PART TWO

185O-19OO

## CHAPTER 8

SECOND EMPIRE PARIS, UNITED ITALY, AND IMPERIAL-AND-ROYAL VIENNA

MANY historians, in despair, have merely labelled the period after 1850 ‘Eclectic’ as if earlier periods of architecture—and notably all the preceding hundred years since 1750—had not also been eclectic, although admittedly to a lesser degree. Within the eclecticism of the late eighteenth and early nineteenth centuries there can readily be distinguished the two major stylistic divisions with which Part I has dealt separately (in Chapters 1-5 and in Chapter 6, respectively). So also in the fifties, sixties, and seventies two principal camps are discernible among the architects. Their programmes were less clear than in the previous half century, and in one case much less widely accepted internationally. Yet the High Victorian Gothic of England, taken together with the later Neo-Gothic elsewhere, on the one hand, and what may be loosely called the international Second Empire mode on the other, subsume between them a fair part of the more conspicuous architectural production of the third quarter of the century.

Both the Victorian Gothic of this period and the Second Empire mode were ‘high’ phases of style. Perhaps for that reason neither of them controlled, in the way that Romantic Classicism had done in the earlier decades of the century, all or even any very extensive segments of building activity; yet between them they gave colour to a very considerable proportion of it. The obvious stigmata of one or of the other, or even of both—external polychromy and high mansard roofs, respectively—are to be found on such modest things as mills and working-class housing blocks as well as on major public monuments. The High Victorian Gothic first developed in Anglican ecclesiastical architecture and always carried with it a rather churchy flavour—sometimes quite ludicrously, as in the case of Gothic distilleries, Gothic public-houses, and Gothic sewage plants. Continental Neo-Gothic was more largely confined to churches, especially in France. The international Second Empire mode found its inspiration in the grandiose extension of a palace in Paris; something of the Parisian and even the palatial clung to it when it was used—as often in the non-French world—for such things as factories and modest suburban villas.

Both the Victorian Gothic and the Second Empire had definite national homes, yet both were also full of elements of Italian origin. In that respect the High Victorian phase of the fifties and sixties was somewhat analogous to the Germanic _Rundbogenstil_, as well as being the direct heir of the earlier and more puristic Gothic Revival of the forties in England. Often the Second Empire mode was even more Italianate, since it was in the main but a pompous modulation of the earlier Renaissance Revival. The one had its roots in the Picturesque, but it differed from earlier Picturesque manifestations in being a ‘style’—or very nearly such—not merely the reflection of a point of view. The other had roots not only in Romantic Classicism but also farther back in the High Renaissance and the Baroque; some qualities of those earlier styles were both continued and revived. But neither High Victorian Gothic nor Second Empire were ‘revivals’ in the sense of those of the first half of the century; they lived with a vigorous nineteenth-century life of their own, not one borrowed from the past. In both cases one may more properly say that they _had_ revived.

The Second Empire mode was the heir, or at least the successor, of the last universal style of the western world, the Romantic Classical. Moreover its wide international sway was hardly terminated by the end of Napoleon III’s reign in France any more than its beginning had waited for his enthronement. Concerning that sway it should be noted, however, that considered as a definite ‘style’ the Second Empire mode is very far from characterizing as much of French production in this age as of that in several other countries. Indeed, somewhat paradoxically, its actual initiation may almost be said to have occurred outside France and before the political Second Empire actually began in 1852. In this chapter and the next, certain alternative developments in succession to the earlier Renaissance Revival have been associated with the Second Empire mode, sometimes a bit arbitrarily perhaps, for lack of a more appropriate place to deal with them.

Although France was less affected by the Picturesque in the first half of the nineteenth century than England, the Renaissance Revival had permitted some straying from the more rigid paths of Romantic Classicism in the thirties and forties (see Chapters 3 and 6). The earliest French work of the twenties that may seem of Italian Renaissance inspiration is very severe and flat, approximating occasionally the effects of the German _Rundbogenstil_ yet consistently disdaining that mode’s tendencies towards either medievalism or originality in detail. Gradually, under Louis Philippe, there were changes: on the one hand, there arose an interest in later periods of the Italian Renaissance; on the other, there came an increasing and less peripheral use of sixteenth-century and even later native models. Common to both these developments was an evident desire for richer and more plastic effects.[178] What above all distinguishes the mature Second Empire mode, even more in other countries than in France, is the elaboration of three-dimensional composition by the employment of visible mansard roofs and of pavilions at the ends and centres of buildings, these last capped either with especially tall straight-sided mansards or, even more characteristically, with convex or concave ones. Such features are rare before 1850 in France and almost unknown elsewhere.[179]

The return of the mansard in France is harder to document than its appearance as a new element of architectural composition in other countries, for in France it had never passed out of use as a practical device for providing usable attics. With the increasing emulation of sixteenth-century French models in the second quarter of the century tall roofs of a more medieval sort began to be used with some frequency. Biet’s ‘Maison de François I’ of 1825 did not have them; but ten years later they are very prominent on the _François I_ house Dusillion built in the Rue Vaneau. Moreover, Lesueur in the late thirties could hardly avoid their use when extending the sixteenth-century Hôtel de Ville (Plate 22A). As noted earlier, it seems to have been H.-A.-G. de Gisors, at the École Normale Supérieure built in 1841-7, who first re-introduced on a prominent building mansards of seventeenth- or early eighteenth-century character, and in association with detailing that suggests, vaguely at least, the _style Louis XIV_. By the late forties the use of such mansards was fairly common in France, although they rarely received much emphasis.

Had Dusillion in 1849-51 built the mansarded mansion for T. H. Hope[180] in Paris rather than in London therefore, or the Danish-born but Paris-schooled Detlef Lienau (1818-87)[181] his mansarded Hart M. Shiff house of the same date in France rather than in America, neither would have been especially notable. But in the England and the United States of the mid century emulation of French models was in itself novel. Dusillion’s and Lienau’s mansards, moderate enough by French standards, suggested to the English and the Americans a way by which edifices of generically Renaissance character could be given something of the bold silhouette that high pointed roofs provided for Victorian Gothic structures. Like Barry’s loggia-topped towers and his corner chimneys, mansards appealed directly to the mid century’s characteristic desire to break sharply away from the flat-surfaced, and nearly flat-topped, cubic blocks of Romantic Classicism. Pavilion composition offered a similar resource for the plastic modelling of façades.

In 1851, following immediately after the Hope house, came the designing of the Great Western Hotel at Paddington in London by the Hardwicks. This was still, one should note, before the Second Empire actually began in France. Gawky though this hotel is, and very uncertain in its use of French precedent, contemporaries generally recognized its inspiration as derived from the period of Louis XIV. The complex massing and the broken skyline, with roofs of different heights and pavilion-like towers at the ends, are much more obviously a premonition of the Second Empire mode in the form the world outside France would shortly adopt it than were the London and New York houses of two years earlier. Unlike Dusillion and Lienau, moreover, the architects of the Great Western Hotel, recognized masters of the dying Greek Revival as well as of the rising Gothic and Renaissance Revivals, were not French-trained.

If the international Second Empire mode had thus, in a sense, beginnings outside France, it is nevertheless true that its spiritual headquarters was in Paris. The prestige of the new Emperor’s capital, a prestige rapidly regained after more than a generation of desuetude, quite as much as the visual appeal of multiple mansards and pavilioned façades, explains the world-wide success of the mode during, and even well after, the eighteen years that the Second Empire lasted.

It was in 1852 that Napoleon, then Prince-President, made himself Emperor. He had already signalized, a few months earlier, his ambition to revive the splendours not alone of his uncle’s rule but those of earlier French monarchs by his decision to complete the Louvre[182]—or more accurately to connect the Louvre with the Tuileries. This was a project over which generations of architects had struggled on paper and at which several abortive starts had already been made. Visconti received the commission, not Duban, who had been engaged since 1848 on what was proving a highly controversial restoration of the old Louvre. Visconti was chosen not for his reputation as a private architect but largely because a succession of public projects for new library buildings in Paris that he had been asked to prepare under Louis Philippe and even under the Second Republic had all fallen through, and it was felt he deserved an important commission from the State. Perhaps also his Tomb of Napoleon I at the Invalides made him especially sympathetic to Napoleon III.

A viable scheme for the New Louvre was produced by the sixty-year-old Visconti with very great rapidity. Counting on the great size of the Cour du Carrousel to obscure the awkward lack of parallelism between the Louvre and the Tuileries, he planned two hollow blocks extending westward at either end of the existing western front of the old Louvre. Beyond these blocks narrower wings, in part built already, would connect with the two ends of the Tuileries Palace in which French rulers usually lived. In the middle of the court fronts of the side blocks there were to be large pavilions, echoing Le Mercier’s in the centre of the west wing of the old Louvre, and other smaller pavilions to mark the salient corners towards the Place du Carrousel. Although the new constructions were intended to house various things—two ministries, a library, stables for the Tuileries, etc.—they were designed comprehensively with no specific indication of what would go on behind the long walls and inside the various pavilions. The New Louvre was not a palace or Royal residence; but like the old Louvre, which by this time housed several disparate activities—most notably the chief art gallery of France—it was meant to be representationally palatial.

In 1853 Visconti died and H.-M. Lefuel (1810-80), a pupil of Huyot, took over. Lefuel very much enriched the design and thereby provided the prime Parisian exemplar of the Second Empire mode, at least as the world outside France came to know it in the late fifties and sixties. Heavily though Lefuel leaned on the precedents provided by the various sections of the old Louvre, it is important to stress that his design did not represent, in the way of the first half of the century, a specific ‘revival’. For one thing, the old Louvre, begun by Pierre Lescot late in François I’s reign and carried forward by a succession of architects in the next four hundred years, offered a wide range of suggestions but no one consistent model. The most characteristic and striking features of the New Louvre, the corner pavilions, were those that were most eclectic in inspiration and in their total effect most nearly original (Plate 68). No part of the old Louvre is as boldly plastic as these pavilions with their rich applied orders set far forward of the wall-plane; only Le Mercier’s Pavillon de l’Horloge on the old Louvre offered precedent for the great height of all the new pavilion roofs and in particular for the convex mansards, like square domes, over the central pavilions flanking the Cour du Carrousel.

Sumptuous as was Goujon’s sculptural investiture of the earliest work in the court of the old Louvre, this was delicate in scale and very flat; much of the sculptural decoration of the new pavilions follows Goujon fairly closely, but even more—some of it nearly in the round—is so bombastically plastic as almost to justify the term ‘Neo-Baroque’. Although there is actual early-seventeenth-century precedent for most of their individual details, the very lush stone dormers set against the high straight mansards of the corner pavilions are particularly novel in effect. For the next thirty years, and even longer, such features of the New Louvre would be imitated all over the western world yet, paradoxically, they had much less influence in France and almost none in Paris.

As far as the outside world—particularly perhaps England and the United States, but hardly less Latin America—was concerned the New Louvre was the prime architectural glory of Second Empire Paris and the symbol, _par excellence_, of cosmopolitan modernity. Burghers in Amsterdam and Montreal, vacationers in Yorkshire and silver-miners in the Rocky Mountains all expected to find echoes of it in the sumptuous new hotels they frequented; Latin Americans continued to emulate it even into the twentieth century. Yet in the real Paris of the Second Empire, the Paris which is largely still extant today, the New Louvre is but one prominent structure among many and, as has been said, not even a very typical one.

The first Napoleon had had no time to carry out any considerable urbanistic reorganization of his French capital. But for the goodwill of his successors, notably Louis Philippe, the architectural projects that he was able to initiate would never have been brought to completion. His nephew, however, vowed to peace and not to war, had nearly two decades in which to build. Well before his reign began, moreover, he had definitely made up his mind to replan Paris more drastically than any great city had ever been replanned before.[183] Only a few fine squares, the Champs Élysées, and the Rue de Rivoli remain in Paris from earlier campaigns of urban extension and replanning; but the Paris of the Second Empire, the Paris of the boulevards and the great avenues, is the urbanistic masterwork of the third quarter of the nineteenth century, a period notably deficient in urbanistic achievement almost everywhere else except in Vienna.

For all the sumptuousness of the individual monuments with which the focal points of Napoleon III’s Paris were ornamented, their settings are generally more distinguished than the ‘jewels’ mounted in them; an exception, of course, is the Place de l’Étoile where, however, the jewel was inherited from an earlier period (Plate 7). This is because of the high standard of design that was maintained in the general run of new blocks of flats that lined the _places_, the boulevards, and the avenues (Plate 75A). Since in Second Empire Paris the urban totality is more significant than the individual buildings, and since over the years of the Empire—or for that matter down even to the eighties—there was very little stylistic development, the Parisian production of this period may well be presented more topographically than chronologically, as if one were outlining a tour[184] of its splendours.

There is one extant railway station of some distinction belonging to the period at which to arrive. Yet this station, Hittorff’s Gare du Nord designed in 1861 and built in 1862-5, is perhaps less advanced than Duquesney’s Gare de l’Est, which was just being completed as the Second Empire opened (Plate 22B). The flat Ionic pilasters of the façade and the great archivolt-surrounded openings between them are evidence of the firm resistance that Hittorff’s generation put up against the lusher tastes of the mid century as expressed in Lefuel’s work on the New Louvre. Even more characteristically Romantic Classical, and probably finer though less famous than the Gare du Nord, was Cendrier’s Gare de Lyon, since demolished, which had been built almost a decade earlier at the same time as his Palais de l’Industrie in the early fifties.

Proceeding from Hittorff’s station one strikes immediately the characteristic broad straight streets, often lined with trees, that were the new Second Empire arteries of Paris. The continuous ranges of grey stone buildings, their even skyline crowned with inconspicuous mansards, generally include shops below and always contain flats above. They are so designed as to attract very little attention to the individual structures,[185] almost as little as do the separate houses in London terraces. There is much less irregularity of outline than along Nash’s Regent Street, for example, and a general consistency in the size and phrasing of the windows. There is also very little noticeable variety in the handling of the conventional apparatus of academic detail so crisply carved in fine limestone. Even where, by great exception, some bolder architect such as Viollet-le-Duc used more original detail, the unity of character is barely disturbed, so consistent are the basic patterns of the façades (Plate 101A).

Since the plan of Paris has remained basically radial, the visitor has the choice of proceeding circumferentially along one of the lines of outer or inner boulevards or of turning inwards to the centre. It is more profitable, on the whole, to advance centripetally, for the outer boulevards are generally very monotonous. The Île de la Cité was the original core of Paris; the east-and-west axis of the Louvre, extended westward along the Champs Élysées all the way to the Étoile, already provided a central tract parallel to the Seine; the new cross axis was to be a north and south artery running from the Gare de l’Est to the Observatoire. On the Île the vast complex of the Palais de Justice, whose restoration and extension had been undertaken by Duc as early as 1840, received a notable Second Empire ornament in its western block, facing the Place d’Harcourt, which was built by Duc assisted by E.-T. Dommey (1801-72) in 1857-68. Rationalistic in its structural expression and Classical in most of its detailing, this façade and the hall behind it reflect the tastes of the period in the heavy scale of the parts and the rather cranky—and certainly studied—awkwardness of the modelling of the various conventional elements of the orders and minor features of detail. Duc’s earlier work at the Palais de Justice, on the other hand, was detailed with very great grace and elegance, it may be noted.

