Sir Christopher Wren’s Ambitious Tower and Spire
Once we reach the eighteenth century, historical records relating to the development of the lantern, both proposed and actual, become more plentiful, although far from complete. Some of the evidence sheds valuable light on earlier vicissitudes too, and this has already been embraced in the previous chapter.
The First Report on the Fabric, 1713
Sir Christopher Wren was appointed Surveyor of the Fabric of Westminster Abbey in 1698. His principal assistant (Under Surveyor) was William Dickinson, who remained briefly in post after Wren’s death in 1723. Wren was continuously engaged, inter alia, on a comprehensive re-roofing campaign from 1699 to 1718, and in 1713 he reported to Francis Atterbury, the newly installed dean, on the current state of the fabric.50 He described the works undertaken to date, set out ‘what yet remains to finish the necessary repairs’, and presented his ‘thoughts and designs, in order to [achieve] a proper compleating [sic] of what is left imperfect, hoping we may obtain for this the continuation of the Parliamentary assistance’.51 Given Wren’s long association with Westminster, it is unfortunate that the Abbey has only sixteen drawings emanating from his office.52 Some of these are pertinent to the crossing.
From an early stage in his surveyorship, Wren clearly set his sights upon building an impressive crossing tower and spire, and in 1710 he had a model made in oak and pear wood demonstrating his proposal.53 [34] The model, which cost £30 and stands 2.36 m high, is still extant.54 It was made to a scale of 4 feet to one inch (i.e. 1:48) and thus represents a proposed structure some 372 ft (114 m) high. The tower was square with attached octagonal corner-turrets, and there were two main stages, the upper being a belfry. It had crenellated parapets, and pinnacles at the corners, and from the centre rose a lofty twelve-sided needle-spire, pierced at two levels by lancet openings in every facet.
Referring to the lantern, Wren noted that the construction of a Gothic tower had been started, ‘the beginnings of which appear on the corners of the cross’, but that it was abandoned ‘before it rose so high as the ridge of the roof ’. He clearly saw the remnants of Henry III’s angle-turrets and incorporated them in his design. Wren added that ‘the vault of the quire under it is only lath and plaister, now rotten’. This description refers to the ceiling over the crossing, which was finished as a ‘dome’ and, as already noted, is shown both on Wren’s plan and section of 1715 [31], and on Dickinson’s section of 1724. [32]
Structural Strengthening to Support a Tower
Wren advocated the building of an impressive steeple, first pointing to the problem that beset the four crossing piers (or pillars, as they were termed): ‘it is manifest to the eye that the four innermost pillars of the cross are bended inward considerably, and seem to tend to ruin, and the arches of the second order above are cracked also’. He was referring here to the triforium arcades. He continued, ‘lest it should be doubted whether the four pillars below be able to bear a steeple, because they seem a little swayed inward, I have considered how they may be unquestionably secured, so as to support the greatest weight that need be laid upon them; and this after a manner that will add to their shape and beauty’. The bowing of the crossing piers was a result of the lateral thrust exerted on each of them, from two directions, namely by both the main arcades and the triforia. He illustrated the engineering principles with a diagram.55 Wren attributed the failing to the absence of the intended central tower, the dead-weight of which, he argued, would have anchored the piers and prevented their bowing. This is the same principle as applies to the design of flying buttresses, where masonry is piled, as a counterpoise, on top of a buttress that has to take the lateral thrust from a flier.
34 Wren, 1710. Timber model of his proposed crossing tower and spire. WA Lib. Coll.
35 Iron tie-bars linking the capitals of the arcade piers to one another, and to the aisle walls. Note the right-hand end of the bar in the foreground retains its original fixing (sandwiched between the abacus and capital), whereas the left-hand end has been refixed to a stirrup attached to the arch moulding above the abacus. View north-west from the south transept. Author
Wren also noted with interest the presence of wrought iron tie-bars at capital level in the main arcades, the purpose of which was to connect all the piers to one another and also to the outer walls of the aisles. [35, 72] He concluded that the function of the bars adjacent to the crossing was to restrain the arches from spreading and applying a lateral thrust to the crossing piers, noting that the medieval master mason, ‘tied these arches every way with iron’, so that ‘this might serve the turn till he built the tower to make all secure, which is not done to this day’.
Wren then made the interesting observation that ‘these irons which were hooked on from pillar to pillar have been stolen away; and this is the reason of the four [crossing] pillars being bent inward, and the walls above cracked’. Today, all the tie-bars linked to the crossing piers are present, which raises the question as to whether some had been removed and were replaced either by Wren or by one of his successors. There is no evidence to suggest that Wren did any work in the crossing, and this fell to Hawksmoor (p. 53).56 Moreover, although the expression ‘stolen away’ might seem to imply that the bars had been removed altogether, that was almost certainly not the case. Wren used the term in the now-archaic sense of implying movement by stealth. In other words, he observed that the hooks onto which the eyes of the tie-bars were threaded had gradually been pulled away from their anchorages in the capitals: several inches of movement would have been possible before the hooks were completely withdrawn, allowing the bars to fall. There is nothing to suggest that happened.
