Old Grosvenor House

CHAPTER XIII

Old Grosvenor House

For most of the nineteenth century and the first one and a half decades of the twentieth the principal London home of the Grosvenor family was a large detached house situated on the south side of Upper Grosvenor Street (where it was latterly numbered 33) but also enjoying a long frontage to Park Lane and wide views to the south and west over Hyde Park (Plates 64, 65, 66, 67b, 67c, fig. 55: see also Plates 18, 20c, 28 in vol. XXXIX). Compared with other of London's dynastic mansions Grosvenor House came rather late into the hands of its eponymous owners, and was already more than seventy years old when the Grosvenors acquired it in 1806. For much of that time it had been occupied as a royal residence. But in outward aspect the house vacated by the family in 1916 and demolished in 1927 was essentially of the nineteenth century. Undeniably grand, it was yet never harmonized by thorough rebuilding, and in its juxtaposition of uncompleted building campaigns it stood a little obliquely to the main channels through which the prestige of the nobility usually expressed itself in London.

It is not clear how a detached house came to be built in Upper Grosvenor Street in the first place. As early as 1721 an agreement had been concluded with Major Joseph Watts of the Chelsea Water Works Company to develop the whole block now occupied by the modern Grosvenor House but this was later surrendered without any building having taken place. (fn. 10) Agreements were subsequently made for developing the western part of the Upper Grosvenor Street frontage and the Park Street frontage. (fn. 11) The greater part of the site was not, however, made the subject of any known agreement, and in June 1731 it was leased to Walter, first Viscount Chetwynd, who was already in process of erecting a house there for his own occupation. (fn. 12) Completed in 1732, (fn. 13) this was a squarish detached building set back some ninety feet from Upper Grosvenor Street. Who the builders were and what the house looked like are both unknown, and though the plan may not have been very different from that surveyed in 1805 (fig. 55) there were certainly more rooms on the ground floor in the mid eighteenth century. (fn. 14) In front of the house Chetwynd laid out a funnel-shaped courtyard with a narrow entrance into Upper Grosvenor Street flanked by a pair of porters' lodges. The unwanted land on both sides of this entrance provided sites for Nos. 31, 32 and 34 Upper Grosvenor Street, which also acted as a screen for the house and its adjoining service wing. A stable block was erected along the Park Lane frontage. (fn. 15)

Lord Chetwynd lived there from 1732 until his death in 1736. (fn. 13) Two years later his brother, the second Viscount, sold the house to the third Duke of Beaufort for £8,000. (fn. 16) In order to protect his views southwards the Duke also took leases of all the land between Mount Street and South Street bounded on the west by Park Lane and on the east by the gardens of the newly rising houses on the west side of Park Street. (fn. 17) Rocque's map published in 1746 shows a formal garden falling away to Mount Street from a terrace on the south side of the house. The Beauforts retained the house until 1761 when they agreed to sell it with much of the contents to H.R.H. the Duke of Cumberland (the commander at Culloden) for £15,865. (fn. 14) The sale was not concluded until 1763 but the Duke took up residence immediately. (fn. 18)

An inventory of 1761 (fn. 14) shows the house then had six rooms on the ground floor, apart from the entrance hall and two staircase compartments, a closet and a water closet. This allowed a drawing-room and a large saloon to be accommodated on that floor, as well as a great and a little eating-parlour, a gilt dressing-room and another 'Dressing or Alcove Room'. (Perhaps the greater number of rooms beyond what are shown on the plan of 1805 (fig. 55) is explained by the existence of more than one room where the later plan shows only a great drawing-room.) The principal staircase compartment probably already had the round-ended shape shown in 1805, and was lit by 'four Globe Lamps on the stairs supported by Twisting Dolphins of Mahogany very Grand and neatly carved'. Above, the 'library' of 1805 was evidently used in 1761 as a drawing-room, but the other apartments were bedrooms (two of them for upper servants) and a dressing-room. The basement included a candle room and a room for powdering wigs, and the domestic offices extended into an adjacent service block.

As was often the case, large pier-glasses and chimneyglasses were important in the decorative effect of rooms, as also were the conventional marble slabs on carved frames, those in the drawing-room and saloon having removable green cloth covers reaching to the ground. In three ground-floor rooms ornamented or tabernacled pictureframes are mentioned, the eight in the saloon, at least, being 'all adapted to the room and hangings'. Of the thirteen main rooms (where, however, the furnishings may be incompletely listed), five have carpets mentioned, four being Wilton (two fitted to the rooms) and one Kidderminster. On the ground floor no unified colour schemes are apparent, but on the first floor the former bedchamber of the Duchess of Beaufort was all in crimson—bedfurnishings, chair-upholstery, wall-hangings and curtains; Lady Harriet's room was green; and another room had 'embossed serge' for both bed-furnishings and curtains. On the ground floor silk damask wall-hangings are mentioned in the gilt dressing-room (where they were blue) and in the saloon, but the drawing-room had tapestry hangings. Blue damask was also hung upstairs, in a dressing-room. There was flock paper in the Duchess's crimson bedchamber (where the chimney board was covered with 'India paper'), and paper hangings in another room. Especially on the first floor, most of the curtains were of the kind called 'festoon'. The movable furniture was often of mahogany, but walnut, oak, deal and beech also occurred in important rooms.

The Duke of Cumberland died here of a haemorrhage in October 1765 when about to take the chair at a meeting of the Cabinet, which under the Rockingham administration used regularly to assemble at the Duke's house. (fn. 19) In 1766 his executors assigned the lease to his nephew William Henry, Duke of Gloucester, a brother of George III, who lived in the house until his own death in 1805, (fn. 20) during which time it was known as Gloucester House.

