Chapter 1: Romantic Classicism Around 1800

DESPITE the drastically reduced production of the years just before and after 1800, between the outbreak of the French Revolution and the termination of Napoleon’s imperial career, there are prominent buildings in many countries that provide fine examples of Romantic Classicism in its early maturity; others, generally more modest in size, give evidence of the vitality of the Picturesque at this time. Since England and America were least directly affected by the French Revolution, however much they were drawn into the wars that were its aftermath, they produced more than their share, so to say, of executed work. French architects before 1806 were mostly reduced to designing monuments destined never to be built or to adapting old structures to new uses.

The greatest architect in active practice in the 1790s was Sir John Soane (1753-1837), from 1788 Architect of the Bank of England. The career of his master, the younger Dance, was in decline; he had made what were perhaps his greatest contributions a good quarter of a century earlier. Whatever Soane owed to Dance, and he evidently owed him a great deal, the Bank[17] offered greater opportunities than the older man had ever had. His interiors of the early nineties at the Bank leave the world of academic Classicism completely behind (Plate 3). His extant Lothbury façade of 1795, with the contiguous ‘Tivoli Corner’ of a decade later—now modified almost beyond recognition—and even more the demolished Waiting Room Court (Plate 4A) showed that his innovations in this period were by no means restricted to interiors.

Soane’s style, consonant though it was in many ways with the general ideals of Romantic Classicism, is a highly personal one. At the Bank, however, he was not creating _de novo_ but committed to the piecemeal reconstruction of an existing complex of buildings, and controlled as well by very stringent technical requirements. Thus the grouping of the offices about the Rotunda, like the plan of the Rotunda itself, goes back to the work done by his predecessor Sir Robert Taylor (1714-88) twenty years earlier; while the special need of the Bank for various kinds of security made necessary both the avoidance of openings on the exterior and a fireproof structural system within. The architectural expression that Soane gave to his complex spaces in the offices which he designed in 1791 and built in 1792-4 had very much the same abstract qualities as those to which older masters of Romantic Classicism, such as Ledoux and Dance, had already aspired in the preceding decades (Plate 3). The novel treatment of the smooth plaster surfaces of the light vaults made of hollow terracotta pots, where he substituted linear striations for the conventional membering of Classical design, was as notable as the frank revelation of the delicate cast-iron framework of his glazed lanterns (see Chapter 7). These interiors have particularly appealed to twentieth-century taste, while Soane’s columnar confections of this period generally appear somewhat pompous and banal.

The Rotunda of 1794-5 was grander and more Piranesian in effect; thus it shared in the international tendency of this period towards megalomania. So also the contemporary Lothbury façade, with its rare accents of crisply profiled antae and its vast unbroken expanses of flat rustication, is less personal to Soane and more in a mode that was common to many Romantic Classical architects all over the Western world. The original Tivoli Corner of 1805, however, was almost Baroque in its plasticity, with a Roman not a Greek order, and a most remarkable piling up of flat elements organized in three dimensions at the skyline that could only be Soane’s.

On the other hand, the reduction of relief and the linear stylization of the constituent elements of the Loggia in the Waiting Room Court of 1804, equally personal to Soane, illustrated an anti-Baroque tendency to reduce to a minimum the sculptural aspect of architecture (Plate 4A). Planes were emphasized rather than masses, and the character of the detail was thoroughly renewed as well as the basic formulas of Classical design that Soane had inherited. This was even more apparent in the New Bank Buildings, a terrace of houses, begun in 1807, that once stood across Prince’s Street. Except for the paired Ionic columns at the ends, conventional Classical forms were avoided almost as completely as in the Bank offices of the previous decade, and the smooth plane of the stucco wall was broken only by incised linear detail.

Perhaps the most masterly example of this characteristically Soanic treatment is still to be seen in the gateway and lodge of the country house that he built at Tyringham in Buckinghamshire in 1792-7 (Plate 6A). There the simple mass is defined by flat surfaces bounded by plain incised lines. The house itself is both less drastically novel and less successful; various other Soane houses of these decades have more character.

Summerson has claimed that Soane introduced all his important innovations before 1800. However that may be, there is no major break in his work at the end of the first decade of the century, nor did his production then notably increase. It is therefore rather arbitrary to cut off an account of his architecture at this point; but it is necessary to do so if the importance of the Picturesque countercurrent in these same years, not as yet of great consequence as an aspect of Soane’s major works, is to be adequately emphasized. His concern with varied lighting effects, however, if not necessarily Picturesque technically, gave evidence of an intense Romanticism; more indubitably Picturesque was his exaggerated interest in broken skylines.