36 Wren, 1715. Ground plan of the crossing, showing a proposal to enlarge the four piers and their bases. South is at the top. Bodleian Library, Oxford: Gough Mss. Wren Society 1934
Wren devised a scheme for restoring strength to the crossing piers, and made a model to demonstrate how this was to be achieved. He did not describe the method in his report, and the model has not survived, but we know that the solution was to be ‘after a manner that will add to their shape and beauty’. We can, however, reconstruct Wren’s proposal in respect of the piers since a surviving plan shows that his solution involved encasing the bowed piers inside new and much larger ones which would of course be plumb.57 His drawing shows that he intended to enlarge the octagonal plinths of the piers from 8½ ft (2.6 m) to 12 ft (3.7 m), and to simplify the mouldings of the shafts, giving them a late medieval profile. [36, 37] As Wren cautioned, ‘this must be first done, otherwise the addition of weight upon that which is already crooked and infirm will make it more so’. Thankfully, the work was never carried out: if it had, the elegant and skilful articulation of the crossing with the four arms of Henry III’s church would have been wrecked.
The crossing piers were plumbed by Hawksmoor in June 1724 and a note appended to Wren’s earlier drawing (1715). Their height was given as 78 ft 10¾ ins (24.04 m), from the pavement to the top of the capitals.58 The amounts by which they were out of upright were recorded,59 but unfortunately it was not noted which face of the pier was measured, there being three possibilities for each. The measurements appear to relate to the deviation from verticality between the top and bottom of each pier, and are not a record of the amount by which they bowed in the middle region. Structural issues will be further explored in the next chapter (p. 57).
37 A. Plan of a thirteenth-century crossing pier and its octagonal plinth, as existing. B. Plan showing Wren’s proposal to encase the medieval pier inside a new and much larger one. Author
Once the piers had been secured, Wren proposed to build a tower ‘according to the original intention of the architect. … In my opinion the tower should be continued to at least as much in height above the roof as it is in breadth’ (i.e. a minimum of 44 ft, or 13.4 m), adding that ‘if a spire be added to it, it will give proper grace to the whole fabric, and the West-end of the City, which seems to want it’. He insisted that he would ‘strictly adhere’ to the Gothic style: ‘to deviate … would be to run into a disagreeable mixture, which no person of a good taste could relish’.60 The timber model, which represented Wren’s proposal, owed much to the central tower and spire of Salisbury Cathedral. However, he added the comment, ‘I have varied a little from the usual form, in giving twelve sides to the spire instead of eight’.61 Wren evidently considered a number of possibilities for detailing the tower and spire, and a plan dated 1715 (drawn to the same scale as the model) shows both the octagonal and dodecagonal options.62 The scheme included detached octagonal stair-turrets at the four corners of the tower. [31] Interestingly, Wren’s assistant titled the drawing ‘plan and section for the “light shell” described in Sir Chr. Wren’s report of 1713’. Given that a major strengthening of the crossing piers was envisaged, the term ‘light shell’ undoubtedly relates to the structure’s function as a lantern, and not to its weight.
It may be noted, en passant, that the phenomenon of bowed crossing piers is by no means unique to Westminster, and it is found in a number of other churches where the piers are tall and slender: e.g. Amiens Cathedral. There too the lofty crossing does not (now) support a masonry tower which would have provided a dead-load: the original termination is unknown, but since the early sixteenth century the crossing has been topped with a relatively lightweight flèche. [6]
William Dickinson’s Contribution
Wren’s ambitious scheme languished, presumably for lack of funds, but also perhaps on account of declining energy in his final years: he died in 1723, aged 90. His successor as Surveyor of the Fabric was Nicholas Hawksmoor, who adopted not only many of Wren’s ideas, but also retained his assistant William Dickinson (Under Surveyor, 1711–25). Several drawings relating to the crossing in the hand of the latter have survived, demonstrating that completing the lantern tower was once again a matter receiving serious consideration. While still working for his old master, Dickinson prepared drawings for at least three schemes: the first two envisaged the erection of a 95-ft (29 m) high square central tower with octagonal corner-turrets and crenellations. Various further options involving a soaring spire and a dome were presented. The third scheme allowed for raising the tower more modestly and surmounting it with a domed lantern and cupola.