For over forty years, therefore, this house was occupied by members of the royal family, yet no record of any alteration or decorative work has survived except for a rumour in 1779 that 'the additional building at Gloucester House was to cost four thousand pounds'. (fn. 21) But by 1794 the front courtyard had certainly been altered, Chetwynd's straight sides being replaced by curving walls (dressed in 1805 if not earlier with shallow recesses and columns), the entrance widened, the porters' lodges removed and, probably, a tetrastyle portico erected in front of the house (as shown in 1805). Inside on the ground floor some of the rooms were evidently thrown together. (fn. 22) Internal decoration by Sir William Chambers has been postulated, on account of the style of the dining-room ceiling in 1889 (Plate 65b), and although this was certainly not by Chambers (partly because that room only acquired its shape in 1871), there is a hint of work by him here in the Duke of Gloucester's riding house in Hyde Park close to Grosvenor Gate, erected in 1768 (Plate 13a in vol. XXXIX), which was virtually identical with Chambers's riding house for George III at Buckingham Palace of 1763–8. (fn. 23) (fn. 1)

When Robert, second Earl Grosvenor, succeeded his father in 1802 he and his wife were living in the Thamesside mansion at Millbank sometime called Peterborough House. (fn. 24) This basically seventeenth-century building had come to the Grosvenors through Mary Davies but had been first occupied by them only in 1719 and vacated in 1755 for No. 45 Grosvenor Square. (fn. 25) In 1789 the first Earl Grosvenor had bought back the lease of the house at Millbank (recently remodelled on neo-classical lines by a tenant) as a home for his son and heir Lord Belgrave, the future second Earl. (fn. 26)

A proposal to build a large penitentiary on the open ground to the south of the house at Millbank, not carried into effect until 1813–16 but under discussion in the late 1790's, (fn. 27) was probably the reason for the Belgraves' decision, in c. 1800, to look for a house in Mayfair. Both the Duke of Gloucester's and Lord Fitzwilliam's (No. 4 Grosvenor Square) were considered, and in August 1801 the Belgraves declared themselves 'decidedly in favour of Gloucester House'. (fn. 28) There was no immediate prospect of its becoming vacant (the lease did not expire until 1830) but by registering their interest they forestalled the grant of a reversionary lease. Henceforward the presumed future status of the house as a Grosvenor family residence would have to be taken into account by the Grosvenor Board when dealing with applications to renew the leases of adjoining properties. (fn. 29)

On his father's death in August 1802 the second Earl Grosvenor chose to remain at Millbank (fn. 24) (rather than occupy No. 45 Grosvenor Square), but the death of the Duke of Gloucester only three years later provided an unexpectedly early opportunity to acquire his house. By the autumn of 1805 negotiations were under way. Lord Grosvenor's surveyor, William Porden, spent four days in November surveying the house. He thought it oldfashioned and valued it at only £16,231 'or nearly Ten Thousand Pounds less than the price required'. Evidently aware of his slightly invidious position when acting for his employer in the unusual role of 'buyer' rather than 'seller' he was at pains to assure Lord Grosvenor that his valuation, which prospective tenants often thought pitched too high on the ground landlord's behalf, was not here too low: 'I have not been inclined to make it less than I fairly ought. On the contrary, I have believed that I should meet your Lordship's wishes more compleatly, if I stretched the line as tightly as possible, as a few Hundred Pounds were not to be placed in competition of such a desirable object.' (fn. 30)

By February 1806 the purchase had been agreed at a compromise price of £20,000. The assignment of the lease to Lord Grosvenor took place in May 1807. (fn. 31)

Porden's survey of Gloucester House in 1805 revealed a compact building of three storeys over a basement, simply planned around a central staircase compartment. The disposition of the rooms on the ground and first floors is shown in fig. 55. Most of the second floor was given over to ten bedrooms. The offices were in the basement and the service wing. (fn. 32)

As it stood the house was not suitable for immediate occupation by Lord Grosvenor. According to Porden the interior was 'very dirty', and 'not so chearful as the situation would lead one to expect'. (fn. 30) The Times later remembered the house as being 'so gloomy, that it appeared to defy all endeavours to render it light'. (fn. 33) Porden advised that the interior should be painted, whitewashed and repapered, and that in addition to other repairs all the 'ornaments' should be regilded, as they were 'too much tarnished to be used with new Furniture'. He also criticized the doors and windows ('low and narrow when compared with the taste of the present Day'), and commented on the small principal staircase, and the 'heavy, antiquated, but respectable stile' of the finishings in the principal rooms. (fn. 30)

There was evidently no question of a complete rebuilding, such as was already taking place at Eaton Hall, but under Porden's direction a costly programme of alterations and redecoration was put in train. Work began in March 1806 and continued for about two years. (fn. 34) When Lord Grosvenor was out of London Porden kept him informed by letter, often sending ideas and suggestions, as well as details of the work in progress. Rather less sanguine in tone were the letters and reports Lord Grosvenor received from his friend and mentor, Professor John Hailstone, the geologist, who visited the house whenever he was in London (where he had rooms in Albany), and, at Lord Grosvenor's prompting, urged Porden to speed up the work. 'I assure your L.', wrote Hailstone in February 1807, 'I seldom miss a day but I walk to the House and sometimes spur the bold P. till he is ready to lash out at me'. (fn. 35) These letters compensate in some measure for the total absence of any drawing or other illustration of Porden's interior work, but many questions remain unresolved as the replies containing Lord Grosvenor's decisions do not usually survive.