The principal Second Empire construction on the east-and-west axis of Paris, the New Louvre, has been described already. Along the north side of the Louvre the Rue de Rivoli was extended eastward in 1851-5 the entire length of the palace with no change in the original Percier and Fontaine design except for the addition of high quadrantal mansards throughout the entire length of the street and its subsidiaries. Even a large new hotel[186] was forced into this framework. Yet because of its island site, the high rounded roofs give this block as it is usually seen from the Place du Théâtre Français to the north something of the new plasticity; it thus provided eventually an appropriate terminus to the Avenue de l’Opéra, after that was finally completed under the Third Republic.

Facing the east side of the Louvre, Hittorff balanced the restored Gothic front of Saint-Germain-l’Auxerrois with the new front of the Mairie du Louvre built in 1857-1861. Characteristic of this period in France is the avoidance of Gothic detail on this secular façade in favour of something vaguely _François I_; yet the pattern of the front of the church is carefully repeated, even to the rose-window in the high-pitched gable, and the new tower by Ballu, on axis between the Church and the Mairie, is Gothic.

Up to the Rond Point, the Champs Élysées is flanked by parked areas on either side and decorated by fountains and other features designed by Hittorff (see Chapter 3). At the Rond Point there are a few very sumptuous _hôtels particuliers_, but beyond that the avenue was built up—or more accurately, for the most part, would eventually be built up—like a very broad boulevard flanked by large blocks of flats with shops and cafés below. In the open area on the left between the main axis, the river, and the new quarter which had taken its name ‘François I’ from Biet’s house, lay the Jardin d’Hiver of 1847 and the Palais de l’Industrie of 1853-4. Here also is the Rotonde des Panoramas of 1857 by G.-J.-A. Davioud (1823-81). Around the Arc de l’Étoile, at the far end of the Champs Élysées, are ranged pairs of dignified houses; these were designed by Hittorff with the collaboration of Rohault de Fleury in 1855 and executed in 1857-8 in a mode so academic as to be almost a revival of the _style Louis XVI_ (Plate 7). The general layout of the _place_ was determined by Haussmann, expanding a much earlier scheme of Hittorff’s.

What is most notable in all this mid-nineteenth-century construction along the main axis of the city is the continuity of taste between the Second Empire period and the period that preceded it. The only real echo of the New Louvre was in the big private houses set back from the Rond Point.

The Avenue de l’Opéra, extending north-westward from the Place du Théâtre Français, has become, since its completion in 1878, the major cross axis, rather than the earlier Boulevard Sébastopol to the east. The Place de l’Opéra, with a short spur of the avenue at its south end, was laid out in 1858; and the façades of the buildings (Plate 70C) around it began to go up in 1860 from the designs of Rohault de Fleury[187] and Henri Blondel (1832-97). The Opéra[188] (more properly Académie Nationale de Musique)—after the New Louvre the most conspicuous product of the Second Empire—was begun in 1861 from the design with which J.-L.-C. Garnier (1825-98), a pupil of Lebas who also worked briefly for Viollet-le-Duc, won the second competition held in that year. Although the Garnier design is often thought to be particularly characteristic of the taste of the Imperial couple, it was actually very unpopular with the Empress Eugénie; she had expected the project of her friend Viollet-le-Duc to be accepted and was furious when it failed to win. Substantially completed externally by 1870, the Opéra was not finally finished and opened until January 1875, so that neither Napoleon III nor Eugénie ever entered it.

Here, at its heart, the contrast between setting and monument in Second Empire Paris is at its most extreme, even though this setting is far richer and more plastic than that provided by the severely flat houses that surround the Arc de l’Étoile. Just as there, however, the use of a giant order on all the big blocks that form the _place_ reveals the distinctly academic taste of the leading French architects in this period; but Blondel’s rounded pavilions, where two major streets come in on either side at an angle, provide an almost Baroque elaboration in the grouping of the various masses by which the complex space is defined (Plate 70C). Certainly the result is very different from the large open areas surrounded by discrete blocks of plain geometrical shape favoured by Romantic Classicism.

The Opéra is sumptuous in a rather different way from the New Louvre (Plate 70B). Yet in Garnier’s work, as in Lefuel’s, a generically Neo-Baroque effect is achieved with elements mostly High Renaissance in origin, but here Italian rather than French. The richly coloured marbles, the admirably placed sculpture by Carpeaux, and above all the fashion in which the masses pile up—from the ornate colonnade crowning the main façade, through the half-dome which expresses the auditorium externally, to the tall stage-house at the rear—is much richer plastically than the somewhat repetitive scheme of the New Louvre. The whole, moreover, is made fully three-dimensional by the comparable organization of the major elements at the sides and on the rear. Thus Garnier provided a visual equivalent to the complex ordering of his extremely elaborate plan, a plan the undoubted virtues of which can be fully appreciated only on paper (Figure 15). Inside the Opéra the great staircase, the foyer, and the actual auditorium drip with somewhat brassy gold and the profusion of detail has a curiously un-Renaissance spikiness and lumpiness (Plate 71). This quality underlines how un-archaeological was Garnier’s approach, how responsive he was (perhaps unconsciously) to the new tastes of the mid century that had produced the High Victorian[189] Gothic in England in the previous decade and fostered generally the international success of the Second Empire mode. When Eugénie asked him what the ‘style’ of the Opéra was—_Louis XIV_, _Louis XV_, _Louis XVI_—he replied with both tact and accuracy: ‘C’est du Napoléon III’.

Like the lushness of the New Louvre, Garnier’s lushness has an undeniably parvenu quality characteristic of the time and place; but the pace he set, however much emulated all over the world in later opera houses, and the peculiar capacity he showed for satisfying the taste for bombastic luxury of the third quarter of the century were never equalled by other architects, least of all by French ones. In the twin theatres flanking the Place du Châtelet,[190] which were built in 1860-2, Davioud, the architect of the Rotonde des Panoramas, made little attempt to vie with Garnier’s Opéra; but they are considerably more successful in their own right than is the Vaudeville in the Boulevard des Capucines of 1872 by A.-J. Magne (1816-85), which does. Garnier’s own Panorama Français of 1882 at 251 Rue Saint-Honoré has only a modest façade to the street.

Only one other work of Garnier himself rivals the Opéra, his Casino at Monte Carlo of 1878. The fine site that this occupies somewhat makes up for its tawdry finish in painted stucco, and the two-towered façade towards the bay has a properly festive air. The Casino and Baths he built at Vittel in 1882, his Observatory at Nice, and the Cercle de la Librairie of 1880 in the Boulevard Saint-Germain in Paris are considerably quieter in design. The Palais Longchamps[191] of 1862-9 in Marseilles by H.-J. Espérandieu (1829-74), who had worked for Questel and for Vaudoyer, two palatial museum blocks joined by a curved colonnade above an elaborate cascade, is more Neo-Baroque than most work of the period (Plate 70A); but much of the credit should go to the sculptor Bartholdi whose earlier fountain project Espérandieu took over.

[Illustration:

Figure 15. J.-L.-C. Garnier: Paris, Opéra, 1863-74, plan ]

Despite what has been said of the houses at the Rond Point, most Second Empire mansions in Paris, at least those built by leading architects, tend to be rather restrained in their general design and often quite archaeologically correct in their detailing. They are likely, moreover, to follow French seventeenth- or eighteenth-century models rather than those of sixteenth- or seventeenth-century Italy. Already, in the Hôtel de Pontalba, Visconti had copied Versailles closely in the interiors, while his exterior followed the line of the early eighteenth-century _hôtels particuliers_. (This was drastically remodelled in the eighties.) Labrouste, in the Hôtel Fould, 29-31 Rue de Berri, which was built in 1856-8, was rather plausibly Louis XIII; while Alfred Armand (1805-88), a pupil of Leclerc and a frequent collaborator with Pellechet, in designing the Hôtel Pereire and its twin in the Place Pereire about 1855 approached the _style Louis XVI_ as closely as Hittorff did round the Étoile. Nevertheless, study of Parisian exemplars inspired many foreign architects to design houses that could hardly be anything else but Second Empire.

This is largely explained by the special character of the publications[192] of C.-D. Daly (1811-93), a pupil of Duban, and of P.-V. Calliat (1801-81), a pupil of Vaudoyer, through which current French work of this period chiefly became known to the outside world. Almost as was the case at the opening of the century, when the volumes illustrating Prix de Rome projects made the higher aspirations of French architects better known to students abroad than their ordinary practice, the publications of this later day seem to have focused attention on certain aspects only of the French architectural scene, aspects prominent enough, but not altogether characteristic as regards public monuments and dominant official taste. Without knowledge of the French architectural past, without the inhibitions instilled early in French architects by their training at the École des Beaux-Arts, foreign architects readily derived from published sources a Second Empire mode considerably lusher than was generally approved for public use in French academic circles and made it very much their own. Even in public architecture foreigners must have seen current work with different eyes from the French.

For example, the Tribunal de Commerce on the Île de la Cité, an agency provided in 1858-64 with a building of its own instead of mere quarters in the Bourse, was supposed by French contemporaries to express in its detailing the Emperor’s personal enthusiasm for the _quattrocento_ buildings that he had lately seen in Brescia. But posterity, like foreigners when the Tribunal was new, notes in this work of A.-N. Bailly (1810-92) the characteristic Second Empire mansards and the almost Neo-Baroque dome—which at Haussmann’s insistence was added to close the vista down the new north-south artery—not the uncharacteristically flat and delicately detailed façades. Far finer is the front of that section of the École des Beaux-Arts facing the Seine which was built by Bailly’s master Duban in 1860-2, finer and doubtless also truer to the most exigent taste of the day. Rather directly expressive of its interior uses—it houses exhibition galleries, etc.—the detailing of this façade is quite original without being at all cranky like Duc’s on the Palais de Justice, and the whole very subtle in composition (Plate 72B). Much of the cold severity characteristic of the previous half-century remains; but Duban was clearly trying to be creative, not archaeological, so that one cannot properly apply stylistic names from the past, not even to the extent that it is possible to do so in the case of the New Louvre and the Opéra. However, such high distinction of design as Duban achieved here was rather rare in Second Empire Paris; it parallels in this period the equally exceptional distinction of Henri Labrouste’s Bibliothèque Sainte-Geneviève of the forties.

The accepted range of stylistic inspiration was so wide that it is often only a certain syncretism that gives buildings of this period, nominally in any one of half a dozen ‘styles’, a recognizably contemporary flavour. So also new methods of construction, rather than superseding masonry in toto and thereby demanding original expression as in Victor Baltard’s Central Markets, were more characteristically fused with it, as in the reading-rooms of Labrouste’s libraries. Of these only the later, that in the Bibliothèque Nationale, was built under the Second Empire (Plate 69). Except for this Salle de Travail of 1861-9 and the Magasin or stacks, both so exciting to posterity, most of Labrouste’s other work at this institution, begun in 1855, is as derivative as his private houses; for the most part it is actually hard to say where the old seventeenth- and eighteenth-century buildings stop and his nineteenth-century additions and those of his successor J.-L. Pascal (1837-1920) begin.

Despite the increasing use of metal in all sorts of buildings, there was undoubtedly less sympathy for it than earlier, and hence less success in finding appropriate expression of its qualities (see Chapter 7). By exception, however, the Central Markets in Lyons of 1858 by Antoine Desjardins (1814-82), a pupil of Duban, have a somewhat Labrouste-like elegance in the arched and pierced metal principals spanning the three naves that is not found in Baltard’s so much larger Central Markets in Paris.

In church architecture something like full eclecticism reigned in Paris under Napoleon III, although Gothic was most popular in the provinces. The new Parisian churches generally occupy focal points where major avenues join or boulevards change direction; but, like the Opéra, they have little visual relation to the sober settings provided by the blocks of flats among which they are placed. Instead, each one seems intended to illustrate an alternative mode quite different from the standard urban vernacular of the day.

Saint-François-Xavier in the Boulevard Montparnasse was begun by the elderly Lusson in 1861 and finished by T.-F.-J. Uchard (1809-91) in 1875. With its basilican plan and cold Early Renaissance detail, this might well have been built under Louis Philippe. Saint-Jean-de-Belleville by Lassus, on the other hand, begun in 1854 and completed in 1859 after his death, while larger and rather better built than his churches of the forties, hardly represents any advance over Gau’s Sainte-Clotilde, completed by Ballu only two years earlier. Neo-Gothic could hardly be duller. However, Saint-Denys-de-l’Estrée (Plate 98), the parish church of the suburb of St-Denis, designed by Lassus’s associate and successor Viollet-le-Duc[193] in 1860 and built in 1864-7, is more comparable in quality to the contemporary High Victorian Gothic churches of England (see Chapter 11).

Victor Baltard’s church of Saint-Augustin, also of 1860-7, is not located, like the Gothic edifices by Lassus and Viollet-le-Duc, in a working-class district or suburb, but occupies a very prominent if awkwardly narrow triangular site in the Boulevard Malesherbes near its intersection with the Boulevard Haussmann. Considering the success of his Central Markets, it is not surprising that Baltard used iron here; but he did so with much less consistency and thoroughness than Boileau had done at Saint-Eugène (see Chapter 7). The arched iron principals of the roof accord very ill with the Romanesquoid-Renaissance design of the masonry structure below. The front, with its great rose window, is somewhat more effective. At least it provides a strong urbanistic focus among the standardized ranges of blocks of flats that line the boulevards in this quarter. Two other big Parisian churches are similar in quality although quite different in appearance. Ballu, in addition to finishing Sainte-Clotilde, built both Saint-Ambroise in the Boulevard Voltaire, which is certainly more plausibly Romanesque than Saint-Augustin, and also La Trinité in the Rue de la Chaussée d’Antin, which is much less plausibly _François I_ than his later work at the Hôtel de Ville. La Trinité was built in 1861-7, Saint-Ambroise in 1863-9. Both are vast and pretentious, but neither has much positive character. Like so many comparable examples of the eclecticism of this period in other countries, it is by their faults and not by any characteristic virtues that they are readily recognizable as products of the Second Empire.

Two Romanesquoid churches less prominently located, and hence less well known, are considerably more interesting. One is the parish church of Charenton, Seine, built by Claude Naissant (1801-79) in 1857-9; this is clearly composed and detailed with a somewhat eclectic elegance not unworthy of Labrouste or Duban. Much larger is Notre-Dame-de-la-Croix in the Rue Julien-Lacroix in the Menilmontant quarter of Paris. Built by L.-J.-A. Héret (1821-99), a pupil of Lebas, in 1862-80, this is a cruciform edifice with the vaulting ribs all of openwork iron like those of Saint-Augustin. For archaeological plausibility it compares not unfavourably with Questel’s church at Nîmes, begun some twenty years earlier, in the design of the masonry portions of the structure.