While Soane’s work at the Bank was proceeding, in these years before and after 1800, James Wyatt (1746-1813), capable of producing at Dodington House in 1798-1808 a quite conventional example of Romantic Classicism, was building in the years between 1796 and his death in 1813 for that great Romantic William Beckford the largest of ‘Gothick’ garden fabricks, Fonthill Abbey in Wiltshire.[18] This was a landmark in the rise of the Gothic Revival. In 1803 S. P. Cockerell (1754-1827), otherwise far more consistently Classical than Wyatt, was erecting for his brother, the Indian nabob Sir Charles Cockerell, a vast mansion in Gloucestershire in an Indian mode. The design of Sezincote was based on early sketches made by the landscape gardener Humphry Repton (1752-1818) and all its details were derived from the drawings Thomas Daniell (1749-1840) had made in India fifteen years before and published in _The Antiquities of India_ in 1800. The ‘Indian Revival’ (so to call it) had little success; in these years only the stables built in 1805 by William Porden (_c._ 1755-1822) for the Royal Pavilion at Brighton followed Sezincote’s lead.

The Neo-Gothic of Fonthill, however, a mode that had roots extending back into the second quarter of the eighteenth century, is illustrated in a profusion of examples by Wyatt, Porden, and many others. None, however, seems to have succeeded as well as Beckford and Wyatt at Fonthill in achieving the ‘Sublime’ by mere dimension. The characteristic Gothic country houses of this period were likely to be elaborately Tudor, like Wyatt’s Ashridge begun in 1808 and Porden’s Eaton Hall of 1803-12, or lumpily Castellated like Hawarden of 1804-9 by Thomas Cundy I (1765-1825) and Eastnor of 1808-15 by Sir Robert Smirke (1781-1867). The last, moreover, differs very little from Adam’s Culzean of 1777-90.

Some Gothic churches were built in these decades, too, as others had been ever since the 1750s. Such an example as Porden’s church at Eccleston of 1809-13, while more recognizably Perpendicular, lacked the brittle charm of the earlier ‘Gothick’ churches of the eighteenth century.

The virtuoso of the Picturesque mode and, after Soane, the greatest architectural figure of these years in England, was John Nash (1752-1835). Working in partnership with Repton for several years at the turn of the century, he turned out a spate of Picturesque houses, many of them rather small, with various sorts of medieval detail: Killy Moon in Ireland, built in 1803, is Norman; more usually they are Tudor or at least Tudoresque: his own East Cowes Castle on the Isle of Wight, which was begun in 1798, for example, or Luscombe in Devonshire, begun the following year. The medieval detail was probably designed by the French émigré Augustus (Auguste) Charles Pugin (1762-1832), whom Nash employed at this time (see Chapter 6). It is rather for their asymmetrical silhouettes and for the free plans that this asymmetry encouraged, however, than for the stylistic plausibility of their detailing that these houses are notable.

Finer than such ‘castles’ is Cronkhill, which Nash built in 1802 at Atcham, Salop. Here the varied forms are all more or less Italianate, and the whole was evidently inspired by the fabricks in the paintings of Claude and the Poussins—literally an example of ‘picturesque’ architecture. Actually more characteristic of the Picturesque at this time, however, is the Hamlet at Blaise Castle. There Nash repeated in 1811 a variety of cottage types that he had already used individually elsewhere, arranging them in an irregular cluster (Plate 50A).

The Rustic Cottage mode, like so many aspects of the Picturesque in architecture, had its origins in the fabricks designed to ornament eighteenth-century gardens. But the mode had by now attained considerable prestige thanks to the writings of the chief theorists of the Picturesque,[19] Richard Payne Knight (1750-1824) and Uvedale Price (1747-1829). Their support was responsible also for the rising prestige of the asymmetrical Castellated Mansion and the Italian Villa; indeed, Payne Knight’s own Downton Castle in Shropshire of 1774-8 is both Castellated and Italianate. The appearance of several prettily illustrated books on cottages[20] in the nineties provided a variety of models for emulation, and from the beginning of the new century the Cottage mode was well established for gate lodges, dairies, and all sorts of other minor constructions in the country.

For larger buildings a definite Greek Revival was now beginning to take form within the general frame of Romantic Classicism. More young architects were visiting Greece and, for those who could not, two further volumes of Stuart and Revett’s _Antiquities of Athens_, appearing in 1787 and in 1794, and the parallel _Ionian Antiquities_, which began to be issued in 1769, provided many more models for imitation than had been available earlier. The Greek Doric order had first been introduced into England by Stuart himself in 1758 in the Hagley Park temple, as has been mentioned earlier; a little later, in 1763, he used the Greek Ionic on Litchfield House which still stands at 15 St James’s Square in London. From the nineties, the Greek orders were in fairly common use, as such a splendid group as the buildings of Chester Castle, of 1793-1820 by Thomas Harrison (1744-1829), handsomely illustrates. However, the handling of them was not as yet very archaeological.

