{\rtf1\ansi\ansicpg1252\deff0\deflang2057{\fonttbl{\f0\fswiss\fprq2\fcharset0 Microsoft Sans Serif;}}
{\colortbl ;\red0\green0\blue128;}
\viewkind4\uc1\pard\cf1\f0\fs20 Wren\rquote s revisions to the design in 1685-86 created a podium for a vast, richly modelled dome, inspired by those of Michelangelo\rquote s St Peter\rquote s in Rome (1547-90) and Jules Hardouin-Mansart\rquote s Invalides church in Paris (1677-91). Engravings of the Invalides church issued in 1687 reveal a timber-frame outer dome and lantern; Wren's earliest engraved section, produced soon after, shows a dome and lantern entirely in masonry, and lower drum that spans nearly the full 160-ft width of the crossing square (see WRE/5/1/2, and Higgott 2009, pp.155-61). By comparison the dome completed in 1708 is far lighter in construction and rises from a drum only 140 ft wide externally. The outer dome is lead-clad timber and the inner wall of the drum is sloped to carry a hidden brick cone just 1 ft 6 inches thick, its base reinforced with a wrought-iron chain. The canted inner drum and cone combine to support the tall stone lantern which appears to rest on the hemisphere of the outer dome. They direct its weight of the lantern towards the fronts of the crossing arches and help contain the overall load of the dome within the piers masses of the crypt. \par
\par
How did Wren arrive at his remarkable final scheme? Construction began at the base of drum of the dome in January 1696. By this time Wren would have decided on the widths and positions of the inner and outer walls of the drum and the structures that rise from them. A payment in December 1695 to Edward Strong\rquote s masons to \lquote Alter and Add to the Modell of the Legg of the Dome\rquote suggests final revisions to the lower part of the dome (Wren Society 15, p.5). These revisions would have included the sloping of the inner wall, a crucial modification which allowed Wren to reduce its thickness. The masonry model altered in late 1695 was probably the \lquote large Modell of 1/8 of the Great Dome\rquote built by Strong\rquote s men in June 1694 (Wren Society 14, p.134). \par
\par
Three years earlier Strong\rquote s men made models of a \lquote quarter\rquote of the dome (January to February 1691) and a \lquote part of a Modell in small stones for part of the Dome\rquote (May 1691) (Wren Society 14, pp.80, 86). To this earlier period, c.1690-91, can be assigned a group of small-scale studies which explore alternatives to the 16-bay scheme of the \lquote Revised design\rquote : WRE/5/1. Amongst the earliest of these is WRE/5/1/3, a composite drawing of c.1690, mostly in Hawksmoor\rquote s hand, which reduces the outer dome to a single shell and changes the design again - on the right - by setting a much larger lantern over a hemispherical dome. This was the visual solution that Wren aspired towards but had yet to discover the structural means to achieve.\par
\par
In a pen-and-ink study, now at the British Museum, Wren transformed the section of his dome to lessen its weight (Higgott 2009, pp.161-64, fig.86). He was concerned to relieve pressure on the substructure in the crypt, where, from about 1689, the internal rubble cores of the piers beneath the crossing were beginning to settle and cause splitting in the external stone cladding. The two half-sections in the study employ a formula devised by Robert Hooke in about 1671 for calculating the curve of a parabolic dome and reducing its thickness. Hooke had explored this curve the three-dimensional equivalent of the \lquote hanging chain\rquote , or catenary arch: the shape of a weighted chain which, when inverted, produces the ideal profile for a self-supporting arch. He thought that such a curve derived from the equation y = x3. On the right side Wren divided the half-section into four parts, each 20 ft wide. These form the \lquote x-axis\rquote of a graph. The points of intersection with the cubes of these numbers (1, 8, 27, 64) are taken from a scale in units of 8 on the vertical \lquote y-axis\rquote and create a parabolic curve which dictates the overall section of the dome. Wren drew this profile more carefully in the left half-section of a 24-bay dome. However, the parabola had to be straight-sided to support a large lantern. Working with Hawksmoor and an unknown draughtsman Wren modified the profiles in his study-design and explored variants of a triple-shell masonry dome with straight-sided internal cones: WRE/5/1/6-11. \par
\par
In the next phase, between about 1691 and 1694, Wren introduced iron chains into the sections of his designs to contain the outward thrusts of the cone and inner dome: WRE/5/2. This innovation enabled Wren to reduce the depth of the peristyle and heighten his internal and outer domes. He increased the bays in the peristyle to 32, as in the Great Model, but infilled every fourth one as a buttress. A sketch for such a dome, WRE/5/2/3, has two slots for the chains. This study led to a series of larger scale designs, including WRE/5/2/6, an eighth-part plan of the dome, shaded in colour to distinguish stone types at two levels. This drawing is probably preparatory for the small-scale masonry model of June 1694.\par
\par
The final phase from c.1694 onwards saw Wren establishing the section of the dome up to the top of the peristyle by the start of construction in 1696, but then modifying it soon after Queen Anne\rquote s accession in March 1702; see WRE/5/3. In the summer of that year Anne\rquote s first Parliament agreed a large increase in funding from coal tax to complete and decorate the cathedral, effective from 1708 (Campbell 2007, pp.67-69; Lang 1956, p.220). This overturned a severe cut in revenue imposed under William III in 1697, when Wren\rquote s salary was halved pending the completion of the cathedral. \par
\par
A full-height section of the crossing and dome, WRE/5/3/9, drawn mostly by William Dickinson, and revised shortly afterwards by Wren, reflects the cathedral\rquote s improved financial circumstances around 1702 and the architect\rquote s renewed ambition to crown his dome with an imposing lantern. Dickinson drew three alternatives for the section above the internal peristyle in brown ink, two on the right and one on the left, the latter beneath the pink-coloured revisions. The brown-ink alternatives show that Wren had not decided on the structural form of the upper part of the dome in 1701-02, when construction was 10 ft above the floor of the peristyle. These variants are of a dome that supersedes a version engraved at this time. The left-hand variant is the most advanced. It eliminates the upper tier of windows in the peristyle by lowering the springing of the inner dome. This version was then revised in freehand pink line and shading over thick pencil. Wren increased the thickness of the cone and fused its base with that of the inner dome to provide a more solid support for a more massive lantern, which rises through the timber-framed shell of a hemispherical outer dome. \par
\par
The lantern was entrusted entirely to Edward Strong and his team of masons, one of whom may have drawn a marked-up half-section, probably for a model completed in January 1707; see WRE/5/3/10. Dickinson\rquote s fully annotated construction drawing for the lantern\rquote s capping, ball and cross, endorsed with instructions for gilding in November 1708, aptly concludes the sequence of drawings for the dome; see WRE/5/3/11.\par
}
Conditions governing access
Access to the Wren office drawings held at London Metropolitan Archives is available only with advance notice and at the discretion of the Heritage Services Director, London Metropolitan Archives, 40 Northampton Road, London, EC1R 0HB.