The only big Paris church of the sixties of much real distinction—the only French church, for that matter—is Saint-Pierre-de-Montrouge at the intersection of the Avenue du Maine and the Avenue d’Orléans. This was built by J.-A.-E. Vaudremer (1829-1914), a pupil of Blouet and Gilbert, in 1864-70. Romanesque and Early Christian—perhaps more specifically Syrian—in inspiration,[194] this basilica is notably direct in its structural expression, nobly scaled, expressively composed, and restrained almost to the point of crudity in its detailing (Plate 72A). Vaudremer’s Santé Prison off the Boulevard Arago in Paris, which was commissioned in 1862 and built in 1865-85, is also Romanesquoid or at least in a sort of very simple _Rundbogenstil_. The still quite Durandesque character of this prison illustrates Vaudremer’s close linkage, through the work of his two masters, who had both specialized in designing prisons and asylums under Louis Philippe, with the classicizing rationalism of 1800. His much later Lycées of the eighties, Buffon and Molière in Paris and those at Grenoble and Montauban, on the other hand, reflect the more Gothic rationalism of Viollet-le-Duc (see

## Chapter 11).

Vaudremer’s work may have had some influence, around 1870, on the American Richardson, who was still a student in Paris when Saint-Pierre-de-Montrouge was begun (see Chapter 13). However, no significant line of development led forward in France from his sort of church design. In a smaller and later Parisian church, Notre-Dame in the Rue d’Auteuil of 1876-83, Vaudremer himself showed no further development of his personal style, though the interior here is not unimpressive in its scale and proportions.

The vast and prominent church of the Sacré-Cœur on Montmartre in Paris was begun by Paul Abadie[195] (1812-84), a pupil of Leclerc, well after the Second Empire was over in 1874, and largely finished before the end of the century by the younger Magne (Lucien, 1849-1916). This is Romanesque in inspiration, too, but painfully archaeological—’painfully’, because its architect, in carrying out the restoration of his principal medieval exemplar, Saint-Front at Périgueux, seems to have sought to provide ‘precedent’ for several of the features that he introduced here! Yet the bold exploitation of the remarkable site of this church, dominating Paris from the heights of Montmartre, and the bubble-like silhouette of its cluster of domes when seen from a distance give the Sacré-Cœur positive qualities lacking in most other French ecclesiastical work of the later nineteenth century except Saint-Pierre-de-Montrouge.

Architecture in France had been a highly centralized profession ever since the late seventeenth century. Under Louis XV a few provincial cities showed some capacity for independent activity, but this subsided during the unproductive years that followed the Revolution. Except to a certain extent in Lyons and Marseilles, local activity did not revive very notably in the first half of the nineteenth century. Under the Second Empire most French cities still remained content to follow the lead of Paris. There is hardly a large provincial town which did not—to stress first the positive side of the picture—lay out broad boulevards or straight avenues and line them with more or less successful versions of the _maisons de rapport_ of Paris; on the negative side, the public buildings and churches were usually derived from, and too often very inferior to, prominent Parisian models.

In the centres of the biggest cities one can well believe that one has not left Paris. Occasionally, however, there are urbanistic entities which have more vitality than the rigidly controlled and tastefully restrained new squares and streets of the capital. The fairly modest square in front of the cathedral at Nantes, with its ranges of high-mansarded blocks, is a case in point. Better known is the rising slope of the Cannebière, continued in the Rue de Noailles and the Allées de Meilhan at Marseilles, with the columnar dignity of the Chamber of Commerce on the left near the Vieux Port at the bottom and the paired Gothic towers of Saint-Vincent-de-Paul closing the vista at the top. Public buildings in smaller cities sometimes have a rather illiterate sort of gusto in their boldly plastic massing and exuberantly coarse detailing closer to Second Empire work abroad than to that of Paris; to some eyes these have a theatrical charm not unlike the period flavour of Offenbach’s operas. They often date from well after 1870.

Espérandieu’s Neo-Baroque Palais Longchamps at Marseilles has been mentioned (Plate 70A). Also at Marseilles is the enormous Romanesco-Byzantine cathedral of 1852-93, which was designed by the younger Vaudoyer (Léon, 1803-72), a pupil of his father and also of Lebas. Espérandieu became _inspecteur_ on the job in 1858 and carried on the work after Vaudoyer’s death. This is hardly superior to Ballu’s Paris churches, much less to Vaudremer’s or even Abadie’s, but it is more striking plastically in its rather redundant combination of domed west towers, crossing dome, and transeptal domes; it is also exceptionally colouristic for France. There is an almost High Victorian Gothic brashness in the treatment of the exterior walls with bands of alternately white and green stone. Here the aggressive assurance of the period speaks with an even louder voice than at the New Louvre and the Paris Opéra; this assurance is echoed, moreover, near by in Espérandieu’s own high-placed church of Notre-Dame-de-la-Garde of 1854-64, a scenic accent of the most brazen Second Empire vulgarity.

The Marseilles Exchange, however, dominating its own tree-lined square, is rather similar to the Chamber of Commerce in the Cannebière as it rises among ranges of houses that are more Provençal than Parisian in the modesty of their painted stucco fronts. Originally begun in 1842 by Penchaud, the Exchange was largely built in 1852-60 by his pupil Coste, but its style remains _Louis Philippe_ rather than Second Empire.

The great elaboration and consequent expensiveness of Second Empire modes of design, as generally executed in France in fine freestone, restricted their full exploitation to the capital and the largest provincial cities. There is a sort of economic striation, from the immense sums the Emperor and, after him, the authorities of the Third Republic—even though relatively impoverished—were willing to put into representational public construction at the top, through the level represented by what Parisian investors spent on blocks of flats or rich provincial cities on their principal monuments, down finally to the niggardly building budgets of small towns and villages. This striation provides a sort of analogue to the breakdown of that earlier stylistic unity which had been so marked and happy a characteristic of French architecture for at least a century and a half. That this breakdown was still relative in France is apparent when one turns to other countries where eclectic taste in this period was bolder and where the variation in expenditure on different sorts of buildings was at least as great.

French architectural prestige revived internationally in the fifties to remain surprisingly high for another two generations.[196] However, the Second Empire mode was gradually succeeded internationally by another Parisian mode to which it is convenient to apply the name ‘Beaux-Arts’, from the École des Beaux-Arts out of whose instruction it stemmed. More and more foreigners went to Paris to study as the second half of the century wore on, until Paris became almost what Rome had been in the eighteenth century. In architectural education the influence of the École was especially strong in the New World; the training of English and most Continental architects was much less affected. The first two architectural schools to be founded in the United States, both by William Robert Ware (1832-1915)—himself, curiously enough, a practitioner of a fairly aggressive sort of Victorian Gothic (see

## Chapter 11)—that at the Massachusetts Institute of Technology in Boston

opened in 1865 and the somewhat later school at Columbia University in New York, were both based on the methods of the École.[197] French winners of the Prix de Rome were increasingly imported to serve as teachers, and three generations later the last of them had not yet left the United States. The influence of the École in Latin America was even more powerful, and the dominance of its ideas has lasted in some countries down almost to the present.[198]

Both in the New World and the Old most cities grew like weeds in the third quarter of the century; the analogy is, indeed, a rather accurate one, for the growth was characteristically rank, uncontrolled, and destructive of earlier architectural amenities. Various European capitals, however, imitating Napoleon III’s re-organization of Paris, took advantage of the clearing away of their fortifications to lay out something equivalent to the _grands boulevards_. Florence during the late sixties, for example, when it was very briefly the capital of Italy, saw the laying out, according to the general plan of Giuseppe Poggi (1811-1901), of a range of avenues and squares that extend around the city to the east, north, and west on the site of the old walls. These districts, built up over the years 1865-77, display little or none of the new Second Empire afflatus. For the most part everywhere in Italy in this period the architecture is of generically Renaissance revival character. Only in the much later Piazza della Repubblica, carved out of the slummy heart of the city in the 80s and 90s, is there a heavy pomposity of scale that is curiously un-Florentine—the centre of nineteenth-century Athens might be Neo-Greek, but it was Munich, not Florence, that became characteristically Neo-Tuscan!

In the old Savoy capital of Turin, where the first half of the century had seen such notable urbanistic projects, a vigorous local tradition continued to control most of the new work.[199] However, at the farther side of the Piazza Carlo Felice the Porta Nuova Railway Station was built in 1866-8, as was mentioned in Chapter 3, by the engineer Mazzuchetti and the architect Ceppi in a rather original sort of _Rundbogenstil_. The vast iron and glass lunette at the front still provides a handsome termination to the long axis of the Via Roma, although the rear of the station has been rebuilt since the War. Along the Corso Vittorio Emmanuele II the earlier arcades of Promis were continued almost indefinitely; but the detailing of the façades grew continually richer in evident emulation of Second Empire Paris. This influence also affected the building up of the contiguous quarter of the city. In the fine new square at the end of the Via Garibaldi, however, balancing the earlier Piazza Vittorio Veneto at the end of the Via Po, the Piazza dello Statuto opened in 1864, the façades by Giuseppe Bollati (1819-69) are not at all Parisian, but recall rather the local Academic Baroque of Juvarra. Especially effective, and rare in Turin, are the warm and tawny colours of the painted stucco walls here.

With the uniting of Italy and the eventual taking over of Rome as the capital of the kingdom of Italy on the downfall of Napoleon III in 1870, a tremendous expansion[200] of the old Papal city began. The two principal new streets extending eastward, the Via Venti Settembre and the Via Nazionale, were laid out in 1871 and built up over the next fifteen years. Vast and tawny-coloured like the Piazza dello Statuto in Turin, but much less distinguished in design, is the Finance Ministry in the former street built by Raffaele Canevari (1825-1900) in 1870-7. Equally grand in scale and much more dignified are the quadrantal façades of the Esedra built by Gaetano Koch (1849-1910) in 1885 at the head of the Via Nazionale facing Michelangelo’s Santa Maria degli Angeli (Plate 76A). With the fine later fountain by A. Guerrieri and Mario Rubelli in the centre this provides a most impressive piece of late-nineteenth-century academic urbanism. It still offers a not altogether unworthy preface to the Baths of Diocletian—of which it actually occupies the site of the largest exedra—and to the new railway station (Plate 183B), both so near, which epitomize between them the ancient and the modern worlds in the architecture of Rome.

Koch’s Palazzo Boncampagni in the Via Vittorio Veneto, now the American Embassy, built in 1886-90, is also very dignified. It represents very well the occasional tendency in that decade towards restraint and sobriety in Renaissance design, a tendency that balances the contemporary stylistic development towards the Neo-Baroque. In the Via Nazionale the two most prominent edifices[201] by Italian architects, the Palazzo delle Belle Arti of Pio Piacentini (1846-1928) begun in 1882 and Koch’s Banca d’Italia of 1889-1904, are both quite academic in a respectable Renaissance way, and in the latter case impressively monumental as well. The same applies _a fortiori_ to the two principal public edifices begun in Rome in the eighties—not the respectability, goodness knows, but the monumentality. The enormous Palazzo di Giustizia, in a new quarter across the Tiber, is an incredibly brash example of Neo-Baroque loaded down with heavy rustication, doubtless of Piranesian inspiration. This was designed by Giuseppe Calderini (1837-1916) in 1883-7 and built in 1888-1910 without the intended high mansards.

But the most overpowering new structure in Rome, dominating the whole city and blocking the view of both the ancient Forum and the Renaissance Campidoglio, is the Monument to Victor Emmanuel II, rising above the much enlarged Piazza Venezia at the head of the Corso. Largely the work of Count Giuseppe Sacconi (1854-1905),[202] who in 1884 won the third competition held for its design, this was begun in 1885 and continued after his death by Koch, Piacentini, and M. Manfredi (1859-1927), being finally brought to completion only in 1911 by the engineer R. Raffaelli. Hardly Second Empire nor yet quite ‘Beaux-Arts’, this most pretentious of all nineteenth-century monuments well illustrates the total decadence of inherited standards of Classicism in Europe towards the end of the century. It can be compared only with Poelaert’s Palace of Justice in Brussels, begun twenty years earlier, and entirely to the latter’s advantage even as regards mere gargantuan assurance.

In general, Italian production of the second half of the century is of relatively slight interest; moreover, it often seriously upsets the balance of earlier urban entities by its heavy scale. The great exception, and the one ranking Italian work of the period, is generally recognized to be the Galleria Vittorio Emmanuele II in Milan. In Genoa, behind the theatre, the Galleria Mazzini of 1871 also exceeds in length, in height, and in elaboration all the galleries and passages built in various European cities in the first half of the nineteenth century, yet it is not essentially very different from them in its scale or its detailing. The vast cruciform Galleria in Milan, however, extending from the Piazza del Duomo to the Piazza della Scala, with a great octagonal space at the crossing, is in concept and in its actual dimensions more a work of urbanism than of architecture (Plate 75B). Built with English capital by an English firm, the City of Milan Improvement Company Ltd, and even, presumably, with some English professional advice—M. D. Wyatt was a member of the English board—this tremendous project more than rivals the greatest Victorian railway stations of London in the height, if not the span, of its metal-and-glass roof. But the actual designing architect was Italian, Giuseppe Mengoni (1829-77), and the Galleria de Cristoforis provided him with at least a modest local prototype. Erected in 1865-77 and now completely restored to its pristine richness and elegance, the Galleria scheme involved the enlargement of the Piazza del Duomo and the lining of two of its sides with related façades—executed only partly from Mengoni’s designs—as also the regularization of the Piazza della Scala. Alessi’s sixteenth-century Palazzo Marino, itself of almost Second Empire lushness, was enlarged to serve as the offices of the municipality and provided with a new façade in Alessi’s extreme Mannerist style across one side of the square facing La Scala. This was carried out in 1888-90 by Luca Beltrami (1854-1933), who had studied in Paris at the École des Beaux-Arts, to serve as municipal offices.

Like all the other most prominent buildings of this period, Mengoni’s Galleria makes its impression by its size, its elaboration of detail, and above all its unqualified assurance. From the triumphal-arch portal, rising as high as the nave of the medieval Duomo, to the gilded arabesques of the pilasters, all is obvious, expensive, and rather parvenu; yet the setting—at once so comfortable and so magnificent—that it provides for urban life, centre as it has always remained of so much Milanese activity, has not been equalled since.[203] The Galleria Umberto I in Naples is a late and rather inferior imitation whose ornate entrance most ungenerously overpowers the San Carlo Theatre across the street. This was built by Emmanuele Rocco in 1887-90.

After Paris the most extensive and sumptuous example of the re-organization of a great city carried out in this period is not in Italy but in Austria. Vienna had been relatively inactive architecturally in the first half of the nineteenth century under Francis I (see Chapter 2). His successor Francis Joseph, however, who came to the throne in 1848, set out in the following decades as _Kaiser_ and _König_ to see that his Imperial and Royal capitals should rival Napoleon III’s Paris. In 1857 the fortifications surrounding the old city of Vienna were removed, and the following year Ludwig Förster (1797-1863) won the competition for the layout of the Ringstrasse that was to take their place. The execution of this project, with many modifications, took some thirty years (Plate 74). Outside the actual walls there had been a wide glacis, and therefore the Ring could be developed not merely as a series of wide tree-lined boulevards like those of Paris but with large open spaces in which major public buildings were grouped. These edifices are even more various in style than the comparable ones in Paris, despite the fact that they were the work of a very closely knit group of architects. None of them is of specifically Second Empire character, though the high mansards and the pavilion composition of the New Louvre were used fairly frequently on private buildings in Vienna and throughout the Austro-Hungarian Empire.