Summerson credits the attack made by the connoisseur Thomas Hope (1770?-1831) in 1804 on Wyatt’s designs for Downing College, Cambridge, with helping to establish a more rigid standard of correctness. However that may be, the winning and partly executed design of 1806-11 for this college by William Wilkins (1778-1839) well illustrates the new ideals. Wilkins had made his own studies of Greek originals in Sicily and Southern Italy, and was publishing them in the _Antiquities of Magna Graecia_ at this very time (1807). The inherited concepts of medieval college architecture, largely maintained through the earlier Georgian period, were all but forgotten at Downing. The group was broken down into free-standing blocks, each as much like a temple as was feasible, and repeated Ionic porticoes provided almost the only architectural features. There was no Soanic originality here, no Picturesque eclecticism; perhaps unfortunately, however, this provided a codified Grecian mode which almost anyone could apply from handbooks of the Greek orders.

Wilkins was also responsible for the first[21] British example of a giant columnar monument, the Nelson Pillar of 1808-9 in Dublin. This 134-foot Greek Doric column in Sackville (now O’Connell) Street, of which the construction was supervised by Francis Johnston (1760-1829), initiated a favourite theme of the period usually, and not incorrectly, associated with Napoleon (see Chapter 3).

The Covent Garden Theatre in London was rebuilt in 1808-9 by Smirke. This pupil of Soane had, like Wilkins, seen ancient Greek buildings with his own eyes and generally aimed to imitate them very closely. His theatre was somewhat less correct than the Cambridge college, but despite the castles he had built it was Smirke rather than Wilkins who carried forward the Grecian mode at its most rigid through four more decades (see Chapter 4). Wilkins, however, at Grange Park in Hampshire in 1809 had shown, as C.-E. de Beaumont (1757-1811) had done at a country house called ‘Le Temple de Silence’ just before the Revolution in France, how the accommodations of a fair-sized mansion could be squeezed inside the temple form (admittedly with some violence to the latter). Grange Park provided an early paradigm of a Grecian domestic mode destined to be curiously popular at the fringes of the western world in America, in Sweden, and in Russia, but very rarely employed in more sophisticated regions (see Chapter 5). The house was much modified by later enlargements of 1823-5 by S. P. Cockerell and of 1852 by his son C. R. Cockerell (1788-1863).

Grecian design descended slowly to the world of the builders. The relatively restricted urban house-building of the two decades before Waterloo maintained a close resemblance to that of the 1780s. Russell Square in London, built up by James Burton (1761-1837) in the first decade of the new century, does not differ notably from Bedford Square of twenty years earlier—probably by Thomas Leverton (1743-1824)—except that the façades are smoother and plainer. But a still greater crispness of finish could be, and increasingly was, obtained by covering terrace houses—as for that matter most suburban villas also by this time—with stucco. In this respect the work of some unknown designer in Euston Square in London, which was built up at the same time as Russell Square, may be happily contrasted with Burton’s (which has in any case been much corrupted by the introduction around 1880 of terracotta door and window casings).

In industrial construction, such as the warehouses by William Jessop at the West India Docks, begun in 1799, and those by D. A. Alexander (1768-1846) at the London Docks, begun in 1802, the grandeur and simplicity characteristic of Romantic Classicism can be seen at their best.[22] These warehouses also presage the importance of commercial building in a world increasingly concerned with business (see Chapter 14).

During the years of the American Revolutionary War, 1776-83, years in which Romantic Classicism was maturing in France and in England, North Americans were not entirely cut off from the Old World. Not only did many earlier cultural ties remain unbroken—while a surprising reverse emigration of good painters from the New World to the Old occurred—but new cultural ties with the French ally were established, and these were maintained and reinforced by several émigrés of ability who arrived in the 1790s. Thomas Jefferson (1743-1826), hitherto as confirmed a Palladian as any English landowner of the mid eighteenth century, was undoubtedly influenced by his friend Clérisseau when he based his Virginia State Capitol[23] of 1785-96 at Richmond very closely on the best preserved ancient Roman structure that he had seen in France, the Maison Carrée at Nîmes, even though he used for the portico an Ionic instead of a Corinthian order. In this first major public monument initiated in the new republic Jefferson’s drastic aim of forcing all the requirements of a fairly complex modern building inside the rigid mould of a Roman temple was more consonant with the absolutism of the French in this period than with the rather looser formal ideals of the English.