Structural alterations were not very extensive, the most significant being the erection of a two-storey bow topped by a balcony in the centre of the south front. (fn. 36) In early views this bow is shown with a verandah at first-floor level, the roof of which was painted like a striped awning (Plate 67c: see also Plate 18a in vol. XXXIX). The shallow segmental arches over the ground-floor windows on the south front were doubtless of Porden's devising, and the whole of the exterior appears to have been stuccoed. (fn. 37)

Inside, the plan was hardly disturbed, though the uses of the principal rooms were changed. On the ground floor the old drawing-room and breakfast parlour became the dining-room and the ante-dining-room; the former ante-room, newly extended into the bow, became a saloon, the old dining-room was turned into a drawing-room, and the bedroom into an ante-room (later a breakfast room). The two old staircases at the centre of the house were replaced by one grand 'geometrical' staircase, where Porden used the vertical panels of Greek-key pattern which he had recently employed in the staircase of Mrs. Fitzherbert's house in Brighton (Plate 64c). New back stairs were built in an extension on the east side. (fn. 38) For all the size of the house Porden encountered a difficulty in finding a place for a water closet on the ground floor, the lack of which he admitted would be inconvenient. His best suggestion, which may or may not have been adopted, was to put one in the back staircase compartment, accessible from the dining-room via the ante-dining-room. As he pointed out to Lord Grosvenor, after dinner the ladies would retire to the saloon or drawing-room, and 'the Anti-room will be left to the Gentlemen'. (fn. 39)

In the principal apartments the work included the renewal or remodelling and heavy gilding of ceilings and cornices, (fn. 40) and the fitting of new mahogany doors to enlarged doorways. Porden offered Lord Grosvenor a choice between sets of 'single' doors, and folding doors which could be turned back into their jambs 'so that when the whole suite of Rooms are thrown open on a gala night, there would be no projections into the Rooms, but company might walk everywhere without obstruction'. (fn. 41)

In the hall the old chimneypiece was replaced by a niche with a corresponding niche on the opposite wall, and in order to complete the symmetry there Porden wanted to add two blank doors of red baize. He thought it prudent, however, to get Lord Grosvenor's agreement as he and Lady Grosvenor were known to 'dislike the appearance of many doors in a Room'. (fn. 42) The glazing in the hall was to include 'figured devices such as Arms, crests etc.' (fn. 41) The scagliola columns mentioned in the accounts (fn. 34) were probably intended to flank the doorway leading from the hall to the staircase.

Much time and effort was put into devising a decorative scheme which would provide a suitable setting for Lord Grosvenor's rapidly expanding collection of pictures (substantially augmented in 1806 by the purchase of Welbore Ellis Agar's entire collection (fn. 43) ). The paintings were to be displayed in the main rooms, where it was proposed that the walls should be lined with crimson damask salvaged from old Eaton Hall. (fn. 44) 'This will be more expensive than Plaister', Porden told Lord Grosvenor, 'but it will be ready for use immediately, and will afford proper means for fixing the Pictures, and for changing their situations, as often as your Lordship thinks it proper. On Plaister walls that is not quite so easy.' The merit of hangings in this respect was that the pictures could readily be fixed to the framework of wooden battens on which the damask was fastened, thus dispensing with the brass picture-rails, which were 'expensive', and the dependent cords, which Porden thought unsightly. (fn. 39)

When the damask finally arrived from Cheshire late in 1807 it was found to be sufficient for three rooms, the best of it being hung in the drawing-room and ante-diningroom, and the second best in the north-west ante-room (later breakfast room) where there was 'less light … to expose defects'. (fn. 45) At Porden's suggestion the saloon was hung with a crimson velvet (Lord Grosvenor having evidently vetoed blue), and the dining-room, little used by daylight, with a scarlet wallpaper. Porden thought scarlet had the important recommendation that 'it lights up by night to more advantage than Crimson'. (fn. 46) He had painted his own dining-room in Berners Street scarlet, and told Lord Grosvenor 'my dingy drawings look quite brilliant in consequence'. (fn. 42) The carpets were either crimson or scarlet. (fn. 41)

The woodwork of the dadoes, the plinths and the mouldings were painted with 'a tint of satinwood' as harmonising best with the mahogany doors, the hangings and the pictures. (fn. 47) Porden reported to Lord Grosvenor that when the colour was tried out the Claude which he had hung against the crimson damask appeared superior to any he had ever seen. This opinion, he said, was shared by others including William Seguier, the future first curator of the National Gallery, whom Lord Grosvenor had recently engaged to look after his own collection. Porden was so confident of 'the propriety and excellence of the effect' of the 'satinwood' paint that, 'were it not unbecoming in me to say so', he would have offered to repaint the whole suite at his own expense if the colour was not approved. An alternative colour scheme for the painted woodwork in pink and white was evidently proposed by the furniture-maker Richard Gillow, who was refurnishing the house, and appears to have had some temporary support from Seguier, but was condemned by Porden. He thought it gave 'a feeble unfinished effect, and is extremely injurious to the pictures'. (fn. 48)

That Lord Grosvenor should have consulted Gillow on such a matter Porden felt as a slight on his professional judgment. He was touchy about anything which might be construed as interference in his own, widely conceived, sphere of work, (fn. 49) but was free with unsolicited opinions on the work of others (here and at Eaton), excusing himself to Lord Grosvenor on the grounds that he did it 'Con amore', and 'from no other motive whatever, but a desire to see everything in either House as perfect as your Lordship's rank and fortune can command'. (fn. 50) Gillow's furniture was particularly liable to this criticism. In Porden's opinion Gillow was 'an excellent workman' but not being 'a man of superior taste' was 'only governed by fashion', whereas an architect like himself, while not entirely disregarding fashion, was guided by 'the principles of general harmony'. In designing a room the architect had 'in contemplation the effect of the whole, the carpet, the hangings, and the furniture as well as the painting, taking the tone from the principal features that are unalterable, or have been determined on'. (fn. 51) At Grosvenor House these were the pictures and the hangings. Writing 'with that frankness which becomes an architect in the confidence of your Lordship' Porden expressed his disapproval of Lord Grosvenor's choice of Gillow's furniture designs, and offered to undertake the designing himself with the assistance of his son-in-law Joseph Kay. (fn. 52) Lord Grosvenor appears to have ignored most of these criticisms, but Porden did succeed in having the draperies put under the 'active direction' of Kay, 'who understands them both as a Painter and an Upholsterer'. (fn. 53)