The earliest major project of Francis Joseph was the construction of the Arsenal, begun in 1849, where most of the leading architects of the period worked (see Chapter 2). All in various versions of the _Rundbogenstil_, this group of buildings culminates in the centrally placed Army Museum of 1856-77 by Förster and his Danish son-in-law Theophil von Hansen (1813-91). On this the very ornate detail is Byzantinesque and Saracenic in inspiration, yet it is not without a distinctive flavour that is unmistakably of this particular period: the brilliant polychromy of the red and yellow brick walls almost seems to echo, like Vaudoyer’s Marseilles cathedral, the bolder effects of the contemporary High Victorian Gothic architects of England.

Ferstel’s bank in the Herrengasse of 1856-60, also _Rundbogenstil_, has been mentioned earlier. The North Railway Station of 1858-65 by Theodor Hoffmann was _Rundbogenstil_ of an even more ornate sort, with only a rather modest iron-and-glass-roofed shed set between its two massive masonry blocks. This was badly damaged by bombing in the last War but not totally destroyed. On the other hand, the South Station, built in 1869-73 by Wilhelm Flattich (1826-1900), a pupil of Leins in Stuttgart, was of rather conventional High Renaissance character.

The typical and, one may suppose, the preferred stylistic vehicle of most Viennese architects in these decades was, indeed, a rather rich High Renaissance mode. This, for example, Hansen used very effectively for the Palace of Archduke Eugene of 1865-7 and for the Palais Epstein at 1 Parlamentsring of 1870-3. He and Förster, and after Förster’s death Hansen alone, as well as many other architects, employed this mode ubiquitously for various big blocks of flats along the Ring and elsewhere (Plate 74). Good examples are such new hotels of the period as the former Britannia, still standing in the Schillerplatz, and the Donau, which once rose opposite the North Station. Both are by Heinrich Claus (1835-?) and Josef Grosz (1828-?) and were built in the early seventies. Their rather Barryesque raised end-pavilions, without mansards, and the heavily sumptuous detailing of the façades are most characteristic. The better known Sacher’s Hotel behind the Opera House, built by W. Fraenkel in 1876, is somewhat smaller and less lush, at least externally. The block at 8 Operngasse, built by Ehrmann in the early sixties, was topped with Parisian mansards, as are also the long blocks in the Reichstrasse behind the Parlament and the University on either side of the Rathaus; these also have open arcades at their base somewhat like those in Turin.

As along the boulevards of Paris, there is a considerable homogeneity in the private architecture that lines the Ring and the many squares and streets that were built up at the same time. Only in the design of public monuments—often by much the same architects, it is worth noting—did a pompous and somewhat retardataire eclecticism rule. Consider the major works of Ferstel: his bank is _Rundbogenstil_; his Votivkirche of 1856-79 is Gothic; his University something else again.

Ferstel’s Gothic must be compared, not with the distinctly original High Victorian churches of its period in England (see Chapter 10), but with Gau’s earlier Sainte-Clotilde in Paris (see Chapter 6): it is certainly a considerable improvement over that in the general justness of the scale and the plausible laciness of the fourteenth-century detail. But in English terms the Votivkirche is still Early rather than High Victorian. The painted decoration by J. Führich and others, somewhat more discreet than that in the chief _Rundbogenstil_ churches of Vienna, relieves effectively the coldness usual in these big Continental examples of Neo-Gothic.

Ferstel’s much later University of 1873-4, which stands next door to his church and balances Hansen’s precisely contemporaneous Grecian Parlament (see Chapter 2), is a richly plastic pavilioned composition of generically Renaissance character. It also has a high convex mansard over the central block like those on the New Louvre, a feature echoed on the Justizpalast in the Schmerlingplatz, built by Alexander Wielemans (1843-1911) after the University in 1874-81. So much for the main works of one leading architect of the period. Not all Ferstel’s contemporaries had quite so varied a stylistic repertory, however.

In Vienna, as in Paris, one of the most conspicuous and also one of the most successful and original of the new public buildings was the Opera House. This was built in 1861-9 by Van der Nüll & Siccardsburg in a mode quite unrelated to their earlier work at the Arsenal but one not easy to define. The Vienna Opera House is a somewhat simpler and less boldly plastic structure than Garnier’s, both in its generally right-angled massing, with pairs of rectangular wings projecting on each side towards the rear, and in the rather flat, somewhat _François I_ detail. Yet the vast curved roof, actually rather like that over the buildings along the Rue de Rivoli, does give it a distinctly Second Empire air (Plate 74). Less grandly sited than the Paris Opéra, it was none the less balanced across the Opernring by one of the largest and handsomest of Hansen’s private works, the Heinrichshof of 1861-3 (Plate 73B). This had a fine glass-roofed passage through its centre and ranges of flats behind the elaborate Late Renaissance façades. It has unfortunately been demolished since the War to make way for a very poor modern block of offices.

Here by the Opera House, as at the Place de l’Opéra in Paris, the Viennese urban achievement of the age was concentrated. The Heinrichshof, with its raised central portion matching the high roof of the Opera House opposite and its corner towers corresponding to the mansarded pavilions of more definitely French-styled blocks of flats, offered a handsomer Austrian equivalent of the Second Empire mode than does the Opera House itself; for the Opera House lacks externally the lushness and bombast characteristic of the period at its most assured, while the auditorium within, re-opened in 1955, is today a much simplified reconstruction by Erich Boltenstern (b. 1896). Yet the masonry exterior of the Opera House is clean and fresh today thanks to Boltenstern’s restoration and, with the great staircase and foyer regilded and refurbished generally, it offers a lighter and more festive vision of the period than do the vast majority of Viennese buildings whose stucco so often badly needs a coat of paint.

Hansen’s Musikvereinsgebäude of 1867-9 in the Dumbagasse is academic in an almost eighteenth-century way, both as regards the general organization of the exterior and the restraint of the detailing. In his still later Parlament of 1873-83, as has been noted earlier, he produced the last grandiose monument of the Greek Revival. More characteristic, however, is his contemporaneous Academy of Fine Arts of 1872-6 in the Schillerplatz. This is externally in the Renaissance mode that he presumably preferred after he left Athens, but it has Grecian detailing inside of a delicacy and elegance that recalls the thirties. Especially handsome is the colonnaded Aula in the centre, even though its rich painted ceiling of 1875-80 by Anselm Feuerbach is inappropriately Baroque in a rather Rubens-like way.

Another Austrian architect besides Ferstel was using Gothic for prominent Viennese edifices in this period (see also Chapter 11). After Ferstel’s Votivkirche the next Neo-Gothic structure was the Academische Gymnasium in the Beethovenplatz; this was built in 1863-6 by Friedrich von Schmidt (1825-91), who had worked earlier under Zwirner on the restoration and completion of Cologne Cathedral. But the school was soon outshone in size and in elaboration by Schmidt’s Rathaus of 1872-83. This stands between Hansen’s Parlament and Ferstel’s University but in a line with the Reichstrasse at their rear. The Vienna Rathaus is certainly not unrelated to G. G. Scott’s Victorian Gothic and that of Waterhouse in England, particularly in the side wings that end, eclectically enough, in high-mansarded pavilions. But the general fussiness of the turreted front recalls rather pre-Puginian Gothic, say Porden’s Eaton Hall of seventy years earlier (see Chapters 6 and 10).

Despite the total visual unlikeness of the Rathaus to its Grecian neighbour, the Parlament, both have a similarly obsolete air. It is as if Francis Joseph’s presumptive intention in the fifties of outbuilding Napoleon III had been succeeded by a belated and rather provincial desire to outrival the larger structures in other countries in the two leading modes of the previous period, the Greek Revival and the Gothic Revival, neither much represented hitherto in Vienna.

Yet an equally prominent public monument of the seventies and eighties, the Burgtheater, which stands just opposite the Rathaus, is of a Late Renaissance, almost Neo-Baroque order, with a distinctly Second Empire flavour to its bowed front and generally very plastic composition (Plate 73A). This, the most distinguished of all the public monuments along the Ringstrasse, was built in 1874-88 by Semper, whose international career in Germany, England, and Switzerland wound up in Vienna after he was called there in 1871 by Francis Joseph to advise on the extension of the Hofburg Palace. Except perhaps in its bowed front, this Viennese theatre does not much resemble the rebuilt Dresden Opera House of 1871-8 which Semper had just designed (see Chapter 9). Perhaps Semper and his Viennese partner Karl von Hasenauer (1833-94), a pupil of Van der Nüll and of Siccardsburg, were somewhat influenced by the plans on which they were working together for the extension of the nearby palace; these were, not inappropriately, in the Austrian Baroque of Fischer von Erlach’s unfinished Michaelertrakt of the Hofburg dating from the second quarter of the eighteenth century. However that may be, the theatre, boldly scaled and tightly composed, is a far more successful building than the very derivative Neue Hofburg projecting out towards the Ring as that was executed in 1881-94 by Hasenauer after Semper’s death. The post-War restoration of the theatre and the rebuilding of its auditorium are by Michel Engelhart (b. 1897).

Semper and Hasenauer’s two vast Museums of Art History and Natural History face each other on a large square across the Burgring from the Neue Hofburg. Of identical design, they were both largely built in 1872-81. In the treatment of the exteriors—they were finished internally only very much later—as also in some of Hansen’s very latest work in Vienna, one senses a conscious rejection of the bold plasticity and the compositional elaboration characteristic of the preceding decades, and most notably of the Burgtheater. The Renaissance detail is by no means sparse, but there is an academic sort of primness and orderliness belonging to the last quarter of the century such as has been noted earlier in Koch’s Roman work.

The Bodenkreditanstalt built by Emil von Förster (1838-1909), Ludwig’s son, in 1884-7 is still more severe in its Florentine _quattrocento_ way, recalling the more Tuscan aspects of the _Rundbogenstil_. With this may be contrasted the unashamed Neo-Baroque of Karl König’s Philipphof of 1883, introducing one of the modes most characteristic of the end of the century in both Austria and Germany.

Budapest, the second capital of the Austro-Hungarian Empire, was also much embellished with public buildings by Francis Joseph. Stüler from Berlin worked here, using a quiet version of the _Rundbogenstil_ for the Academy of Sciences in 1862-4. But the later and more ornate _Rundbogenstil_ of Berlin and Vienna had already been echoed in Budapest by Frigyes Feszl (1821-84) in the Vigado Concert Hall of 1859-65. This could easily be by Ferstel, so similar is it to his bank in Vienna. The leading Hungarian architect of the period, Miklós Ybl (1814-91), who was trained in Vienna, also used the _Rundbogenstil_, but of a rather more Romanesquoid sort, for the Ferenczváros Parish Church which he built in 1867-78. However, his Renaissance Revival Custom House of 1870-4 is more nearly up to the best Vienna standards of the day as maintained by Hansen. The Opera House that Ybl built in 1879-84, with its boldly convex mansards, vies in its rich plasticity with Garnier’s, but none too successfully. The Szent Lukásh Hotel by R. L. Ray (1845-99), a Swiss-born pupil of Gamier, is one of the largest mansarded Second Empire hotels anywhere in the western world. On the whole, the dominant influences in Hungary were Austrian and German, however, not Parisian, as is hardly surprising. No autochthonous note was struck; as is true of all Eastern Europe, the architecture of this age is as essentially colonial in character as in the outlying British Dominions or in Latin America, although the models emulated were rather different.

------------------------------------------------------------------------

## CHAPTER 9

SECOND EMPIRE AND COGNATE MODES ELSEWHERE

IN the cities of Germany and of Northern Europe generally there were in this period no such comprehensive urbanistic developments as in Paris and Vienna. Some individual public monuments are, perhaps, not inferior to those that Napoleon III and Francis Joseph obtained from their architects; but these are rarely grouped into such coherent entities as the Marktplatz in Karlsruhe of the first quarter of the century or the Ludwigstrasse in Munich of the second quarter. The domestic building of the period is also considerably less consistent in character than in Paris and Vienna.

The architectural scene in Germany was overshadowed by the distinguished achievements of the previous period. The Schinkel tradition, although increasingly corrupted, lasted on almost indefinitely not merely in Prussia but in most German states. Stüler, Schinkel’s ablest disciple in Berlin after the death of the short-lived Persius, remained an internationally respected practitioner. He was employed in Sweden and in Hungary, as has been noted, not to speak of German cities, down to his death in 1865. By him and by many others the _Rundbogenstil_ was employed quite as late as in Austria-Hungary both in the various German states and also in the Scandinavian countries. Such a very large and prominent public building as the Berlin Rathaus of 1859-70 by H. F. Waesemann (1813-79) well indicated the long-continued hold of this mode on German officialdom. Nor was this particularly inferior in quality to much similar work produced in the earlier heyday of the _Rundbogenstil_ before 1850. As in Austria, however, alternative modes were growing increasingly popular, even though none rose to a local dominance comparable to that of revived Renaissance in Vienna. The taste of the period for elaboration, both in general composition and in detail, is everywhere evident regardless of the mode employed.

French influence was not absent; indeed, specifically Second Empire features were perhaps more common than in Austria. G. H. Friedrich Hitzig (1811-81), a former assistant of Schinkel’s, had actually studied in Paris. After Stüler, he was the most prominent and successful architect of the period in Berlin, and in the fifties he built a few mansarded houses there. Along the new Viktoriastrasse in the Tiergarten quarter, where he did a great deal of work in 1855-60, one house among the eight that he built was mansarded; the others and most of those he was erecting near by in the Bellevuestrasse, the Stülerstrasse, and other streets at the same time were, however, in a much elaborated Schinkelesque vein. Suburban houses of the sixties occasionally followed Parisian modes also; but far more were clumsy variants of Schinkel’s and Persius’s Italian Villas, or else in some sort of equally clumsy Gothic.

Public buildings in Germany were only occasionally designed in the mansarded mode and, in general, only after the mid sixties. The Baugewerkschule in Stuttgart, built in 1866-70 by Josef von Egle (1818-99) its director, had projecting centre and end pavilions with crudely Parisian detailing. It is curious to realize that it was contemporary with Leins’s belated but rather distinguished Grecian Königsbau there. In Cologne the High School of 1860-2, and the Stadttheater of 1870-2 by Julius Raschdorf (1823-1914), both destroyed in the last War, were heavily mansarded and very plastically modelled; the latter, at least, on which H. Deutz collaborated with Raschdorf, had some real compositional interest in the tight interlocking of the masses (Plate 77B). Despite their very evidently French character, both were considered by contemporaries to be ‘German Renaissance’—as, for that matter, was Wieleman’s Justizpalast in Vienna—because of the specific precedent of much of the detail; German Renaissance was by this time the latest fashion, but to later eyes these buildings in Cologne were no more characteristic examples of it than the one in Vienna. Raschdorf is better known in any case for his much later Neo-Baroque work, notably the Berlin Cathedral, for which he prepared the design in 1888, although it was not built until 1894-1905.

The Military Hospital by F. Heise in Dresden of 1869 was considerably more French in the strong articulation of the mansarded centre and end pavilions and also in its quite Parisian detailing than Raschdorf’s contemporary buildings in Cologne. More prominent in Dresden by far, however, is the Hoftheater, which is not at all French in character. This was designed in 1871 by Semper after his earlier theatre there had been destroyed by fire; its construction was supervised by Semper’s son Manfred after he settled in Vienna, and completed in 1878. Gone was most of the festive grace and delicacy of his Hamburg and Dresden work of the forties, even though the auditorium was not dissimilar to the one that had been destroyed. Yet in the arrangement of the interior and the disposition of the masses this rivals in clarity of organization the opera-houses of Garnier in Paris and of Van der Nüll & Siccardsburg in Vienna. The plans undoubtedly owed a great deal to the elaborate studies Semper had made for Ludwig II in 1865-7 for an opera-house to be built in Munich especially for the production of Wagner’s operas.