Jefferson was not able to impose so rigid a Classicism on the new Federal capital of Washington at its start, despite the efforts of various French and British engineers, architects, and amateurs who participated in the competitions of 1792 for the President’s House (White House) and for the Capitol and who worked on the latter during its first decade of construction. The White House[24] as designed by the Irish architect James Hoban (_c._ 1762-1831) was still quite in the earlier eighteenth-century Anglo-Palladian manner, and Jefferson’s own project was based on Palladio’s Villa Rotonda. Neither the English amateur William Thornton (1759-1828) and his professional assistant who was also English, George Hadfield (_c._ 1764-1826), nor their French associate É.-S. Hallet succeeded in giving the Capitol[25] a very up-to-date character (Plate 82A). Yet it is these major edifices that still occupy two of the focal points in the Washington city plan,[26] which was prepared by the French engineer P.-C. L’Enfant (1754-1825) before his dismissal from public service in 1792.

It was Benjamin H. Latrobe (1764-1820), an English-born architect of German and English training, who finally brought to America just before 1800, and shortly to Washington, the highest professional standards of the day and a complete Romantic Classical programme. Indeed, he almost succeeded in making Romantic Classicism the official style in the United States for all time; at least it remained so down to the Civil War in the sixties, and a later revival lasted, as regards public architecture in Washington, from the 1900s to the 1930s (see Chapter 24). A pupil of S. P. Cockerell, Latrobe emigrated in 1796 and was soon assisting Jefferson on the final completion of the Virginia State Capitol as well as undertaking the construction of canals as an engineer. Not inappropriately Latrobe’s first important American building, the Bank of Pennsylvania begun in 1798, was also an Ionic temple, but with an order that aspired to be Greek. This Philadelphia bank included a great central hall whose saucer dome, visible externally, made it a more complex and architectonic composition than the Richmond Capitol. The flat lantern crowning the dome recalled, and may derive from, those over Soane’s offices at the Bank of England. Characteristically, Latrobe at this very same time was also building a country house, Sedgley, outside Philadelphia, with ‘Gothick’ detailing. By 1803 he had taken charge of the construction of the Capitol, nominally under Thornton, with whom he had continual rows. Most of the early interiors there were his, notably those in the south wing, fine examples of Romantic Classicism with French as well as English overtones; moreover he was still in charge of rebuilding them after the burning of the Capitol in 1814 down to his forced resignation in 1817.

In 1805 Latrobe submitted alternative designs for the Catholic cathedral in Baltimore. The Gothic design is one of the finest projects of the ‘Sublime’ or ‘High Romantic’ stage of the Gothic Revival; yet in its vast bare walls, carefully ordered geometry, and dry detail it is also consonant with some of the basic ideals of Romantic Classicism. The Classical design that was preferred and eventually built is perhaps less original; but internally, at least, this is one of the finest ecclesiastical monuments of Romantic Classicism, combining a rather Panthéon-like plan with segmental vaults of somewhat Soanic character (Plate 5). The cathedral was largely completed by 1818. The portico, though intended from the first, was added only in 1863, but the present bulbous terminations of the western towers are not of Latrobe’s design.

Near by in Baltimore the Unitarian Church of 1807 is by a Frenchman, Maximilien Godefroy (_c._ 1760-1833),[27] who was also responsible for the first Neo-Gothic ecclesiastical structure of any consequence in North America, the chapel of St Mary’s Seminary there, also of 1807. The Unitarian Church is a monument which might well have risen in the Paris of the 1790s had the French Deists been addicted to building churches. The triple arch in the plain stuccoed front below the pediment comes straight from Ledoux’s _barrières_; the interior, unhappily remodelled in 1916, was originally a dome on pendentives of the purest geometrical order. So also Godefroy’s Battle Monument of 1814 also in Baltimore, with its Egyptian base, might easily have been erected in Paris to honour some general prominent in Napoleon’s campaign on the Nile.[28] Another Frenchman, J.-J. Ramée (1764-1842), active since the Revolution in Hamburg and in Denmark, also came briefly to America. In 1813 he laid out Union College[29] in Schenectady, N.Y., on a rather Ledolcian plan and began its construction before he returned to Europe. His semicircle of buildings still crowns the hill—although two only are original—and Ramée here initiated a tradition of college architecture as remote from that of earlier American colleges, with their free-standing buildings set around a ‘campus’, as Wilkins’s Downing at Cambridge was from earlier English colleges.

The French eventually departed leaving no line of descent; but Latrobe had a pupil, the first professionally trained American in the field and, like Latrobe, almost as much an engineer as an architect. By 1808 Robert Mills (1781-1855) was supervising for Latrobe the new Bank of Philadelphia, Gothic (or at least ‘Gothick’) where his earlier Bank of Pennsylvania had been Grecian, and also building on his own the Sansom Street Baptist Church, a competent but not distinguished essay in Romantic Classicism. In the same year another Latrobe pupil, William Strickland (1788-1854), designed for Philadelphia a Gothick Masonic Hall; this was built in 1809-11, and later rebuilt, but according to the original design, after a fire in 1819-20.