In all this Porden did not forget that the rooms were to form a setting not only for pictures on the walls but for 'the living pictures which will frequently adorn them', and reminded Lord Grosvenor of 'the difference that will be made in the appearance of the finest forms and faces when moving on a back Ground of feeble and unharmonious colours; and one that by its contrast and harmony will define the form, and give brilliancy to the face'. (fn. 48)

The internal embellishments were accompanied by a remodelling of the garden under the direction of a Mr. Gray, doubtless of Messrs. Gray and Wear, nurserymen, of Brompton Park, who received £100. (fn. 54) A number of trees were transplanted from Millbank, where the Grosvenors had been enthusiastic gardeners. (fn. 55) In front of the house the courtyard was lowered and newly gravelled, and the gates there repaired. (fn. 56)

By January 1808 the work in the house was sufficiently advanced for a start to be made on hanging the pictures. This, of course, was Seguier's province, and Porden, as he explained, did not interfere, 'farther than to lend assistance if necessary, and to advise if anything occurs that will improve the Arrangement'. (fn. 57) Out of London, Lord Grosvenor fretted over Seguier's disposition of the paintings, fearful that 'many good pictures' would have to be omitted, but perhaps comforted by Porden's assurance that one of the Claudes 'looks divine'. (fn. 58)

The house appears to have been occupied by Lord Grosvenor from April 1808 (fn. 59) and two months later it was thrown open to the 'Fashionable world' amid much 'critical acclaim'. The Times reported that it was 'now transformed into a residence, which combines, in a superior degree, the several qualities of magnificence, elegance, and convenience… . Indeed, it has undergone a metamorphosis, which, under the influence of a pure and solid taste, has produced a splendour, that possesses the happy medium between the cumbersome finery of a former period, and the fillagreen frippery, or motley vagaries of the present day.' (fn. 33) The Morning Post admired the hangings, the 'truly magnificent' chandeliers and Grecian lamps, the vast and beautiful mirrors, the grand staircase, 'superbly illuminated' and 'adorned with the most rare specimens in the art of sculpture', and the richly gilded ornamental ceilings and cornices. (fn. 60) Porden received, by name, his fill of praise. Not everybody was enthusiastic however: Lord Lonsdale, who visited the house in the company of the architect Robert Smirke, found it 'most expensively furnished, but in a bad taste'. (fn. 61)

It was rumoured that the house, together with the alterations, the furniture and pictures, had cost Lord Grosvenor £120,000; (fn. 62) but the true figure seems to have been nearer £80,000. Of this about £35,000 had been spent on pictures, the greater part (£31,000) on the Agar collection. The house itself cost £20,000; the alterations and decorations a further £16,670; and Gillow's furniture over £7,000. (fn. 63) Apart from £7,992 spent on carpenter's, bricklayer's and mason's work, the alterations required large payments for decorative plasterwork to Francis Bernasconi (£2,097), and for painting, glazing and gilding (£2,371). (fn. 2) (fn. 34) Porden's commission added another £833 9s. 8d. to the total cost. (fn. 64) For some reason Lord Grosvenor seems to have queried Porden's charges, provoking a pained response from the architect: 'I beg leave … to assure your Lordship most solemnly that I have not intentionally made any charge … but what I think myself strictly warranted in doing by the practice of my profession'. (fn. 65) Whatever the problem was it did not deter Lord Grosvenor from having Porden design and supervise the erection of a picture gallery here in c. 1817–19.

The earliest surviving reference to the gallery is in a letter from Porden in January 1818, and by then work on what he termed the 'New Room' was well advanced. (fn. 66) Completed in 1819 it consisted of a single toplit compartment just over fifty feet long and of double-storey height attached to the west side of the house. The circumstances in which this unbalanced arrangement was accepted are unknown, but it tended to complicate the task of giving symmetry to any extended south front when a few years later adjacent leases were to fall in.

No illustration is known but the external walls appear from plans to have had a pilastered treatment. Skylights let into the 'Waggon headed ceiling' provided the main source of natural illumination. There was also a small circular window in the west wall, while from a similar window in the east wall Lord Grosvenor could look into the gallery from his first-floor library. (fn. 67)

The natural lighting, about which Seguier had been consulted, was evidently a success and in 1820 Porden was reminded of its merits by the 'mean and I think insufficient' lighting of Soane's picture gallery at Dulwich College. (fn. 68) Artificial lighting for evening receptions was provided by a cut-glass gas chandelier 'of the most chaste and beautiful description', the gas for which was supplied by the Gas Light and Coke Company from their main in Park Street. (fn. 69) (Porden had evidently been unaware of this source when in 1818 he counselled against gas lighting: 'As the Gass-light Company have no Station near Grosvenor Square I fear it cannot be done, without your Lordship making a Gassometer for yourself, which would be a considerable expense in making, and managing'. (fn. 66) ) At Lady Grosvenor's grand rout on 17 May 1819 the chandelier was much admired for its 'mild yet brilliant light' and 'sun-like brightness'. (fn. 70)

Of the pictures in the new gallery the most impressive were four huge religious paintings by Rubens, the smallest measuring 14 feet by 14½ feet, which Lord Grosvenor had bought from the Danish Envoy, Edmund Bourke, in March 1818 for £9,205. (fn. 71) Indeed, it was thought by some that the gallery had been erected expressly to receive them. (fn. 72) This seems unlikely, however, as building was well advanced by March 1818, and in announcing his 'grand purchase' to Hailstone in April, Lord Grosvenor merely says the pictures 'will be in the new Gallery'. (fn. 73) A catalogue of the 143 paintings in the collection published in 1821 shows that the other principal apartments remained plentifully hung with old masters, of which only twentyone, doubtless on account of their size, were to be found in the gallery. (fn. 74)