The relative importance of Berlin was, of course, rising well before its establishment as the imperial capital in 1871. Friedrich Hitzig’s most considerable public building in Berlin, the Exchange, built in 1859-63 at the same time that the Rathaus was in construction, was neither Schinkelesque nor _Rundbogenstil_ but in a rather academic sort of Late Baroque (Plate 77A). Hitzig seems to have been consciously recalling what Knobelsdorf built for Frederick the Great and thus presaging the more overt Neo-Baroque of the last decades of the century. His later Reichsbank of 1871-6, on the other hand, was in general considerably more Classical despite its banded and diapered walls in two colours of brick.

The public buildings of Martin K. P. Gropius (1824-80) are also indicative of the general stylistic stasis of this period in Germany. His Museum of Decorative Art in Berlin, begun in 1877 and completed in 1881 by Heinrich Schmieden (1835-1913), resembled Hitzig’s houses of the fifties in its Grecian elaboration; it also recalled Klenze’s Hermitage Museum, built more than a generation earlier in Petersburg. Gropius & Schmieden’s still later Gewandhaus in Leipzig of 1880-4, however, is less reminiscent of Schinkel or Klenze and more conventionally academic. This concert hall was renowned for its superb acoustics.

It is easy to forget how much the architects of these decades, apparently obsessed with stylistic elaboration, were also concerned to incorporate in their buildings all sorts of technical advances. Iron may show less than in the previous period, but it was quite consistently used behind the scenes. Central heating, extensive sanitary equipment, vertical transportation, and various other things that are taken for granted today first became accepted necessities in these decades. But it was only in the commercial field—and in England and the United States above all—that such technical innovations influenced architecture very positively or visibly (see Chapter 14), however much they must actually have preoccupied architects who seem today so imitative and retardataire. The Anhalter Bahnhof in Berlin by Franz Schwechten (1841-1924), however, built in 1872-80, did represent a real advance over the principal English railway station of this period, St Pancras in London of 1863-76, in the clarity and coherence of its organization. One can hardly say that the shed roof of the Anhalter Bahnhof was in the _Rundbogenstil_; yet it is much more happily related in scale and shape to the masonry elements of the station than are the two parts of that in London, world-famous nonetheless until the nineties for the unrivalled span of its shed.

Architectural activity in Bavaria was of a very different order. The Ludwigsschlösser,[204] the country palaces that Ludwig II of Bavaria erected for his private delectation after he succeeded Maximilian II in 1864, are the playthings of a monarch mad about Louis XIV. Linderhof, built in 1870-86, revived a local Bavarian sort of Baroque, and was thus even more premonitory of a favourite German mode of the eighties and nineties than Hitzig’s Berlin Exchange (Plate 84). Herrenchiemsee, first projected as early as 1868 but begun only in 1878, is a direct imitation of Versailles. Neuschwanstein, on the other hand, is a wild Wagnerian fantasy of a medieval castle occupying a superb mountain site.

It must be assumed that the architect of the first two, Georg von Dollmann (1830-95), was little more than the draughting agent of his master’s dreams of grandeur. More interesting than the exteriors are the incredibly rich interiors of Linderhof, operatic recreations of the Bavarian Rococo. Appropriately enough these were designed by Franz von Seitz (1817-83), then director of the Munich State Theatre, who was famous for his stage-sets. At Herrenchiemsee, however, many of the interiors were exact copies of the main apartments of Louis XIV at Versailles. These were executed by Julius Hoffmann (1840-96), who began to work under Dollmann in 1880 and succeeded him in 1884. More original were certain other rooms at Herrenchiemsee designed by F. P. Stulberger after 1883 in an even more elaborate and fantastic Neo-Rococo than those by Seitz at Linderhof.

Ludwig II had another obsession besides the majesty of Louis XIV, and that was the genius of Richard Wagner. This cult is almost nauseatingly reflected at Neuschwanstein, for which Riedel, who had built Schloss Berg in 1849-51, prepared the original design in 1867. Construction there began in 1869, was taken over by Dollmann in 1874, and only completed as regards the exterior in 1881; much of the decoration is still later. Despite Ludwig’s romantic love of the real Romanesque of the Wartburg, Neuschwanstein really differs very little from the fake castles of the first half of the century, except in its very ingenious adaptation to a most precarious site. It is the later interiors, designed by Hoffmann in the early eighties, that attempt to realize the Wagnerian legends both in the architectural detailing and in endless murals. The whole culminates in the Byzantinesque throne room of 1885-6 intended by Ludwig to be a sort of ‘Grail Hall’ from _Parsifal_. The results of his other obsession are more gratifying to the eye.

Never again would any ruler, however, not even in Germany, be so spendthrift a patron of architecture. Considering the deterioration in quality evident in these palaces and castles of the seventies and eighties from the work done for Ludwig’s predecessor Ludwig I or for Frederick William IV of Prussia in the thirties and forties, this was just as well. Fortunately the activities of William II were less related to the building arts; and Hitler, a thwarted architect, had too little time.

Far more typical of the turn German architecture in general was taking in the seventies than the Ludwigsschlösser were such things as the von Tiele house in Berlin by Gustav Ebe (1834-1916) and Julius Benda (1838-97). In its crawlingly rich German Renaissance detail and its irregularly gabled silhouette this prepared the way far more definitely than Raschdorf’s contemporary Cologne buildings for a veritable flood of such coarse work all over Germany in the next decade. This characteristic German mode has analogies with the English style-phase of the seventies and eighties somewhat perversely known as ‘Queen Anne’; more specifically it often resembles very closely what is called ‘Pont Street Dutch’ in England. But leadership comparable to that provided in England by Webb and Shaw was entirely lacking, and even lesser talent of the order of George’s or Collcutt’s (see Chapter 12).

Usually executed in dark-coloured brick with stone trim, this prime manifestation of the bourgeois ambitions of the Bismarckian Empire produced a spate of buildings of all sorts that have come to look very grim indeed with the accumulated smoke of years. Old photographs indicate that many of them once had a certain lightness and even a quite festive air, Wagnerian in the _Meistersinger_ vein rather than in that of the _Ring_ as at Neuschwanstein. But the materials used were always hard and mechanically handled and the execution of the detail at once fussy and metallic. No positive originality in general composition or in planning made up, as with much comparable work in England, for the anti-architectonic character of the basic approach.

A prominent late example is the Rathaus[205] in Hamburg built in 1886-97. This vast and turgid edifice contrasts most unhappily with the suave High Renaissance design of Wimmel & Forsmann’s contiguous Exchange built in the thirties. Its tall tower, moreover, has neither the richness of outline of Scott’s on the Nikolaikirche nor the simple directness of de Chateauneuf’s on the Petrikirche, with both of which it still disputes the central position on the Hamburg skyline.

The nationalistic ‘Meistersinger mode’, so to call it, had only too long a life, lasting well into the twentieth century. But it was early challenged by a new modulation of German taste in the eighties, parallel to that which the English also experienced, towards an eighteenth-century revival—here in Germany definitely Neo-Baroque—of which Linderhof was probably the first really sumptuous and striking example. Ebe & Benda early deserted the German Renaissance for a German Baroque at least as chastened as that of Hitzig’s much earlier Exchange when they built their Palais Mosse in Berlin of 1882-4. In 1882 Paul Wallot (1841-1912), who had also worked earlier in the Meistersinger mode, won the competition for the Reichstag Building with an overpoweringly monumental Neo-Baroque project recalling Vanbrugh more than Bernini or Schlüter. Erected by him in 1884-94, this was soon matched at the inner end of Unter den Linden by Raschdorf’s cathedral.

[Illustration]

[Illustration:

Figure 16A and 16B. Vilhelm Petersen and Ferdinand Jensen: Copenhagen, Søtorvet, 1873-6, elevation ]

Unlike Napoleon III and Francis Joseph, the German emperors William I, Frederick I, and William II did not succeed in making their capital an important exemplar of nineteenth-century urbanism. Moreover, the influential position that Germany had occupied in the international world of architecture in the first half of the century was less and less maintained after the death of Stüler. Not until the twentieth century did Germans again make a significant contribution to European architectural history (see Chapter 20).

With the deterioration of German leadership in the seventies and eighties went also a general decline in the architectural standards of the Scandinavian countries that had so successfully based their later Romantic Classicism and their _Rundbogenstil_ on German models of the thirties, forties, and fifties. In Denmark the work of Meldahl was increasingly inferior to that of Herholdt. Although he was only nine years younger than Herholdt, his direction of the Copenhagen Academy, beginning in 1873, coincided with the feeblest and most eclectic period in Danish architecture, from which recovery started only in the nineties with the early work of Martin Nyrop (1849-1925) in Copenhagen and of Hack Kampmann (1856-1920) in Aarhus (see Chapter 24).

A characteristic urbanistic development of the seventies in Copenhagen, the Søtorvet built in 1873-6 by Vilhelm Petersen (1830-1913) and Ferdinand Vilhelm Jensen (1837-90), is French not German in its ultimate inspiration. This grandiose pavilioned and mansarded range of four tall blocks forms a shallow U-shaped square along a canal (Figure 16). Its definitely Second Empire character may not, all the same, have derived directly from Paris but via German or English intermediaries, so much more typical is this of the international than of the truly Parisian mode of the third quarter of the century.

[Illustration]

[Illustration]

As late as 1893-4 the much more conspicuous Magasin du Nord department store, built by A. C. Jensen (1847-1913) and his partner H. Glaesel in the Kongens Nytorv in Copenhagen, also carried the high mansarded roofs of the new Louvre, both flat-sided and convex-curved, above its end and centre pavilions. The detailing was chastened, however, by memories of local palaces and mansions in the nearby Amalie quarter of the city, where Jensen had worked on the completion of the eighteenth-century Marble Church. The Magasin du Nord thus combines two characteristic aspects of the architecture of the period, evident in most countries but rarely thus joined: a reflection of Napoleon III’s Paris, elsewhere reaching its peak around 1870, and a revival of the style of the seventeenth and eighteenth centuries, generally beginning about a decade later.

In Sweden also there was some Second Empire influence, although nothing very notable resulted from it. The Jernkontovets Building in Stockholm erected by the brothers Kumlien (A.F., 1833-?; K.H., 1837-97) in 1873-5 has a high mansard and pavilions combined with a respectably academic treatment of the façades that is quite different from the bombast of the Søtorvet. Bern’s Restaurant in Stockholm of 1886 by Åbom, whose more conservative Renaissance Revival theatre of thirty years earlier has been mentioned, is similarly Parisian, particularly in the decorations that were provided by Isaeus.

With I. G. Clason (1856-1930) the tide of eclecticism in Sweden turned more nationalistic. The Northern Renaissance of his Northern Museum, built in 1889-1907, parallels somewhat belatedly the Meistersinger mode in Germany; but it also shows a more refined and delicate touch, somewhat like that of George and of Collcutt in England. As in most other countries, the revival of the native sixteenth-century style was soon succeeded by a revival of the Baroque, here rather academically restrained. This phase is most conspicuously represented in Stockholm by the grouped Parliament House and National Bank of 1897-1905 by Aron Johansson (1860-1936). In the nineties Ferdinand Boberg (1860-1946) was also initiating a new movement somewhat comparable to that led by Nyrop in Denmark (see Chapter 24).

The modes of Second Empire Paris left rather more mark on Holland than did those of the First Empire, particularly in the work of Cornelis Outshoorn (1810-75), whose iron-and-glass Paleis voor Volksvlijt in Amsterdam of the late fifties has been mentioned earlier. That is long gone, but the related Galerij, a U-shaped range of mansarded blocks linked by a sort of veranda of cast iron, till lately bounded the south of the Frederiksplein. His enormous Amstel Hotel, near by on the farther side of the Amstel, was built in 1863-7. At Scheveningen the Oranje Hotel (1872-3), also by him, was one of several typical resort establishments there of an international Second Empire order, as is also his hotel at Berg-en-Dal near Nijmegen (1867-9). Fairly generally high mansards rose in the sixties and seventies over the narrow house-fronts in the new quarters of Dutch cities. However, the opposing Neo-Gothic is more significant historically in Holland, and the secular work of Cuijpers as well as his churches, although rather like Clason’s, is better considered in that connexion (see Chapter 11). As in the Scandinavian countries, the nineties saw new beginnings in Holland, in this case with the appearance of Berlage and Kromhout (see Chapter 20).

The principal Anglo-American developments in the second half of the century were in the specialized fields of domestic and commercial building (see Chapters 14 and 15). England, moreover, had from 1850 to the early seventies a lively stylistic development of her own, the High Victorian Gothic, rather different from the later Neo-Gothic of the Continent, which was also very influential in the Dominions and in the United States (see Chapters 10 and 11). Nevertheless, the international Second Empire mode flourished on both sides of the Atlantic among Anglo-Saxons to a greater extent, perhaps, than anywhere in Europe. It is not, of course, possible to subsume all non-Gothic work of these decades in England under the Second Empire rubric any more than on the Continent. Yet, with certain notable exceptions, the most vigorous and conspicuous buildings of a generically Renaissance character were clearly inspired by Paris, and often specifically by the New Louvre, as Prosper Mérimée noted and wrote to Viollet-le-Duc while on a visit to London in the mid sixties.

The most considerable English public monument built just after the mid century, the Leeds Town Hall of 1855-9, is by Cuthbert Brodrick (Plate 78A). That Brodrick was an architect markedly French in his leanings has already been noted in describing his Leeds Corn Exchange, which is later in date but earlier in style than his Town Hall (see Chapter 4). But this major early work, for which Brodrick won the commission in a competition in 1853, is not easily pigeon-holed stylistically. The great hall inside derives quite directly from Elmes’s in Liverpool, designed almost a quarter of a century earlier, though not opened until 1856. The exterior recalls in its grandiose scale the English Baroque of Vanbrugh more than it does anything that had even been projected since the megalomaniac French projects of the 1790s. The Leeds Town Hall is certainly no longer Romantic Classical, no longer Early Victorian; yet except for the rather clumsy originality of some of the detail and the varied outline of the tower—a late emendation of the original project of 1853—it is hard to say how or why it is so definitely High Victorian, and rather a masterpiece of the High Victorian at that. Wallot in Berlin in the eighties approached Brodrick’s mode of design in the Reichstag but had little of his command of scale or his almost Romantic Classical control of mass.

When Brodrick designed his town hall very little was known in England of Visconti’s project of 1852 for the New Louvre, and Lefuel had not yet begun to elaborate the design. So vigorously individual an architect as Brodrick was hardly likely, moreover, to find inspiration in the Hope house of Dusillion or the Hardwicks’ Great Western Hotel. But the wave of Second Empire influence arrived in England well before the Leeds Town Hall was finished. When the English swarmed to Paris to visit the International Exhibition of 1855 the character of the New Louvre became generally known to architects and to the interested public. The Crimean War in the mid fifties served, moreover, to bring English and French officialdom into close contact. To English ministers and civil servants, even more than to architects and ordinary citizens, the existing governmental accommodations in Whitehall contrasted most unfavourably with those Napoleon III was providing in the New Louvre. When a competition was held in 1856-7 for a new Foreign Office and a new War Office to be built in Whitehall, it is not surprising that most of those entrants who were not convinced Gothicists should have modelled their projects more or less on the work of Visconti and Lefuel.