Far more successful than either of these, if now overshadowed by the megalomaniac Classicism of the twentieth-century Philadelphia Museum of Art by Horace Trumbauer and others on the hill above, are the waterworks begun in 1811 on the banks of the Schuylkill. These are probably but not certainly by Mills rather than by the engineer Frederick Graff, whose name is signed to the drawings. These very utilitarian structures are most characteristic of the beginnings of Romantic Classicism in America, where Latrobe, Mills, and also Strickland were all three engineers as well as architects. Moreover, it is evident that engineering considerations often influenced their approach to architecture, just as architectural considerations gave visual distinction to much of their engineering. Thus they may be compared with engineers like Telford and Rennie in England as well as with the English architects of their day.

In this so-called ‘Federal’ period, when Romantic Classicism centred in the Middle Atlantic states thanks to Latrobe, Godefroy, Mills, and Strickland, the leading architect outside this area, the Bostonian Charles Bulfinch (1763-1844), was a late-comer to Romantic Classicism. His great public monument of the 1790s, the Massachusetts State House in Boston, had been designed originally as early as 1787-8, and even as executed in 1795-8 it derived principally from the Somerset House in London of Sir William Chambers (1726-96) and in one interior from Wyatt. His Boston Court House of 1810 first showed evidence of a change in his style, notably in its smooth ashlar walls of cold grey granite. That was a local material destined to lend particular distinction to the principal Romantic Classical buildings of Boston from this time forward (see Chapter 5).

The Frenchmen who came to America at the end of the eighteenth century or in the early 1800s (and shortly left again) could hardly import the French architecture of those decades; on the one hand, they had all been trained before the Revolution, from which most of them were in flight; on the other hand—and more consequently—there was almost no later architecture for them to reflect. Between 1789 and 1806 French building was at a standstill. Architects were mostly busy, if at all, with the decoration of various revolutionary fêtes and the accommodation of new political agencies in old structures.

One major example of the accommodation of an older structure to a new purpose deserves particular mention. In the years 1795-7 J.-P. de Gisors (1755-1828), E.-C. Leconte (1762-1818), and the former’s brother A.-J.-B.-G. de Gisors (1762-1835) built within the old Palais Bourbon the Salle des Cinq Cents, the legislative chamber of the First Republic. This hemicycle, at least as rebuilt along much the original lines by Joly in 1828-33, still serves as the Chamber of Deputies of the Fourth Republic. Such a chamber, so different in plan from the college-chapel arrangement of the British House of Commons with facing benches for Government and Opposition, is characteristically Romantic Classical in form, but this form has unfortunately proved to be conducive to an indefinite shading of multiple parties from right to left. The British model, suited to two-party rule only, was rarely imitated; the French one has been rather frequently, beginning with Latrobe’s House of Representatives in the Washington Capitol. Leaving aside the apparent political effect of the plan—not so notable in Washington as elsewhere—Gisors’s chamber seems to have been respectable if not especially distinguished. Covered with a segmental half dome and a barrel vault, both top-lighted, the smooth though rather richly decorated surfaces of the walls and the vaults made clear the interesting geometrical form of the interior space. The prototype was the lecture theatre of the École de Médecine in Paris erected in 1769-76 by Jacques Gondoin (1737-1818), one of the most advanced interiors of its day.

There was some private building in the Paris of the 1790s and early 1800s before public building eventually revived at Napoleon’s fiat. Typical and partly extant is the Rue des Colonnes, most probably by N.-A.-J. Vestier (1765-1816), although sometimes attributed to Poyet, who may have had some urbanistic control. This has an open arcade at the base carried on Greek Doric columns, here very modestly scaled, and cold flat walls above that are almost without any detailing whatever. This Paris street, as much as the arcaded ones of medieval and Renaissance Italy, may well have been the prototype for Napoleon’s first and greatest urbanistic project, the work of his favourite architects Charles Percier (1764-1838) and P.-F.-L. Fontaine (1762-1853). From his acquisition of La Malmaison in 1799 he kept them busy remodelling the interiors of his successive residences as First Consul and Emperor but rarely gave them new buildings to erect. This extensive planning scheme includes the Rue de Castiglione, running south out of the Place Vendôme, the Rue and Place des Pyramides, and the Rue de Rivoli facing the Tuileries Gardens. This last street was eventually extended to the east well beyond the Louvre by Napoleon III. The opening of the Rue de Castiglione was ordered in 1801; construction began the next year, and the execution of the rest went on, with long interruptions, for more than half a century.