Porden resigned in 1821, and by 1824 Lord Grosvenor was having his successor as surveyor, Thomas Cundy, sketch plans for improvements to the house. (fn. 75) Lord Grosvenor was a man of great and growing wealth who seems to have been at once free-spending and parsimonious. Thus although the additions to the house made by him were important they fell far short of many of the schemes prepared by Cundy and (more numerously) his son Thomas Cundy II. An element in the rather obscure picture of Lord Grosvenor's and his surveyors' intentions was, of course, the physical presence of the old house with its sunlit, bow-fronted rooms on the south, and the fact that it was nearly but not exactly central between Park Street and Breadalbane House. This was significant because the schemes proposed a Grosvenor House extended eastward and westward (Plate 66a: see also Plate 18b, 18c in vol. XXXIX). Whether or not with this in view, the demolition of adjacent properties had been envisaged since at least 1801 in respect of Breadalbane House stables and since 1813 in respect of houses in Park Street. (fn. 76) Any really fine, symmetrical deployment of a new south front between those limits and properly related to a corresponding interior would thus have required (as many plans proposed) the total rebuilding of the old house, while, for those fond of the old house, a less comprehensive rearrangement was not quite ruled out. The biographer of Lord Grosvenor's daughter-in-law, Lady Elizabeth Belgrave, has quoted a letter written by her when incomplete (and never-to-be-completed) changes had been made, in which she spoke of Lady Grosvenor's reluctance to have the old house rebuilt. (fn. 77) Lady Elizabeth thought 'there is no fear of her dissuading Lord Grosvenor', but the old house survived, and possibly Lady Elizabeth under-estimated her mother-in-law.

Although never followed through in a general reconstruction, the chief work done for Lord Grosvenor was not long delayed, and was impressive so far as it went. This was principally the remodelling and extension of Porden's picture gallery to create a longer and (presumably) grander west wing, for early duplication on the east. The demolition in 1825–6 of Nos. 108–111 Park Street and of the stables at No. 107 Park Street, at Nos. 29–30 Upper Grosvenor Street, and at Breadalbane House made this feasible, (fn. 78) and at the same time permitted the construction of a carriage drive between Park Lane and Park Street below the south front of the house, which was thereby given an importance comparable to that of the north front. (The carriages were brought from new and expensive stables on the south side of Reeves Mews built, with gas laid on, in 1826–7 and 1829–30 at a cost of over £6,000. (fn. 3) (fn. 79) )

The main work included general decoration and repairs to the old house (fn. 80) (in the course of which the plasterer put some Wedgwood tiles in Lord Grosvenor's bath (fn. 81) ), but chiefly related to the picture-gallery wing. (fn. 82) It was carried out in 1826–7, that is, under the supervision of Thomas Cundy II as his father's successor in the position of estate surveyor since the latter's death in December 1825. Internally, Porden's gallery was remodelled with a new ceiling rising in a cove to a large rectangular lighting-compartment with glazed sides. It was very closely similar to the ceiling Thomas Cundy senior had recently designed for the ceremonial staircase at Northumberland House, (fn. 83) but its ornateness seems, at least in illustrations, to anticipate Victorian classicism (Plate 64b). What its construction was is uncertain but it was provided, like the similar ceiling of the other, new, apartment beyond it, not by a plasterer but by William Croggon, the Coade stone maker and scagliolist, who did the work, by the use of moulds, for £493. (fn. 84) The new apartment beyond was the square Rubens gallery to which Porden's gallery was united by a wide opening dressed with fluted scagliola columns and responding pilasters, called 'Yellow Antique' in the bills (fn. 84) but appearing brown in Leslie's well-known painting of 1831. Scagliola (made on the site) was used also for the dado and a 'doorway' in these rooms. (fn. 85) The Rubens gallery carried this wing nearly up to Breadalbane House but by its slightly greater width than Porden's gallery gave the south front a forward break sufficient to make the proximity of that house less apparent.

This south front of the picture galleries was cased in Bath stone and given a palatial aspect (Plate 67b, 67c: see also Plate 18a in vol. XXXIX). Above a horizontally channelled podium the unwindowed walls were dressed with blind aedicular and pedimented window-frames, each having over it a decorative panel containing a festoon (similar to that in the attic pavilions at No. 53 Davies Street). The bays were defined by a sequence of Corinthian columns, free-standing before the wall of Porden's gallery and engaged against that of the Rubens gallery, which rose from the podium and supported a deep continuous entablature. This followed the break in the two parts of the front, and over Porden's gallery was surmounted by an urn-topped balustrade. The Rubens gallery, however, had an attic storey above the entablature, where the bays were defined by pilasters, against which were set figures emblematic of the arts and sciences. These south-facing evocations of antiquity were carved, at a cost of £50 for each of six, by Joseph Theakston. (fn. 86)

When this work was in hand the intention to continue the rebuilding eastward was sufficiently firm for the eastern end of the new front to be left as an irregular edge of masonry where it stood slightly forward of the old house. (fn. 87) Even more significantly, a considerable volume of work was done on a new eastern wing to contain a new dining-room, including the making of the foundations across part of the former curtilage of No. 107 Park Street. (fn. 88) This was not very expensive, but a further £1,428-worth of mason's work was executed in such a form that it could be held against future use. (fn. 89)

The whole work cost nearly £23,000. (fn. 4) (fn. 90)