Barry, the head of the profession, did not enter the competition; but unofficially—for he was still an employee of the Government at the Houses of Parliament—he prepared at this time a comprehensive scheme for the development of the whole length of Whitehall from Parliament Square to Trafalgar Square. In this project he crowned all his façades—including that of his already executed Treasury—with mansards, introduced stepped-back courts like that of the New Louvre, and marked the corners and the centres of the court façades in the most Louvre-like way with pavilions crowned by still taller mansards. Had this project of Barry’s been followed, London would rival Paris and Vienna in the extent, the consistency, and the boldness of her public buildings of this period. In fact, practically nothing ever came of it nor, indeed, of the official competition; for by this period earlier traditions of urbanism had all but completely died out and architectural initiative was largely in private hands.

When the competition was judged in 1857, the designs that received the top prizes both for the War Office and for the Foreign Office were in the pavilioned and mansarded manner; they derived, however, at least as much from the Tuileries as from the New Louvre. It was the rising prestige of Napoleon III, of course, that called public attention at this time to the Tuileries which was his residence—as it had been, for that matter, the residence of earlier nineteenth-century French monarchs. Otherwise no one in England would probably have thought of reviving any of the various periods, covering some four centuries, represented in its conglomerate mass or of emulating its pavilioned and mansarded composition.

Since neither of these projects for ministries was ever executed, and their respective architects—Henry B. Garling (1821-1909), on the one hand, and H. E. Coe (1826-85) and his partner Hofland, on the other—never built much else of consequence, it is not necessary to linger over them. However, their designs and other Second Empire ones that received minor premiums were extensively illustrated in professional and general periodicals, and they provided favourite models in the sixties both in England and in the United States. The Paris originals, on which graphic data was not only scarcer but also less readily accessible, were not on the whole so influential. This helps to explain why French influence _appears_ to have been stronger in the Anglo-Saxon world than on the Continent, even though there was probably less direct contact with Paris.

There was also in England at this time a general tendency, even more notable than in Austria or Germany, to enrich and elaborate plastically the long-established Renaissance Revival mode. This is less specifically inspired by Paris. An excellent example is provided by the extensive range of terraces, designed by Sancton Wood (1814-86) in 1857, that flank Lancaster Gate in the Bayswater Road in London with their boldly projecting bay windows linked by tiers of colonnades. In other examples, such as the National Discount Company’s offices at 65 Cornhill built by the Francis Brothers in 1857, the capping of the whole block with a boldly dormered mansard[206] is more obviously of Second Empire inspiration, though the façades below are merely of a much enriched _palazzo_ order.

When the Moseley Brothers designed in 1858 the vast Westminster Palace Hotel near Westminster Abbey at the foot of Victoria Street, a caravanserai intended to exceed the Hardwicks’ Great Western Hotel of 1850-2 in international luxury, they took over its pavilioned and mansarded design. To judge from the relative dignity and sobriety of their detailing, they would seem to have studied contemporary Parisian work—not the New Louvre but the quieter _maisons de rapport_ along the boulevards—rather than merely basing themselves on the prize-winning Government Offices projects as so many others were content to do at this time. This hotel, which proved a failure, now serves as a block of offices, and has been remodelled almost beyond recognition.

The next year Barry designed the Halifax Town Hall, his last work. He did not himself propose to cap this, like the Government Offices in his Whitehall scheme, with French mansards; those that were executed are an emendation by his son, E. M. Barry, who carried the building to completion in 1862 after his father’s death in 1860. But the richly arcaded articulation of the walls and the emphatic forward breaks of the great tower and of the more modest pavilion at the other end clearly emulate, without directly imitating, the sumptuous plasticity of the New Louvre. Nevertheless, the boldly asymmetrical composition, dominated by a single corner tower, is more in the Italian Villa vein (Plate 78B).

This tower—but not the site—was lined up with the axis of Prince’s Street, which enters Crossley Street at this point. The assured quality of its design and above all that of its tremendous spire, more than worthy of Wren in the ingenuity with which the silhouette of a Gothic steeple was built up out of Renaissance elements, makes the Halifax Town Hall thoroughly English and one of the masterpieces of the High Victorian period. Totally devoid of Gothic elements, it has more Gothic vitality than Barry’s Houses of Parliament, at this time just approaching completion nearly thirty years after they were first designed.

E. M. Barry went on to crown two London station hotels, that at Charing Cross in 1863-4 and that at Cannon Street in 1865-6, with mansards; but these were far from being masterpieces, and that at Charing Cross has lately been much modified. The Grosvenor Hotel, built beside the new Victoria Station in 1859-60 by Sir James T. Knowles (1831-1908), is far more original. He covered the whole enormous mass with a very tall convex mansard, giving further emphasis to the broad pavilions at the ends by carrying their roofs still higher and capping them with lanterns. Beyond this nothing was French. The detail indeed, defined by its architect as ‘Tuscan’, i.e. _Rundbogenstil_, is highly individual, partaking of the coarse gusto and even somewhat of the naturalism of the most advanced Victorian Gothic foliage carving of the period (see

## Chapter 10).

Similar mansards, but flat-sided not bulbous, and similar detail characterize a pair of tall terraces that Knowles built in 1860 on the north side of Clapham Common, south of London. These constituted a subtle suburban attack on Early Victorian traditions of terrace-design that soon had metropolitan repercussions. His Thatched House Club in St James’s Street in London of 1865 has a great deal of very rich carving by J. Daymond in the naturalistic vein, but is less interesting in general composition.

Knowles’s Grosvenor was still new when John Giles outbid it with the Langham Hotel, begun in 1864. Given a much finer site than Knowles’s at the base of the broad avenue of Portland Place across from Nash’s All Souls’, Langham Place, Giles rose boldly—most people now think too boldly—to the occasion (Plate 80A). Certainly he overwhelmed Nash’s delicate and ingenious steeple by the rounded projection and the tall square corner tower—now bombed away at the top—with which he faced it. Equally certainly his massive north façade, with its boldly modelled flanking pavilions and its profusion of lively animal carvings, would overwhelm the urbane refinement of the nearby Adam terraces flanking Portland Place had these not by now been replaced by far inferior buildings. For all its gargantuan scale and the somewhat elephantine playfulness of the detail (not to speak of the dinginess to which the ‘Suffolk-white’ brickwork and the stone trim have now been reduced), the Langham is a rich and powerfully plastic composition, most skilfully adapted to a special site, and more original than most of what was produced in the sixties in Paris. The carved animals at the window heads, so varied and so humorous, deserve an attention they rarely receive; these scurrying creatures almost seem to come out of Tenniel, but may actually derive from Viollet-le-Duc.

That this degree of architectural originality, presented with such bold assurance and even bombast, should within a decade or two have come to seem tasteless and actually ugly—as, indeed, it has seemed to many ever since—is not of major historical consequence. The age that achieved it rejected as tasteless and insipid the architectural production of the previous hundred years, and most notably Late Georgian work of the sort to which the Langham stood in close proximity. What _is_ of consequence is that such High Victorian buildings, even when not Gothic, possessed a vitality and a contemporaneity within their period that was very largely lacking in parallel work on the Continent, most of which in any case is a decade or more later in date. In their parvenu brashness, the Grosvenor and Langham balance the contemporary achievement of the Gothic church architects—an achievement generally more acceptable even today as it was already to highbrows and aesthetics in the sixties—without necessarily equalling it (see Chapter 10).

In the English hotel boom of the early and mid sixties which these big London hotels set off, some variant of the anglicized Second Empire became the accepted type of design; indeed, a mansarded French mode continued to be used as late as the nineties[207] for such a big London hotel as the Carlton built by H. L. Florence (1843-1916) in 1897. Many heavily mansarded London hotels of the seventies and eighties are now gone or have been turned, like the earlier Westminster Palace and the Langham, to other uses—among these the former Grand Hotel in Trafalgar Square of 1878-80 by H. Francis and the front block of the former Cecil in the Strand built in 1886 by Perry & Reed may at least be noted here, since they remain so conspicuous and are so exasperatingly unavailable to travellers.

It is a resort hotel, however, the Cliff (now the Grand) at Scarborough in Yorkshire, built by Brodrick at the height of the boom in 1863-7, just before he retired to live in France, that remains internationally the most notable example of the type (Plate 79). And the type could be found in such remote spots as the famous ‘ghost town’ of the Comstock Lode, Virginia City, Nevada, where the large and elaborate hotel is no more, or Leadville, Colorado, where the more modest and much later Vendome Hotel, built by Senator Tabor for his ‘Baby Doe’, is still in use, as well as in big European cities such as Amsterdam, Frankfort, Brussels, and Budapest.

The site of Brodrick’s Grand Hotel is a superb one on the edge of the Scarborough cliffs above the North Sea, as different as possible from the setting of the New Louvre. Its corner pavilions are capped, not with ordinary high mansards, but with curious roofs like pointed domes, richly crowned with elaborate cornices. In the intricacy of their silhouette these are not unworthy rivals of Barry’s Halifax tower. The massive walls are not of freestone in the manner of Paris nor yet of pallid Suffolk brick with light coloured stone or cement trim as in London. Instead, they are of warm red brick with incredibly lush decorative trim of tawny terracotta—a combination that M. D. Wyatt also used on the most elegant Second Empire mansion in London, Alford House, which stood from 1872 until 1955 in Prince’s Gate at the corner of Ennismore Gardens (Plate 83A).

Public and private architecture could hardly hope to rival the sumptuousness of the new hotels, and in Britain rarely attempted to do so. At Liverpool T. H. Wyatt in 1864-9 carried a U-shaped range of ornately pavilioned and mansarded blocks that housed the Exchange around the open space at the rear of the Town Hall, somewhat as Outshoorn carried his Galerij around the Paleis voor Volksvlijt in Amsterdam; but that is now all gone.

In the English countryside, the Bowes Museum at Barnard Castle in County Durham, built in 1869-75 by J.-A.-F.-A. Pellechet (1829-1903), and Waddesdon Manor in Buckinghamshire by another French architect, G.-H. Destailleur (1822-93), largely of 1880-3, are unique examples of extensive mansions completely in the Second Empire mode (Plate 76B). In London Montagu House, designed in 1866 by the elderly Burn for the Duke of Buccleuch, once raised in Whitehall the mansarded pavilions that Barry and the winners of the Government Offices competition had proposed in 1857, but this has now been demolished.

The most notable Second Empire ensemble in London, however, still partly survives (Plate 80B). Facing the gardens of Buckingham Palace and extending southward from the group of Late Georgian monuments around Hyde Park Corner, are the terraces of Grosvenor Place. These were designed[208] in 1867 and built in the following years. They provide one of the more striking features of the London skyline inherited from the Victorian period. Rivalling the high roofs and, almost, the tall steeples of the Victorian Gothic, the mansards over the end houses are carried to fantastic heights and capped with pointed upper roofs, providing several storeys of attics; while the centre houses have convex mansards like square domes taken straight from the New Louvre.

Below these Alpine crests, elaborated at the base with rich stone dormers, the enormous houses are all of fine Portland stone—hardly to be found in any earlier nineteenth-century London terraces except those of Ennismore Gardens—and detailed with a plausibly Parisian flair—it is even said that draughtsmen were sent to Paris to study Second Empire work at first hand. English are the porches, however, which make plain that these pretentious ranges are rows of dwellings like those in nearby Belgrave Square. English, also, are the red stone bands, novel touches echoing the fashionable ‘structural polychrome’ of the contemporary Victorian Gothic, just as the tall mansards echo its pointed roofs (see

## Chapter 10).

Beyond the first two blocks of Grosvenor Place the new construction of the sixties stops; but it starts again at the farther end and surrounds the two triangles of Grosvenor Gardens, of which Knowles’s hotel occupies part of the farther side. It is characteristic of the Parisian inspiration of the whole that on the east side of the Gardens great blocks of flats—’mansions’ in a Victorian euphemism—replaced the usual London terraces of individual tall houses, but these now serve as offices as do all the extant houses in Grosvenor Place. For one of these blocks red brick was used, but set like a mere panel-filling within stone frames according to a French rather than an English tradition.

There are no other comparably pretentious examples of Second Empire terraces in London except Cambridge Gate (1875) by Thomas Archer and A. Green (?-1904), an unhappy intrusion among Nash’s stuccoed Regent’s Park ranges despite its handsome execution in fine ashlar of Bath stone. Characteristically, London domestic architecture of the late fifties and sixties merely elaborated the Renaissance Revival formulas of the previous decade. Not only were the chosen models generally later and richer as in Vienna; wherever possible bolder plastic effects were achieved by a more extensive use of ground-storey colonnades, first-storey porches, and projecting bay windows, as on Wood’s magniloquent terraces at Lancaster Gate or those of 1858 by C. J. Richardson (1800-72) that followed them in Queen’s Gate.

The high standards of the earlier period were maintained only in business _palazzi_, not those of London’s City, but those in big Northern towns like Bradford and in Scotland. There good freestone was readily available and a certain cultural lag, as well as a regional sobriety of temperament, led to the maintenance of a more Barry-like tradition. Notable everywhere for their academic virtues are the various National Provincial Bank buildings by Barry’s pupil John Gibson (1819-92). The earliest, but not the most typical, is the head office in Bishopsgate, which was begun in 1863.

A special school of Renaissance design is associated with Sir Henry Cole’s Department of Practical Art, and this produced the various buildings that he sponsored in the new London cultural centre in Brompton (now usually called South Kensington). The Exhibition of 1862, on the southern edge of the estate belonging to the Commissioners of the Great Exhibition, was housed in a structure designed by Francis Fowke (1823-65), an army engineer. As at the Paris Exhibition of 1855, the metal and glass construction of this was masked externally with masonry walls, but, unlike Cendrier’s and Viel’s Palais de l’Industrie, the whole was pavilioned and mansarded in the Second Empire mode. A still more elaborate Second Empire project was prepared by Fowke for the Museum of Science and Art (later Victoria and Albert), Cole having evidently accepted all too abjectly the criticism of his earlier temporary structure, the notorious ‘Brompton Boilers’ (see Chapter 7). As Fowke died at this point the Museum (Plate 83B), begun in 1866, as also the associated Royal College of Science (Huxley Building), built in 1868-71, were carried out in a much less French vein under another army engineer, H. G. D. Scott (1822-83). The walling material is a fine smooth red brick, very rare in the London of the nineteenth century, beautifully laid up with thin joints. With this is combined an enormous quantity of elaborately modelled pale cream terracotta, as on various Central European buildings deriving from Schinkel’s Bauakademie in Berlin of 1831-6.

In these South Kensington structures, planned by an engineer, the emphasis is on the sculptural embellishment designed and executed by Godfrey Sykes and other artists associated with the Department. This team-work, by-passing as it did over-all control by an architect, was not very successful in achieving the coherence of Knowles’s and Giles’s hotels, although those were built for much less sophisticated clients. Much the same team, but with still more sculptors collaborating, was responsible for the Albert Hall, the vast circular auditorium built in 1867-71 on the northern edge of the Commissioners’ Estate facing the most characteristic monument of the age, G. G. Scott’s Victorian Gothic Albert Memorial. The engineer Scott’s really notable achievement here in the metal construction of the vast dome is unfortunately swamped by the profuse investiture of sculptural detail in terracotta, intrinsically elegant though much of that is.