Percier and Fontaine’s façades are characteristic of Romantic Classicism in their coldness of detailing and their infinite repetition of the same formula; but their Italianism, thin and dry though it is, recalls the plates in _Maisons et palais de Rome moderne_, which the two architects had published in 1798 before their professional star had risen very high (Plate 6B). With Nash’s Cronkhill, although in a very different and even opposed spirit, this scheme presages the international Renaissance Revival of the second quarter of the century. The very effective high curved roofs, filling out completely the ‘envelope’ allowed by the Paris building code, were added in 1855; more conventional two-pitched mansards were provided originally.

But the Empire mode, particularly as elaborated by Percier and Fontaine in the service of the Emperor, was primarily a fashionable style for interiors, and found perhaps its most characteristic expression in furniture, usually of dark mahogany with much ornate decoration of a character resembling gold embroidery on uniforms. Such flat decorative work is also found carved on exteriors, not only in France but wherever Napoleonic influence penetrated. Indeed in furniture and interior design generally non-French work is often of the highest quality, especially when executed for such clients as Napoleon’s sister Caroline Murat at Naples.

Yet the character of French leadership in the arts had changed since the 1780s. The architects at the end of the _ancien régime_ had been truly revolutionary in their aesthetic and their social ideals. Napoleon’s designers, almost like Hitler’s and Mussolini’s and Stalin’s in our century, were flatterers and time-servers. Emulation of their work abroad was chiefly a matter of following well-publicized fashion; creative French influence still flowed, however, from men of the older generation now so largely forgotten at home. Thus it was at this point that Ledoux’s projects became generally available to others, thanks to his book published in 1804 and dedicated to Napoleon’s Russian ally of the moment, Alexander I.

Extensive building activity in Paris under Napoleon’s aegis began only in 1806, but once it started there came a positive flood of projects in conscious emulation of Louis XIV’s architectural campaigns. There was also the expectation that this activity would absorb unemployment in the building trades. But Napoleon, like later dictators who have initiated vast building projects, actually bit off a great deal more than he could chew. He was, however, more fortunate than Mussolini and Hitler in that the regimes which succeeded his in the decades between the First Empire and the Second were surprisingly willing to carry his unfinished monuments to completion. Still later, his nephew Napoleon III emulated him in an even more concerted programme of urbanism and monumental construction carried out over nearly two decades in a very different style—indeed in several (see Chapter 8).

The Colonne de la Grande Armée, replacing the statue of Louis XV at the centre of the Place Vendôme, is a properly symbolic monument of its epoch—first to be designed of the many giant columns that would arise all across the Western world from Baltimore to Petersburg within the next quarter century. Wilkins’s Nelson Pillar in Dublin, actually completed before the Paris example, has already been mentioned. The column in Paris is Trajanesque not Grecian, however, and was entirely executed with the bronze of captured guns. It well represents the Imperial Roman megalomania already evident in many projected memorials of the 1790s. Gondoin, its architect, with whom was associated J.-B. Lepère (1761-1844), provides a real link with the past, since his already-mentioned École de Médecine was one of the earliest major edifices in which Romantic Classical ideals were carried beyond the transitional stage of Soufflot’s Panthéon.

Even before the Colonne Vendôme was finished in 1810, a smaller and somewhat less typical monument, but equally Roman and also the first of a considerable line, had been completed by Percier and Fontaine. The Arc du Carrousel of 1806-8—once a gate to the Tuileries from the Place du Carrousel, now unhappily floating in unconfined space—has much of the daintiness and, in the use of coloured marbles, the polychromy of its architects’ contemporary palace interiors. Indeed, the richness of the detailing is far less characteristic of Empire taste in architecture than are their façades near by in the Rue de Rivoli (Plate 6B); the Arc du Carrousel must have provided a rather fussy pedestal for the superb Grecian horses stolen from St Mark’s in Venice that were originally mounted upon it.

Far more satisfactorily symbolic of imperial aspiration is the enormous Arc de Triomphe de l’Étoile, which looks down the entire length of the Champs Élysées today to overwhelm its brother arch even at that great distance (Plate 7). J.-A. Raymond (1742-1811), a pupil of Leroy, first received the commission; but with him was associated J.-F.-T. Chalgrin (1739-1811), the master of the younger Gisors, who soon took over and imposed his own astylar design. Chalgrin, like Gondoin, was an architect already well established under the _ancien régime_. His major innovation had been the reintroduction of the basilican plan[30] at Saint-Philippe-du-Roule in Paris in the 1760s, henceforth one of the favourite models for Romantic Classical churches in France and elsewhere on the Continent. Like many of the monuments of that earlier period by Chalgrin’s contemporaries, his Arc de l’Étoile reverts less to Roman antiquity than to certain aspects of the architecture of Louis XIV. Even its megalomaniac grandeur can be matched, relatively at least, in the Porte St Denis in Paris built in the 1680s by François Blondel, and it follows almost line for line the square proportions of that masterpiece. The arch was slowly brought to completion after Chalgrin’s death, first by his pupil L. Goust from 1811 to 1813 and from 1823 to 1830; then by Goust’s assistant, J.-N. Huyot (1780-1840), advised by a commission that included François Debret (1777-1850), Fontaine, and the younger Gisors; and finally from 1832 to 1837 by G.-A. Blouet (1795-1853). It owes its unmistakably nineteenth-century character partly to the crisp, hard quality of its imposts and entablatures and partly to the great Romantic figural reliefs executed in 1833 by Rude, Etex, and Cortot. These take the place on the piers of the more conventional trophy-hung obelisks on Blondel’s seventeenth-century arch. A certain post-Empire quality derives from the plastic complexity of Blouet’s attic; but on the whole the Arc de l’Étoile, if less original and less influential than Saint-Philippe-du-Roule, is Chalgrin’s masterpiece and Napoleon’s finest memorial.

