised repointing
of failed joints and the renewal of a few very
badly eroded stones; Kentish Rag in the rubble
masonry parapets and Caen in the lower walls.
These had been rather beautifully repaired in
the 1920s using SPAB-inspired tile, brick and
mortar repairs, including repairs to the main
stringcourse above the clerestory windows.
These were left wherever they were sound
but the stringcourse was also given a lead
weathering to give it a further lease of life.
The mortars used for all these works were
based on a moderately hydraulic lime, rated
NHL 3.5, mixed with sharp sand and, in the case of some joints, a little stone dust for colour.
No Portland cement was used. Relatively small
quantities of mortar were required and the
ability to knock up batches of mortar quickly
using this dry, bagged lime was much more
convenient than using slaked lime in putty form.
At the same time, the early 19th-century
lead hopper heads and downpipes on the
clerestories were overhauled and splits repaired
with discrete welded lead patches. The pipe
collars had originally been nailed to wooden
blocks set into the masonry. The iron nails had
rusted and the blocks had decayed so the pipes
were re-attached with stainless steel fixings.
Once the scaffolding was struck and the
dust had (literally) settled, the overall result
was a very inconspicuous repair, retaining
all the patina and irregularity with which
people were familiar, but ensuring a robust,
weather-tight fabric for the coming decades.
The new leadwork to string course and hood
mouldings and the repaired rainwater pipes
should ensure the safe shedding of rainfall. The
tipped iron window bars will not now damage
the stonework into which they are fixed. The
inherent problems of the masonry have been
reduced by ensuring that the replacement stone
is laid on the correct bed to get the best possible
weathering properties from the Chilmark stone.
article is reproduced from Historic Churches, 2008
IAN STEWART is a director of Carden &
Godfrey Architects and has been Surveyor of
the Fabric of Rochester Cathedral since 2000.
He has been well acquainted with the cathedral
since the 1970s when he worked as an assistant
to the then surveyor, his late partner W Emil
Godfrey. He is AABC accredited and is a Fellow
of the Society of Antiquaries of London.
PRODUCTS AND SERVICES
services (stone masons)
Communications Limited 2010