In the sixties there was some coherence in the planning of the Commissioners’ Estate as a whole, with a garden court surrounded by a great hemicycle of terracotta arcading by M. D. Wyatt lying behind the 1862 Exhibition Building and below the Albert Hall. In Vienna the cultural edifices were admirably grouped along the Ringstrasse with plenty of open space between them, however much they may have lacked intrinsic architectural quality. In sad contrast is the way the following decades allowed this considerable tract to become clogged up until almost no urbanistic organization at all remains.

Other European countries tended in this period, like Denmark, Sweden, and Holland, to follow Paris and Vienna rather than London. Only a few works of the sixties and seventies need be singled out from the welter of pretentious public and private construction that turned Brussels, for example, into a ‘Little Paris’.[209] The Boulevard Anspach as a whole suggests the Cannebière in Marseilles, although the mansards on the buildings that line it are more plastically handled; the Exchange, in its own square half-way down the boulevard, was built by L.-P. Suys (1823-87) in 1868-73, and this provides the focus of the mid-nineteenth-century city, as does Garnier’s Opéra in Paris. A provincial variant of the Opéra in many ways, despite its quite different function, this is somewhat more academic in composition yet also rather coarser in its profuse ornamentation. Brussels as a whole is dominated, however, by one of the grandest and most original monuments erected anywhere in this period.

The Palace of Justice,[210] built by Joseph Poelaert (1817-79) in 1866-83, occupies so high a site and is mounted on so mountainous a substructure that almost the whole of its gargantuan mass is visible from all over the city. Although generically Classical, a good deal of the external treatment has an indefinable flavour of the monuments of the ancient civilizations of the East, somewhat like that of the exotic churches Alexander Thomson built in the late fifties and sixties in Glasgow (Plate 81). Even more than Thomson’s relatively small and delicately scaled work, the Palace of Justice also suggests the megalomaniac architectural dreams of such a Romantic English painter as John Martin. Heavy and almost literally cruel, it has a Piranesian spatial elaboration and a plastic vitality of the most exaggeratedly architectonic order. Thus it quite puts to shame the urbane Renaissance costuming of most Continental public architecture of this period and the usual Neo-Baroque of the next.

The existence of this extraordinary edifice in a minor European capital prepares one a little for the important part that Brussels was to play in the nineties, even though there could hardly be two architects further apart in spirit than Poelaert and Victor Horta, who initiated there the Art Nouveau (see Chapter 16). So also in Glasgow, the originality of Thomson’s Queen’s Park Church of the sixties at least opened the way for the notable international contribution to be made by the Glaswegian C. R. Mackintosh in the nineties. But it was Alphonse Balat (1818-95), not Poelaert, who was Horta’s master and also in these decades professor of architecture at the local Academy. Balat’s Musée Royale des Beaux Arts of 1875-81 already represents a reversion to a more restrained and academic classicism with none of Poelaert’s force and vitality. Yet this building is not without a certain correct elegance of detail and conventional skill in composition for which his houses of the sixties, with their Barry-like handling of the High Renaissance _palazzo_ theme, prepared the way. The real eclecticism of this period lies less significantly in the variety of nominal styles employed than in the variety of ways of employing them. It is this, rather than the concurrent multiplication of fashionable modes, that makes it so difficult to characterize broadly the production of the period between the mid century and the nineties.

In several other European countries the situation was made even more complicated than in Belgium by a very considerable cultural lag such as has already been noted in Scandinavia. While the Rütschi-Bleuler House in Zurich of 1869-70 by Theodor Geiger (1832-82) had the fashionable Second Empire mansard, here high and concave, at nearby Winterthur Semper’s Town Hall of precisely the same date, with its dominating temple portico, might at first sight be taken for a provincial French public edifice of the second quarter of the century. At the Zurich Polytechnic School, where Semper became a professor in 1855,[211] the large building begun in 1859 that he erected with the local architect Wolff is equally retardataire in style. His Observatory there of 1861-4 is a delicate and rather picturesquely composed exercise in the _quattrocento_ version of the _Rundbogenstil_, rather like his Hamburg houses of twenty years earlier.

If a German architect of established international reputation could be thus affected by the conservative tastes of his Swiss clients, it is not surprising that in the Iberian peninsula almost nothing of interest was built in this period. It may, however, be mentioned that the building for the National Library and Museums in Madrid, designed in 1866 by Francisco Jareño y Alarcón (1818-92) and almost thirty years in construction, while still of the most conventional Classical character as regards its façades, has convex mansards over the end pavilions of quite definitely Second Empire character. Characteristically, the Chamber of Commerce in Madrid, completed in 1893 by E. M. Repulles y Vargas (1845-1922), illustrates the general return of official architecture to still more conventional academic standards towards the end of the century. But in the seventies there began in Barcelona the career of a Spanish—or more accurately Catalan—architect, Antoni Gaudí, who was destined to produce around 1900 some of the boldest and most original early works of modern architecture. Gaudí’s real links in the seventies and eighties, spiritually if not so much actually, are with the High Victorian Gothic not the Second Empire, although the earliest project on which he worked reflected the Palais Longchamps at Marseilles (see Chapter 11).

The situation in the United States was naturally most like that in England. As has already been noted, a French-trained Danish architect, Lienau, prefigured the Second Empire mode in the Shiff house in New York as early as 1849-50. By the mid fifties mansards of rather modest height, often with shallow concave slopes, had appeared in Eastern cities on many houses not otherwise particularly Frenchified. Richard M. Hunt (1827-95),[212] the first American to study at the École des Beaux-Arts and actually an assistant as well as a pupil of Lefuel, returned from Paris to America in 1855. But he brought with him no lush Second Empire mode but rather the basic academic tradition of the French official world, despite the fact that he had himself worked in 1854 on the New Louvre. Although some of the earliest work of H. H. Richardson, who returned from Paris a decade later after working for several years for Labrouste’s brother Théodore, was of Second Empire character, he showed himself from the first more responsive to influences from contemporary England (see Chapters 11 and 13). On the whole, the Second Empire mode, as it was practised in America through the third quarter of the century, derived almost as completely as the local Victorian Gothic from England. Most American architects were kept informed of what was going on abroad through the English professional Press, and so they naturally followed the models that were offered in the _Builder_ and the _Building News_ rather than those in the publications of César Daly.[213]

The Civil War of 1861-5 did not bring architectural production to a stop; indeed, it seems to have had a less inhibiting effect than the aftermath of the financial crash of 1857 in the immediately preceding years. In Washington the building of Walter’s new wings of the Capitol, initiated in 1851,[214] and of his cast-iron dome, designed in 1855, continued until their completion in 1865, right through the war years at President Lincoln’s express order (Plate 82A). There is nothing specifically French about this new work at the Capitol, even though Walter had the assistance from 1855 of the Paris-trained Hunt. On the other hand, the original more-or-less Romantic Classical edifice that had finally been brought to completion in 1828 by Bulfinch after so many changes of architect was largely submerged. The new wings echo in their academic porticoes the broader portico of the original late eighteenth-century design; but the cast-iron dome (see Chapter 7), rivalling in size the largest Baroque domes of Europe, has a high drum and a Michelangelesque silhouette of the greatest boldness in contrast to the Roman saucer shape of that designed by Latrobe and not much raised in execution by Bulfinch.

It was not in Washington that the Second Empire mode was first introduced for public buildings; Washington, indeed, would never again be the centre of architectural influence that it was in the Romantic Classical period, although the new state capitols begun in the sixties and seventies were mostly capped with imitations of Walter’s dome. A ‘female seminary’ on the Hudson River, endowed by a brewer, and the new City Hall in Boston, Mass., both dating from the opening of the sixties, are the first monumental instances of the new mode that dominated the field of secular public building until the financial Panic of 1873 brought the post-war boom to a close. James Renwick,[215] who designed the very extensive Main Hall for Matthew Vassar’s new college at Arlington near Poughkeepsie, N.Y., in 1860, was specifically instructed by his client to imitate the Tuileries—not the New Louvre—and so he did in an elaborately pavilioned composition of U-shaped plan crowned by various sorts of high mansards. This overshadows in significance his earlier Charity Hospital of 1858 on Blackwell’s Island in New York, already mansarded but very plain, and his Corcoran Gallery of 1859, now the Court of Claims, in Washington, with a rich but muddled façade still rather flatly conceived.

Renwick was at least as eclectic as such Europeans as Ballu and Ferstel. Having made his first reputation with the building of the Anglican Grace Church in New York in 1843-6—if not very Camdenian, this is at least a fair specimen of revived fourteenth-century English Gothic—he continued in the Gothic line with the Catholic St Patrick’s Cathedral in New York, begun in 1859 and completed (except for the spires) in 1879. That vast two-towered pile, however, is Gothic in a very Continental way, resembling Gau’s and Ballu’s Sainte-Clotilde in Paris and Ferstel’s Votivkirche in Vienna more than anything English of the period. In the late forties Renwick had also been the agent of Robert Dale Owen’s ‘Romanesque Revival’ aspirations in designing the Smithsonian Institution in Washington (see Chapter 6).

For such things as the Smithsonian and his churches Renwick had plenty of visual documents on which to lean, either archaeological treatises on the buildings of the medieval past or illustrations of contemporary foreign work. But for Vassar College, very evidently, he was dependent for his inspiration on rather generalized lithographic or engraved views of the Tuileries. Nor could he, at this relatively early date, borrow much from published illustrations of contemporary English work in the new international Second Empire mode. The particular plastic vitality of the Americanized Second Empire is already notable in this early example, however, even though the rather crude articulation of the red brick walls is remote from anything French of any period from the sixteenth century to the nineteenth. Later buildings by Renwick in the same mode are richer and closer to Parisian standards, but their architectonic vitality is considerably less.

The Boston City Hall,[216] built by G. J. F. Bryant (1816-99) and Arthur D. Gilman (1821-82) in 1862-5, is a smaller but suaver edifice. Although it is a compactly planned block, the articulation of the walls by successive Roman-arched orders, coldly but competently executed in stone, is boldly plastic below the crowning mansards. However, just before this, for the Arlington Street Church of 1859-61, the first edifice erected in the Back Bay district that Gilman was just laying out,[217] he had turned not to France but to eighteenth-century England for inspiration, basing himself chiefly on the same churches by Gibbs that had been the most popular American models in later Colonial times.

A leading opponent of the Greek Revival, Gilman, like most Continental architects of the day, evidently knew better what he meant to leave behind than whither he wished to proceed. His Boston church initiated no national wave of Gibbsian church architecture; indeed, the sixties were the heyday of Victorian Gothic design for churches in the United States. His City Hall, on the other hand, set off a nation-wide programme of public building in the Second Empire mode; for Boston was now for a score of years the artistic as well as the intellectual headquarters of the country in succession to Philadelphia. In this programme municipalities, state authorities, and the Federal Government all

## participated actively during the decade following the Civil War. In the

case of many Federal buildings, only nominally the work of the office of the Supervising Architect, where A. B. Mullet (1834-90) succeeded Rogers in 1865, Gilman acted in these years as consultant, and was probably the real designer rather than Mullet or his assistants.

These vast monuments were mostly constructed during General Grant’s presidency. Parisian in intention, yet American in their materials, they are withal rather similar to Second Empire work in England. Few were completed before the mode went out of favour as changes in architectural control sometimes make evident. In the case of the New York State Capitol in Albany, for example, begun in 1868 by Thomas Fuller (1822-98) and his partner Augustus Laver (1834-98), both arriving from England via Canada, Eidlitz and Richardson took over jointly in 1875, modifying the design of the building very notably above the lower storeys towards the Romanesquoid. Thus it was finally brought to completion by them and others in the following twenty years. The very tall tower on the Philadelphia City Hall, begun in 1874, was finished over a decade later. This tower, whose crowning statue of William Penn still tops the local skyline, has hardly anything in common with the Louvre-like pavilions below; yet the whole is nominally the work of one architect, John McArthur, Jr (1823-90), the grandfather of General Douglas McArthur.

Undoubtedly the association of these prominent buildings with the unsavoury Grant administration and the fact that there were—at least in the two cases mentioned above—major financial scandals involved in their slow and incredibly costly construction played an important part in the early rejection of a mode so associated with the public vices of the decade after the Civil War. Not many of them are extant today other than the Boston, Albany, and Philadelphia structures just mentioned and the old State Department Building in Washington (Plate 82B).

In New York, Boston, and other large cities the vast granite piles in this mode that long served as post offices are all gone. In Chicago the Cook County Buildings built by J. J. Egan in 1872-5 have also long since been replaced. In San Francisco Fuller & Laver’s extensive group of Municipal Buildings was destroyed in the fire that followed the earthquake of 1906. This must have been the largest, the richest, and plastically the most complex production of the whole lot, with its triangular site, boldly articulated massing, and central dome.

Though threatened by every new administration, the State, War and Navy Department Building built by Mullet in 1871-5 still stands, overshadowing the nearby White House. This is perhaps the best extant example in America of the Second Empire—or as it is sometimes called locally, the ‘General Grant’—mode (Plate 82B). The tiers of Roman-arched orders in fine grey granite, borrowed by Gilman as consultant architect and presumptive designer from his earlier Boston City Hall rather than from Paris, tower up storey above storey to carry mansards of various different heights above the complex pavilioned plan. Cold and grand, almost without sculptural decoration, this could hardly be less like the New Louvre or the old Tuileries in general texture; nor is there any of the playful semi-Gothic detail of Knowles’s and Giles’s London hotels or of the festive colouring and lush ornamentation of Brodrick’s at Scarborough.

The contrast of the old State Department Building with its _pendant_ on the other side of Lafayette Square, Mills’s Grecian Treasury, finally completed by Rogers a decade earlier, is shocking to most people. Yet it is fascinating to read here the representational aspirations of an age that found its most significant expression, not in its public buildings, but in the new skyscrapers which first rose in New York at just this time, Hunt’s Tribune Building and the Western Union Building by his pupil George B. Post. Both, incidentally, were heavily mansarded, and the one by the American-trained Post was much more typically Second Empire than is the French-trained Hunt’s (see Chapter 14).

In urban domestic architecture, both on large mansions and on the more usual terrace houses, mansards became characteristic but not ubiquitous in the late fifties and remained so down to the mid seventies and even later in the West. Boston’s Back Bay district, laid out by Gilman in 1859, has a few mansions along Commonwealth Avenue that resemble somewhat the _hôtels particuliers_ of Paris, and also several mansarded terraces by Bryant & Gilman and other architects in that avenue and in Arlington and Beacon Streets. The materials used are un-Parisian—brownstone like Gilman’s nearby church or dark-red brick with brownstone trim—and the detail is rarely very plausibly French. In general, inspiration still came from London, even if nothing so extensive and spectacularly monumental as Grosvenor Place and Grosvenor Gardens was ever produced. In New York Lienau’s finest terrace, that built in Fifth Avenue between 55th and 56th Streets in 1869, was rather more sumptuous than the Boston examples, being of white marble with very literate ranges of superposed orders. Hunt’s New York work was often so authentically Parisian as quite to lack the bombast of the international Second Empire mode. Especially interesting were his Stuyvesant Flats in 18th Street, New York, of 1869-70. This block was a very early example of an apartment house of the Parisian sort in America, where they did not generally flourish much before the late eighties.