The renovated and extended picture galleries were opened (although not quite finished) in June 1827, (fn. 91) but the further intended progress had already been brought to a halt. Early that year Lord Grosvenor had written to his agent in London putting a brake on expenditure—'I conclude Cundy has stopped the left [east] wing at its level, and let him understand directly I have no wish to hurry the finishing of the Galleries … tell Cundy I am particularly anxious nothing should be done or undone and no unnecessary expense, no undue haste particularly'. (fn. 92) How far this check, which was perhaps related to Lord Grosvenor's great outlay about that period on land purchases in other parts of Great Britain, was intended to be final is unknown, but so it turned out, as far as the east wing was concerned. Possibly a factor in this was a damaging criticism of Cundy's proposed eastward extension sent to Lord Grosvenor later in the same year by Lord Farnborough, a politician with a reputation as a connoisseur. (fn. 93) He professed to admire the new west wing but thought the proposed east wing, of which he had been shown plans, would not 'accord' with it. This probably means that the plans were among those which retained the nucleus of the old house and either (like one surviving scheme, Plate 66a) did not repeat the elevation of the Rubens room (fn. 94) or (like others) achieved a false semblance of symmetry by compressing the intermediate elevation, corresponding to that of Porden's gallery, into a lesser frontage. (fn. 95) Lord Farnborough had taken Robert Smirke to look at the possibilities, (fn. 93) and so nineteen years after Smirke had assisted one patron to an adverse view of Porden's work he was given the opportunity by another to put Cundy right. Quite unsolicited by Lord Grosvenor, Smirke sent him two schemes, one armoured in rich unfaltering Corinthian and the other showing a more expressively articulated use of the same order. (fn. 96) (fn. 5) The former gave a ground-floor plan of unembarrassed magnificence, with a domed and circular centrally placed salon, a great D-shaped staircase compartment, and a stupendous 160-foot drawing-room along the south front. Perhaps indicative, however, that Lady Grosvenor was indeed a factor in the situation is the contrivance to which Smirke was driven to fit his first-floor plan to his other elevation (for which alone it exists). Cundy, in a complete scheme made in the previous year, (fn. 97) and doubtless influenced by the style of his lately deceased father, had similarly produced a calm north front (Plate 66c) which on the first floor concealed an awkward expedient behind its widely spaced windows. The areas of difficulty in Cundy's and Smirke's plans were differently located, but each was where Lady Grosvenor was to have her dressing-room.

Whether or not he sensed the imminence of criticism, Cundy was in fact already producing a variety of plans adapted to widely differing intentions on his employer's part and ranging from rather ignoble devices to retain the old house, with its apparent centre pushed westward, to extensive grandiose rebuildings. (fn. 98) Of the latter, most envisaged, like Smirke's plans, a circular domed hall, sometimes designated for sculpture. Without exception, all the plans retained Porden's gallery and the Rubens gallery. In a number of schemes, however, these were to be masked by a whole new range of apartments along the south side (Plate 18b in vol. XXXIX). Other plans gave a large new wing extending southwards: in one version this is developed into a Fonthill-type cruciform arrangement, (fn. 99) and in others it becomes part of a great splay-sided courtyard facing towards Park Lane. (fn. 100) Some schemes, in 1829, included slightly Cheltonian elevations towards Park Lane of the Corinthian order, pleasantly enlivened with carved panels and balconies or verandahs (Plate 66b), but perhaps over-high and cramped in their proportions. (fn. 101)

Nothing happened until 1833–4, and then possibly not very much. Cundy prepared plans and specifications at the request of the Marquess of Westminster (as Lord Grosvenor had become in 1831) for the completion of the house to a limited programme. (fn. 102) The staircase of the old house was in bad structural condition; nevertheless the existing nucleus of the house was to be retained with a rebuilt, flat-fronted centre and a new east wing mirroring, except for slightly squeezed lateral dimensions, that on the west. By 1834 this scheme had dwindled to one by which the old house would have been given a new bowed front set forward in modified replica of the old, and perhaps dressed with columns already prepared for the east wing and lying at Belgrave Wharf. In May 1834 the house was evidently unoccupied and 'out of repair', (fn. 103) but although some work was done which included the provision of new gas supplies its extent is unknown. (fn. 104) The existing south front survived, however, without significant alteration.

It is said that in 1835 Anthony Salvin did work on the house to the value of £5,900, (fn. 105) but it is not known where. In any case this seems not to have signified any supersession of Cundy, who in 1842–3 was responsible for a major work on the north side of the house. One of the earliest 'Cundy' plans, not later than January 1826 and perhaps expressing Thomas Cundy I's manner, had provided for a new square courtyard opening in its full width to Upper Grosvenor Street, from which it was to be screened by a colonnade with two main entrances, in the manner of the then still-existing Carlton House screen. (fn. 106) Now a modified version of this earlier 'Cundy' screen was erected by Thomas Cundy II, substituting a Roman Doric for an Ionic order (Plate 20c in vol. XXXIX). The entrances for vehicles, under open pediments bearing the Grosvenor arms, were closed by splendidly modelled cast-iron gates (Plate 64a), each pair weighing six tons and painted like bronze. Similarly painted were the seven cast-iron gas candelabra between the columns. Sculptured panels of the Four Seasons were set over the two smaller entrances, two panels facing the street and two the courtyard. (fn. 107) The builder was evidently John Elger of South Street, who was paid nearly £10,000 in 1843 for unspecified work at Grosvenor House. (fn. 108)

In that year The Athenaeum found fault with the supposed motive of economy that had dictated the use of painted cast iron for the gates and lamps. 'Why were they not of bronze? Surely a hundred pounds more or less could be no consideration.' (fn. 109) With this reservation the screen was enthusiastically received by periodicals for its dignity without pretension. (fn. 6) The Art-Union did, however, draw attention to the contrast between the screen's monumentality and the 'paltry' treatment of the entrance courtyard, (fn. 110) where the walls were stuccoed and decorated in a simple Belgravian manner with a blind arcade which continued across the front of the house.

Cundy had in fact prepared yet more plans in 1842 for a rebuilding of the house 'at a future period' which would have extended to an elaborate reshaping of the courtyard with fountains, statues in quadrant arcades, and a recessed entrance from Park Street. The plan of the house itself would have been predominantly in a square of 140 feet composed around a great central staircase in an arrange ment similar to that at the then recently completed Stafford House. The three-storeyed elevations in an early-Victorian Palladian style looked wealthy but lacked the old-fashioned gravity of the screen. (fn. 111) The garden, enlarged by the demolition of the old stables along Park Lane in 1827, (fn. 112) was proposed to be in the formal manner of Barry and Nesfield, and contrasted with the irregular, romantic plantation of conifers and ornamental trees sketched on the plan perhaps of January 1826. (fn. 113)

Again, except for the entrance screen, none of these suggestions was taken up, either by the first Marquess or, after his death in February 1845, by his son the second Marquess. Cundy prepared an 'amended' design in May 1845 for a new house (Plate 66d), giving a great continuous sequence of ceremonial rooms all round the first floor, (fn. 114) and alterations by him to the old house were undertaken in 1847–8 by the builder Reading Watts of Motcomb Street. Evidently, however, they were not very far-reaching, although a staircase was rebuilt. (fn. 115) In 1859 the Marquess had Frederick Ransome of Whitehall Wharf, Cannon Row, apply his patented preservative to the stonework of the portico and screen, (fn. 116) and this is the sum of known work for the second Marquess, elsewhere so implacable an improver of his Mayfair estate.