38 Dickinson, 1722. Scheme 1 for a crossing tower and spire, showing two options for the fenestration (left and right). WAM (P)907
Scheme 1
Dickinson’s first scheme is dated 1722, and has much in common with Wren’s model. It shows a conventional embattled and pinnacled tower with octagonal corner-turrets. [38, 39] Options for the treatment of the fenestration and parapets are shown to left and right. The tower is also fitted with a tall needle-spire with lucarnes, taking the overall height to 280 ft (85 m). The dodecagonal spire is divided into three registers by decorative bands which are fussy and fragment its outline.63
Scheme 2
This is also dated 1722, and five basic versions were drawn, together with various sub-options. The options were presented by arranging a series of flaps on the drawing which could be opened to reveal further variants.64
Version (i) shows a conventional square tower with crenellations, octagonal angle-turrets and pinnacles, but no spire; it had arcading and lancet windows around the lower two stages, all in the Early English style, and tall, ogival-headed windows in the uppermost stage, based on fourteenth-century prototypes. [40] Drawn alongside this, on a separately attached sheet (not a flap), is an alternative version (ia) for topping the tower with a twelve-sided, ribbed and crocketed lantern-dome surmounted by a spirelet, and having small pinnacles at the four corners. It comprises a mixture of Gothic and classical detail, the latter including acorn and pineapple finials.65 [41]
39 Dickinson, 1722. Two options for the plan of the crossing tower (above eaves level) shown in Figure 38; a thin-walled structure is represented on the left and a thick-walled one on the right. Alternative plans for a spire are also indicated very faintly. WAM (P)907A
40 Dickinson, 1722. Scheme 2, version (i): a crossing tower of three stages with a parapet and pinnacles, but without a spire or dome. WAM (P)909
Version (ii) shows the same tower surmounted by a dodecagonal lantern with a ribbed lead dome, and a cupola with a spirelet. It represents a development of the dome shown in (ia). All this is overlaid on the fainter outline of a tall, plain needle-spire with twelve facets. [42] Two further variations (versions iia and iib) show minor differences to the lantern and cupola.66
41 Dickinson, 1722. Scheme 2, version (ia): option for a dome and spirelet instead of a crenellated parapet. WAM (P)909
Scheme 3
The third scheme is dated 1722 and is more restrained. This envisaged an octagonal drum-like lantern with two tiers of fenestration, modest pinnacles and a lead-covered dome pierced by small windows.67 [43] It had semidetached octagonal stair-turrets at the corners of the tower, rising only to roof-ridge height. What is undoubtedly the plan relating to this scheme is attached to the survey drawing of 8 June 1724, and separately bears the same date.68 [44] The plan is plainly derivative from Wren’s drawing of 1715 (p. 25). [31]
42 Dickinson, 1722. Scheme 2, version (ii): a crossing tower with a lead dome (a variation on Figure 39) and spirelet. A needle-spire is shown as an alternative in the background. WAM (P)909
43 Dickinson, 1722. Scheme 3: an octagonal drum-shaped lantern tower and lead dome over the crossing. WAM (P)908
Scheme 4
An undated ink drawing, evidently by Dickinson, shows in meticulous detail his final scheme (prepared for Wren) for restoring the north transept gable and, above it, faintly drawn in pencil, is an equally detailed elevation of the proposed crossing tower.69 It is square in plan, has two stages with windows (the upper a belfry), crenellated parapets and ornate pinnacles. [45] The drawing, which bears close comparison with Figure 38, may represent Dickinson’s earliest proposal for the lantern tower.
Although a good deal of effort obviously went into formulating these proposals, nothing materialized because Dickinson died in 1725. John James was appointed to succeed him as Under Surveyor,70 eventually becoming Surveyor of the Fabric himself in 1736. It is difficult to draw clear lines of demarcation between the various proposals advanced by Wren, Dickinson and Hawksmoor. Dickinson was in the employ of Wren by 1696, and worked up schemes for him at Westminster, introducing his own proposals in the early 1720s. Hawksmoor, who had joined Wren’s office in c. 1680, borrowed from both men, merging their ideas with his own; he also worked with Sir John Vanbrugh from the late 1690s, adding another source of inspiration to his design repertoire.
Finally, a common source of confusion needs explanation: the two views of Westminster Abbey engraved by Paul Fourdrinier in 1737, purporting to show the spire ‘as designed by Sir Christopher Wren’, bear no resemblance to the 1710 model, or to any drawings known to have emanated from Wren’s office. Allusion to Wren is largely honorific, and acknowledges the fact that he was the originator of the proposal to erect a tower and spire over the crossing. It was Hawksmoor who developed the final scheme for completing not only the lantern but also the western towers. The engravings have been reproduced many times, and the supposed connection with Wren much rehearsed. One of the views is by the hand of John James, and was engraved in the year following Hawksmoor’s death.71 [84] Similarly, the drawing showing the western towers and lantern spire, published by William Maitland in 1739, bore a misleading legend, attributing the design to Wren.72
44 Dickinson, 1724. Plan of the proposed octagonal lantern and spire illustrated in Figure 41. The semi-detached octagonal corner-turrets shown here are smaller than the medieval originals. The larger octagons drawn in outline at the corners represent the plinths of the crossing piers at church floor level. WAM (P)911