The Place de la Concorde, projected by A.-J. Gabriel (1692-1782) at the end of the Baroque Age, continued to lack, even after a half century and more, appropriate monuments to terminate the cross axis. The building of a big church at the head of the Rue Royale to close the vista between Gabriel’s two colonnaded ranges on the north side of the square had bogged down well before the Revolution; across the river the much earlier Palais Bourbon, set at an angle, was even more awkward than before, now that the roof of the Salle des Cinq Cents rose above it. Since the amelioration of this southern terminal required only a tall masking façade set at right angles to the axis, this was promptly provided. Poyet in 1806-8 used the most obvious Romantic Classical solution for such a problem, a high blank wall with a ten-columned temple portico at its centre. The result is certainly an urbanistic success, if without any particular intrinsic interest; the raising of the portico above a high range of steps ensured, for example, its visibility from the square across the bridge. The form of the pediment was slightly modified and the sculpture by Cortot added in 1837-41.

In 1761 Pierre Contant d’Ivry (1698-1777) and, after his death, G.-M. Couture (1732-99) had made successive projects for a church dedicated to the Magdalen at the head of the Rue Royale, the latter already proposing that it be surrounded by a Classical peristyle. This structure, which was as yet barely begun, Napoleon now decided should be not a church but a Temple de la Gloire—he reversed his decision in 1813 after the Battle of Leipzig and the loss of Spain. For such a temple he understandably preferred, in the competition held in 1806, neither the first nor the second premiated design, both of church-like character, but one by Pierre Vignon (1763-1828) that proposed the erection of an enormous Corinthian temple on a high Roman podium. Inside, a series of square bays covered with domes on pendentives supported by giant Corinthian columns provided a structural solution technically Byzantine but as imperially Roman in scale and detailing as the exterior.

Construction of the Madeleine, begun in 1807, dragged on interminably. J.-J.-M. Huvé (1783-1852) succeeded Vignon as architect in 1828 and, like the Arc de l’Étoile, the edifice was finally finished only under Louis Philippe in 1845. The interior has a somewhat funereal solemnity, more characteristic of the post-Napoleonic regimes than of the period of its initiation. The rather obvious temple form of the exterior is redeemed by the superb siting, the really grand scale, and the rich pedimental sculpture by Lemaire. Like Chalgrin’s arch, Vignon’s Madeleine has continued to provide a major monumental nexus in the urbanism of Paris ever since.

Also proposed in 1806 but not initiated until 1808 was the Bourse by A.-T. Brongniart (1739-1813), another architect who had, like Gondoin and Chalgrin, made his mark long before the Revolution (Plate 8B). Again a free-standing peripteral structure like the Madeleine, the Bourse has suffered somewhat from its enlargement in 1902-3 by J.-B.-F. Cavel (_c._ 1844-1905) and H.-T.-E. Eustache (1861-?). Nearly square originally and unpedimented—and also set much closer to the ground—it must always have lacked the monumental presence of the Madeleine. But the interior with its ranges of arcades, derived almost as directly from a Louis XIV monument—in this case the court of the Invalides by Libéral Bruant—as Chalgrin’s arch was from that of Blondel, is very characteristic of the sort of reiterative composition generally favoured by Romantic Classicism. L.-H. Lebas (1782-1867) was associated with the elderly Brongniart from the start, and after Brongniart’s death the building was finished in 1815 by E.-E. de Labarre (1764-1833). Labarre was responsible also for the Colonne de la Grande Armée at Boulogne; this was proposed in 1804 and begun in 1810, but, like so many Napoleonic monuments, not finished until Louis Philippe took up its construction again in 1833. It was finally completed by Marquise in 1844.