On his death in October 1869 the house passed to his eldest son, the third Marquess and, from 1874, the first Duke of Westminster. Perforce abandoning an intention to take a house just built for him in Grosvenor Place in the Second Empire style of Thomas Cundy III, he removed into the family mansion from his home in Princes Gate. The Marquess brought with him from South Kensington a willingness to use his great wealth in the patronage of the arts and sciences, particularly the arts of building. His thirty-year reign had a great effect on the architectural character of his estate, both in Mayfair and the country, but although one of his first works was the alteration of Grosvenor House itself, this again left the original house still standing, and the exterior not essentially very different from what it had been like in 1843.

It was significant of the new Marquess's attitude, and perhaps of the age, that he did not entrust these alterations to the estate surveyor, now Thomas Cundy III, who was relegated to building a staircase against the east wall of the house (fn. 117) —a foretaste of the circumscribed view of his role taken by his new master. For the major changes an independent architect was called in. The choice of Henry Clutton (fn. 117) was not perhaps a very obvious one, but he had recently worked for the Marquess's father-in-law, the Duke of Sutherland, at Cliveden, and was, moreover, a discerning designer for the less straightforward type of commission, with a capability extending well outside the ecclesiastical French Gothic for which he is best known.

The interior work began in 1870, with Messrs. I'Anson as the principal contractors, and continued until 1873, although the main apartments were in use by the summer of 1872. (fn. 118) The best evidence of what was done by Clutton is photographs taken in 1889, (fn. 119) and a few comments in periodicals and elsewhere. Direct documentary testimony to the extent and nature of the work (or its cost) is lacking, and it is chiefly negative evidence which indicates that Clutton did not rebuild the old house, and made few changes to its upper floors, where a glimpse of the library in 1875 seems to have reminded one architectural commentator how Clutton's manner below differed from his predecessors'. (fn. 120) Clutton's work did strengthen one existing characteristic of the house, particularly as extended by Porden and Cundy, that its sequence of state rooms, and its chief grandeur, was on the ground floor. Apart from the library, the rooms on the first floor and above were bedrooms or suchlike.

On the ground floor at least Clutton made some alterations to the existing plan (fig. 55). He enlarged the entrance hall by extending it forward, and at the same time added a tetrastyle Roman Doric porte cochère on the courtyard front. On the east of the hall a complete rearrangement provided the previously lacking servery and a longer dining-room differently lit (Plate 65b). This was one of the few rooms given more space for pictures on its walls, for although the collection of old masters remained a great glory of the house the changes brought about some modification of Grosvenor House's physical shaping as a mansion with a picture gallery attached to it. Change was in fact greatest in the two parts of the picture gallery, which were separated more distinctly as two large apartments of the normal reception-room type (Plate 65a, fig. 55: see also Plate 28b in vol. XXXIX). The south wall of the gallery and the south and west walls of the Rubens room were opened up by long windows, those on the south being shaded by special blinds 'braided to patterns as Mr Holford's [at Dorchester House] with orange webbing'. (fn. 121) Cundy's toplighting was abolished, and ceilings powerfully designed by a hand probably Clutton's were inserted. To give access to these rooms additional to that through the old house Clutton made a corridor on the north side communicating with the entrance hall via an ante-room. This was evidently in recognition of the semi-public access to these rooms encouraged by the Marquess. The Rubens room was much used for meetings and charity concerts (fn. 122) and it was perhaps on account of the latter that a 'Green room' was built nearby a few years afterwards. (fn. 123) In August and September 1876 'designers, artisans and the like employed in any branch of art applied to productive industry' were being allowed to inspect the collection every afternoon. (fn. 124)

This South Kensingtonian viewpoint found some expression in the decorative style of the interiors. Chiefly, however, the rooms emerged from Clutton's hands showing no one manner of designing more closely defined than by an opulent classicism, although the lack of earlier illustrations of most of the rooms makes it difficult to be sure how far this was the result of Clutton's discrimination or of the part survival of previous schemes.

The walls, hung with silk and damask, or, in the diningroom and saloon, with stamped leather, were not architecturally treated, the pictures in their gilded frames forming the chief decoration. The ceilings, however, were elaborately worked. That in the dining-room was reminiscent of Sir William Chambers, whereas those of the saloon (Plate 65c) and drawing-room were of a Raphaelesque type that might have emanated from the followers of the late Prince Albert, and that in the ante-room a more conventional Victorian classic design, retaining, however, the painted panels and arabesques that betokened the artistic use of the room. This ceiling was very carefully coloured, and probably designed, by J. G. Crace to respect the paintings hung against walls of dark red silk below it. As Crace's son wrote later, 'The general colouring of the ceiling is in low tones of cream-colour, verging on drab, with gold' so that 'the quiet general tone gives full value to the skies and lights in the pictures'. (fn. 125) The more vivacious drawing-room ceiling (and, no doubt, that of the saloon) was also by Crace. (fn. 126) The ceiling in the entrance hall, of sober early-Georgian type, rose above a Doric entablature in a cove painted, presumably by Crace, in staid Victorian panels (Plate 64c). In the gallery and Rubens room it seems probable that the forceful and virile modelling of the ceiling was Clutton's. The Albertine ideal was represented in the gallery by a deep frieze painted on canvas and illustrating the arts and sciences. It was executed (and signed) in 1872 by Félix Joseph Barrias. Among his works is said to have been some painting in 'the chapel of the Jesuits in London', (fn. 127) which conceivably gave him a link with Clutton, while his painting of a ceiling in the Drapers' Hall in 1869 must have brought him into contact with Crace who was also involved in the redecoration there.

According to The Architect in May 1871 the ceilings in the gallery and Rubens room were 'entirely made of Desachey's fibrous plaster upon a strongly constructed frame'; but only a few months earlier the same magazine had described the gallery ceiling, under manufacture at Messrs. Jacksons, as of papier mâché. Both ceilings were 'quite independent of the general construction of the building', being suspended from iron girders, and had the remarkable property that 'a series of wheels and pulleys are arranged by which the ceilings can be lifted bodily when required without trouble or expense'. Supposedly (although not very intelligibly) the purpose was to permit the rooms to be heightened by some four feet in a future and larger rebuilding. To the same end 'every portion of the joiners' work and fittings has also been prepared so as to be moveable' (fn. 128) —that is, as the house-carpenter said much later, they were all 'screwed up'. (fn. 129) On the same good testimony, 'there couldn't be better work and materials anywhere than in these rooms—they are A 1'. (fn. 129)

Indeed, much of the effect, particularly in the sequence of drawing-room, gallery and Rubens room, came from the quality and finish of the materials and furnishings —the heavy and profuse gilding, the high French-polish of the Spanish mahogany doors and the sheen of the brass candlestands and chandeliers. This brassware was partly of English and partly of French make: the enormous candelabrum in the Rubens room is said to have held 190 candles and to have weighed two and a half tons. (fn. 129) (fn. 7)

At first the illumination was by candles and oil lamps (fn. 129) but the Marquess (or rather the first Duke, as he had become) was an early convert to electricity. He described its merits with engaging enthusiasm to his daughter-in-law after seeing a demonstration at the Edison company's office in January 1882. 'Edison's Electric lighting', he wrote, 'is the best thing out, and apparently perfect for house lighting everywhere, I mean for rooms, passages, everywhere, no more lamps nor candles nor steam nor nothing! and all perfectly safe you may lay hold of the wires with perfect impunity—delightful. (fn. 130) Electricity had been installed by 1889 (and the great candelabrum in the Rubens room sold (fn. 129) ). In these first days of electric lighting the bulbs were left unshielded and photographs show them dotted along the lines of the cornices or ceiling compartments as well as clustered in the candelabra.

New (and a little second-hand) furniture and furnishings were supplied by Holland and Sons to the value of some £6,481. (fn. 131)

In 1876–7 Breadalbane House, which had hitherto masked part of the western end of Grosvenor House, was demolished. Clutton thereupon designed a 'loggia' here, built by Cubitts in 1880–2, in the form of an open hemicycle dressed with Roman Ionic columns two deep and enriched with swags dependent from the volutes (fn. 132) (Plate 28a in vol. XXXIX): possibly there was some echo of schemes by Thomas Cundy II for an apsidal treatment of the proposed eastern wing dressed with columns. (fn. 133) Conspicuous from Park Lane, this piece of pomp enhanced the palatial effect of the exterior, even if it came rather close to the kind of thing built under the influence of palatial aspirations in the City. (fn. 8) The cost of what was then done seems to have been about £7,811 paid to Cubitts and £2,570 to Bailey and Sons for the Park Lane railings. Possibly this addition was meant to compensate for some simplification of architectural effect on the south front. Clutton had brought forward the wall-face of the Rubens room and gallery to engage such of Cundy's columns as were previously free-standing, and channelled the stone work horizontally, with a rather French effect, but dispensed with Cundy's statues, urns and decorative panels (Plate 66e: compare Plate 18a in vol. XXXIX). Southward lay the garden, 'green and pleasant and full of pigeons'. (fn. 134)

Little change seems to have been made by 1914. The house then still preserved the traditional arrangement of maidservants' bedrooms on the top floor, and menservants' bedrooms in the basement, which was an extensive region on the old pattern but with specialities like a 'fruit room' and a 'visitors' valets' brushing room'. (fn. 135) On the outbreak of war the second Duke of Westminster offered to put Grosvenor House and Eaton Hall at the disposal of the Government. Two years later, in December 1916, Grosvenor House was occupied by Government departments. (fn. 136) The Duke meanwhile took up residence at Bourdon House in Davies Street, and even before Grosvenor House was released from Government occupation, in 1920, he had decided not to live there again, but to stay in Davies Street. (fn. 137)

Although it did not seem quite certain at the time, this sealed the fate of the old mansion. The steps by which its intended replacement took shape as a great commercial development over the whole site are traced on pages 270–1. In 1921 The Times was prophesying that this is what would happen (fn. 138) (and some of the great pictures were in that year sold to Duveen (fn. 139) ). It was autumn 1924 when a rebuilding lease for the site was agreed, the keys of the empty house were handed over in April 1925, and after standing for a year or two in the advancing shadow of the new building old Grosvenor House was pulled down in the autumn of 1927. (fn. 140) (fn. 9)

Occupants include: 1st Viscount Chetwynd, 1732–6: his brother, 2nd Viscount, 1737–8. 3rd Duke of Beaufort, 1738–45: his brother, 4th Duke, 1745–56: the latter's wid. and their son, 5th Duke (a minor), 1756–60. William Augustus, Duke of Cumberland, son of George II and commander at Culloden, 1761 5. William Henry, Duke of Gloucester, son of Frederick, Prince of Wales, 1766–1805: his wid., Maria, Dow. Duchess, 1805–6. 2nd Earl Grosvenor, latterly 1st Marquess of Westminster, 1808–45 : his son, 2nd Marquess, 1845–69: the latter's son, 3rd Marquess, latterly 1st Duke of Westminster, 1869–99: the latter's grandson, 2nd Duke, 1899–1916.