In 1799 a fire made it necessary to rebuild the Théâtre de l’Odéon; but the original design of M.-J. Peyre (1730-88) and Charles de Wailly (1729-98), dating back to 1779, was repeated in 1807 with little change, as was also the case in 1819 when it was rebuilt again after another fire. This provides excellent evidence of the continuity of Romantic Classical style in France before and after the Revolution (see Chapter 3).

Napoleon had in mind the erection of various less monumental and more utilitarian structures than the Bourse and the Odéon; some of these were started, and one or two even finished, before the Empire came to an end. Behind one section of the façades in the Rue de Rivoli an enormous and rather dull General Post Office was begun in 1810 and eventually completed to serve as the Ministry of Finance under Charles X in 1827. Another ministry (Foreign Affairs) on the Quai d’Orsay was designed in 1810 by J.-C. Bonnard (1765-1818) and even begun in 1814; this was eventually carried to completion by Bonnard’s pupil Jacques Lacornée (1779-1856) in 1821-35. With its rich ordonnance of columns and arches, Bonnard’s façade had an almost High Renaissance air, or so it would appear from extant views of a structure long ago destroyed.

The Marché St Martin of 1811-16 by A.-M. Peyre (1770-1843), the Marché des Carmes of 1813 by A.-L.-T. Vaudoyer (1756-1846), and the Marché St Germain of 1816-25 by J.-B. Blondel (1764-1825), with their clerestory lighting and open timber roofs, are typical of the more practical side of Romantic Classicism.[31] The simple masonry vocabulary of these Parisian markets, so straightforward and without Antique pretension, was considered to be Italian (see Chapter 2).

The Napoleonic building flurry barely reached the provinces before its short course was over. The theatre in Dijon, begun about 1805 by Jacques Célérier (1742-1814), may be mentioned; but such plain square blocks with frontal porticoes could have been, and were, built in almost precisely the same form thirty years before—for example Ledoux’s theatre at Besançon of 1775-84. At Pontivy in Brittany, then called Napoléonville, the younger Gisors built a Préfecture in 1809 and a Palace of Justice with associated prisons two years later. A rather dull church, Saint-Vincent at Mâcon, repeating a model that had been new at Saint-Philippe-du-Roule forty years earlier, was also erected by him in 1810. The pair of front towers was a novelty suggested by an earlier project of Lebas.

It is quite characteristic of this period, so ready (as the French have been ever since) to employ elderly architects and so content with stylistic innovations that dated from before the Revolution, that Mathurin Crucy (1749-1826) rebuilt in 1808-12 the theatre in Nantes—very like that at Dijon—in exactly the same form as it had originally been designed by him in 1784-8; while he also finished in 1809-12 the Bourse and Tribunal de Commerce there which he had begun in 1791, just after the Revolution started, with no change in the original design. The setting of his theatre in the Place Graslin provided by continuous ranges of five-storey houses is presumably contemporary; despite the rather high roofs, the façades are notably crisp and smooth. The rusticated arcuation of the lower storeys might make plausible a date in the 1780s, but the rather thin and geometrically detailed iron balcony railings suggest rather the first or second decade of the new century, when the theatre was rebuilt.

If the imperial effort in France barely extended outside Paris except for the interior alterations that Percier and Fontaine carried out in the royal châteaux at Versailles, Compiègne, Saint-Cloud, and Fontainebleau—major examples of Empire decoration but not of architecture—the emperor and his nominees left their mark on most of the great cities of continental Europe. The Palazzo Serbelloni in the Corso Venezia, where Napoleon stayed in Milan, had been built by Simone Cantoni (1736-1818) in 1794. Similar to French work of the 1780s, it would probably have impressed the Emperor as still quite up-to-date. He ordered in 1806 the laying out in Milan of the Forum Bonaparte, according to the designs of Giannantonio Antolini (1754-1842), and the erection of a conventionally Roman triumphal arch, the work of Luigi Cagnola (1762-1832?), which was finally completed in 1838.

In Rome the development of the Piazza del Popolo, like the Forum Bonaparte a work of urbanism rather than of architecture, was based by Giuseppe Valadier (1762-1859), an Italian despite his French name and ancestry, on a project he had made as early as 1794. This project was modified by him under the Empire to incorporate ‘corrections’ by the younger Gisors and L.-M. Berthault (1771?-1823). Execution of the project actually began only in 1813 after Pope Pius VII returned from his Napoleonic captivity; Valadier carried it forward to ultimate completion in 1831. Valadier’s Roman church work, such as his new façade for San Pantaleone of 1806, just off the present-day Corso Vittorio Emanuele, is mostly too dull to mention; his domestic work was somewhat more interesting, but with little personal or even Italian flavour.

In Naples Leconte, who had worked with the two Gisors on the Salle des Cinq Cents in Paris, remodelled the San Carlo opera house in 1809 for Murat—it was, however, refronted in 1810-12 and rebuilt in 1816-17 (see