For the more characteristic free-standing houses that were built outside cities, in suburbs, in towns, and even in the country, the Second Empire mode was also very popular. Interpreted in wood, painted brown or grey stone colours, these have a distinctly autochthonous character. Generally symmetrical and tightly planned, they did not advance the development of the American house in the way of the rival ‘Stick Style’; but in their emphasis on complicated three-dimensional modelling, especially the modelling of the roofs, they prepared the way for one important aspect of the later and more original ‘Shingle Style’ (see

## Chapter 15).

The Second Empire episode in the United States is a curious one. On the one hand, it was a consciously ‘modern’ movement, deriving its prestige from contemporary Paris, not from any period of the past like the Greek, the Gothic, or even the Renaissance Revivals—of which last, of course, it was in some limited sense an heir. On the other hand, the considerable originality of the mode as it was actually employed was largely unconscious and due to the lack of accurate visual documents, or even a codified body of precedent, to be followed. At this time contemporary conditions demanded, as in Europe, the construction of many public edifices, Federal, state, and municipal, to house a complexity of functions. It would have been almost impossible to compress these within the rigid rectangles of the Greek Revival even had the Greek Revival not already been rejected by most critics twenty years or more earlier.

Yet the Second Empire episode was necessarily brief, lasting little more than a decade. The crass assurance it reflected, particularly the special arrogance of the post-war politicians in Washington, the state capitals, and in the bigger cities, was much shaken by the Panic of 1873. The mode did not therefore, as in much of Europe, continue in America into the eighties and nineties.

The episode has a longer-term significance, nevertheless. Slight as was the actual relationship to the Second Empire mode of the first two Americans to be trained at the École des Beaux Arts, Hunt and Richardson, their personal influence and their prestige encouraged a growing trek of architectural students to Paris; their recommendations alone would hardly have had much effect had not fashion already established Paris rather than London in the public mind as the centre of modern architectural achievement and inspiration. From the mid eighties on, the long-maintained dependence on England in architectural matters began to be notably weakened; for a generation and more very many American architects would seek their roots abroad, but henceforth in France, or even Italy, not England.

It is not surprising that in the British Dominions there was no such direct French influence in this period as in Latin America. Urban entities like the Colmena and its terminal square in Lima, Peru, pavilioned and mansarded throughout, rival European examples like the Søtorvet in Copenhagen or the Galerij in Amsterdam. Before they gave way to skyscrapers, the _hôtels particuliers_ along the Paseo de la Reforma in Mexico City were more numerous and more plausibly Parisian than along Commonwealth Avenue in Boston or Bellevue Avenue at Newport. But both in Canada and in Australia the Second Empire mode arrived from England late and in a more corrupted form than in America. The mansarded Windsor Hotel of 1878 in Montreal hardly rivalled the Palmer House of 1872 in Chicago by J. M. Van Osdel (1811-91), to which the rich merchant Potter Palmer was as proud to give his name as to the incredible fake castle that he built for his own occupancy a decade later. The Princess Theatre in Melbourne, Australia, built by William Pitt in 1877, with its three square-domed mansards, has an appealing nonchalance, like that of the contemporary edifices of the mining towns high in the American Rocky Mountains—the hotel in Virginia City, Nevada, that has been mentioned earlier, or the much more modest Opera House in Central City, Colorado, for example. But the public architecture of the third quarter of the century in Australia was more restrained in design just because it was generally so very retardataire.

The Parliament House in Melbourne, begun in 1856 by John G. Knight (1824-92) and completed in 1880 by Peter Kerr (1820-1912), has academic virtues not unworthy of Kerr’s master Barry, though its giant colonnades recall rather those of Brodrick’s contemporary Town Hall in Leeds. The Treasury Buildings in Melbourne, by John James Clark (1838-1915) of 1857-8, are not unworthy of comparison with High Renaissance work of the period on the Continent. Other public buildings of the sixties and seventies are of more definitely Victorian character, but Early Victorian rather than High. For example, Clark’s Government House of 1872-6 in South Melbourne is a towered Italian Villa consciously modelled on Queen Victoria’s Osborne House of a generation earlier. Both in Australia and in Canada the Victorian Gothic had more vitality in this period (see Chapter 11).

There is little profit in pursuing farther in the outlying areas of the western world evidence of direct influence from Paris (of which there is, for example, some in Russia) or autochthonous variants of the Second Empire mode. In this generally rather unrewarding period the best work mostly falls under the High Victorian Gothic rubric, or else it illustrates specifically the development of commercial and domestic architecture in the Anglo-American world (see Chapters 10 and 11; 14 and 15). In an attempt to give an over-all picture too many buildings of low intrinsic quality and little present-day interest have already been cited.

What makes especially difficult the proper historical assessment of the widespread influence of Paris in the decades following 1850 is that this influence, whether direct or indirect, rarely produced buildings on the Continent of real distinction or even of much vitality. Only in England and the United States, where the mode was quite reshaped by a different cultural situation and the bold use of local materials, is it of much independent interest. The more plausibly Parisian the work outside France, the less vigour it usually possesses. Some of it can be very plausible indeed, as for example the street architecture of Mexico City and Buenos Aires, even if what appears to be carved French limestone in the Argentine capital is usually but a triumph of imitative craftsmanship on the part of stucco-workers imported from Italy. In general, Mexican and Argentine Second Empire is very dull, as dull as in Belgium, say, with no Poelaerts to redress the balance. Yet along the Malecón in Havana, Cuba, where the traditional galleried house-fronts were reinterpreted in a generically Second Empire way with Andalusian lushness, the results are much more notable, not least because the soft local stone has been very richly weathered by the strong sea breeze. As was mentioned earlier, the use of _azulejos_ in extraordinary tones of brilliant green and purple gives autochthonous character to similar work in Brazil.

The international Second Empire mode has so far found no historian or even a sympathetic critic. Perhaps no other mode so widespread in its acceptance and so prolific in its production has ever received so little attention from posterity. Yet beside it the contemporary stream of the Victorian Gothic mode, which has been recurrently studied, must seem more than a little parochial and also excessively dependent on the individual capacities—not to say the caprices—of its leading practitioners. Within the areas in which the Victorian Gothic was employed, however, an area effectively confined to the Anglo-Saxon world geographically and to certain kinds of buildings typologically, it was capable of major architectural achievement. Moreover, thanks to the line of spiritual descent from the leaders of the generation of architects

## active in the third quarter of the century to those of the next, the

more creative aspects of the architecture of the turn of the century derive in not inconsiderable part from the later Victorian Gothic.

The Lefuels and Hansens, or such men as Brodrick, Poelaert, and Gilman, trained no worthy pupils. But the disciples of the Victorian Gothic leaders not only include such very able young men who actually worked in their offices as Webb and Shaw and Voysey but also, in some sense at least, so great an American architect as Richardson, whose formal training had been wholly Parisian (see Chapters 11, 12, and 13). The advance of domestic architecture in the second half of the nineteenth century and, to a somewhat lesser extent, also that of commercial architecture therefore owed a great deal to the Victorian Gothic, at least in England and America (see Chapters 14 and 15).

------------------------------------------------------------------------

## CHAPTER 10

HIGH VICTORIAN GOTHIC IN ENGLAND

BY 1850 Neo-Gothic was accepted as a proper mode for churches throughout the western world. Only in England, however, had it become dominant for such use. Moreover, Gothic was a more than acceptable alternative there to Greek or Renaissance or Jacobethan design for many other sorts of buildings also. Only in the urban fields of commercial construction and of terrace-housing was its employment still very rare. On the Continent the nearest equivalent in popularity and ubiquity to the Victorian Gothic was the German _Rundbogenstil_. Neo-Gothic, although used more and more everywhere after 1850 for churches, attracted few architectural talents of a high order (see Chapter 11).

There are several reasons why the Gothic Revival was able in England, and almost only in England, to pass into a new and creative phase around 1850. One was certainly the ethical emphasis of its doctrines, an emphasis more sympathetic to Victorians than to most Europeans of this period, but not without its appeal on the Continent towards the end of the century. Another reason was the informality, not to say the amateurishness, of architectural education in Britain, encouraging personal discipleship and the cultivation of individual expression rather than providing for the continuation of an academic tradition.

Related to this is the private character of architectural practice in England as compared to its more public responsibilities and controls on the Continent. The desirable professional positions in France, and to almost the same degree in many other European countries, were those offered by the sovereign or the State. But after the time of Soane and Nash official employment ceased to carry either prestige or opportunity in England, the Houses of Parliament notwithstanding—it was not Barry’s work there but his clubs and mansions that established his high professional reputation. As in the eighteenth century, a social and aesthetic _élite_ still provided both critical esteem and the most desirable commissions for Victorian architects; by 1850 a large part of that _élite_ was very church-minded and thoroughly Gothicized. Not until the mid sixties was there any significant change; even then those responsible for this change, both the architects and their patrons, had all been brought up in the churchly Gothic Revival tradition.

The High Victorian Gothic opened with the building of a London church. All Saints’, Margaret Street, designed in 1849, largely completed externally by 1852, and consecrated in 1859, was the result of no imperial fiat, like the Votivkirche in Vienna or the big churches of the sixties in Paris, nor did it occupy like them an isolated site approached by wide new boulevards. Intended as a ‘model’ church by its sponsors, the Ecclesiological Society, and financed by private individuals, All Saints’ is set in a minor West End street at the rear of a restricted court flanked by a clergy house and a school (Plate 6A). But for its tower, the tallest feature of the mid-century London skyline, it would have been hard to find; but once found, it could never be ignored.

The architect of All Saints’, Butterfield, had been for some years, together with Carpenter, the favourite of the ecclesiologists because of the Pugin-like ‘correctness’ of his revived fourteenth-century English Gothic. Now, quite suddenly, he and his sponsors embarked on new paths. As soon as the walls began to rise, their startling character became apparent; for the church is of red brick, a material long out of use in London, and that red brick is banded and patterned with black brick, a theme varied on the tower by the insertion of broad bands of stone. ‘Permanent polychrome’, achieved with a variety of materials, thus made its debut here. In the interior, moreover, the polychromatic effect was even richer and more strident, with marquetry of marble and tile in the spandrels of the nave arcade and over the chancel arch, not to speak of onyx and gilding in the chancel itself (Plate 85). The very exiguous site forced any expansion upwards; the nave is tall, the vaulted chancel taller, and the subsidiary structures flanking the court are even higher and narrower in their proportions.

While the construction of All Saints’ proceeded there was much concurrent and complementary activity in the English architectural world. In 1849 a young critic, John Ruskin (1819-1900), had brought out an influential book, _The Seven Lamps of Architecture_, in which many of the recommendations ran parallel to, if indeed they did not influence, Butterfield’s latest stylistic innovations. Notably, Ruskin urged the study of Italian Gothic: if All Saints’ is, in fact, not specifically Italian in the character of its polychromy, it seemed so to most contemporaries. The real foreign influences here, as in the profile of the fine plain steeple, are German if anything. Butterfield’s moulded detail continued to follow quite closely English fourteenth-century models.[218]

In this same year 1849 Wild[219] was building on an even more obscure London site in Soho his St Martin’s Northern Schools with pointed arcades of brick definitely derived from Italian models. Moreover, he was being acclaimed for doing this by the very ecclesiological leaders who had ten years before condemned his Christ Church, Streatham, as ‘Saracenic’. With the publication of the first volume of Ruskin’s next book, _The Stones of Venice_, in 1851 (the two less important later volumes came out in 1853) and the appearance of _Brick and Marble Architecture of the Middle Ages in Italy_ by G. E. Street (1824-81) in 1855, Italian influence increased. Street’s name, moreover, introduces the third of the three men most responsible for the sharp turn that English architecture was taking in the fifties.

Without depending on polychromy, Butterfield designed in 1850 and built in 1851-2 St Matthias’s off Howard Road in Stoke Newington, a London suburb, another church of novel character. Unconfined by a closed-in urban site, this also showed in its great scale and the bold silhouette of the gable-roofed tower—still standing today above the bombed ruin of the church—how the timid Early Victorian Gothic of the forties could be invigorated. Moreover, at St Bartholomew’s at Yealmpton in Devonshire, built in 1850, Butterfield introduced in a country church striped piers of two different tones of marble and considerable coloured marquetry work. A former fellow assistant of Street in G. G. Scott’s office, William H. White (1826-90), at All Saints’ in Talbot Road, Kensington, in London, begun in 1850, also used the new polychromy that soon became the principal, though by no means the only, hallmark of High Victorian Gothic.

A large country house of stone by S. S. Teulon (1812-73), Tortworth Court in Gloucestershire, built in 1849-53, has no polychromy, although its architect was soon to be the most unrestrained of all in its exploitation. His patrons, moreover, would be notably ‘lower’ in their churchmanship than the members of the Ecclesiological Society who employed Butterfield. But in the boldly plastic massing of Tortworth, leading up to a tall central tower of the most complex silhouette, Teulon exemplified the new architectural ambitions, ambitions that would soon be finding as striking expression in secular work as in ecclesiastical building whether ‘high’ or ‘low’.

Street had been a favourite of the High Church party since he first began building small churches and schools of a most ‘correct’ sort in Cornwall on leaving Scott’s office.[220] He was also the author of several critical articles published in _The Ecclesiologist_, notable for their cogency. In these he commented, for example, on the applicability of the arcades of Wild’s school to commercial building; he also attacked the curious habit of the forties, most prevalent with the ecclesiologists, of designing urban churches on confined sites as if they were to sprawl over ample village greens. Street began his first important church with associated school buildings, All Saints’, Boyn Hill, at Maidenhead, in 1853. Here he employed red brick and almost as much permanent polychrome as Butterfield at All Saints’, Margaret Street. He also handled the detail, particularly on the schools, with something of the same sort of brutal ‘realism’ (to use the catchword of the period) that Butterfield used on his subsidiary buildings.

In the same year in London Street’s former employer Scott, long established as the most successful, if hardly the most ‘correct’, of Early Victorian Gothic practitioners, and since 1849 Architect to Westminster Abbey, built in Broad Sanctuary contiguous to the façade of the Abbey a Gothic terrace. That the use of Gothic should have been encouraged here by the Abbey authorities is not surprising. But they themselves may well have been surprised at what their architect produced; for this is no flat range of Neo-Tudor fronts in stock brick, but a plastic mass of stonework bristling with oriels and turrets and capped with a broken skyline of stepped gables. Nothing here recalls the rather French thirteenth-century Gothic of the Abbey itself; instead the effect is Germanic, recalling the medieval houses of the Hansa cities. The work was executed with a boldness of detail doubtless less personal in character than Butterfield’s or Street’s, but quite as striking to the casual observer.

Scott’s houses had little influence, however. Gothic terraces were no more popular in the fifties and sixties in England than in the preceding decades. In residential districts the flood of more-or-less Renaissance stucco continued to spread, little affected by the High Victorian Gothic. As we have seen, the Second Empire mode also had only a very limited success in this field of construction, a field dominated not by architects but by builders.

In 1853 also Scott provided for the Camden Church in the Peckham Road in South London—Ruskin’s own family church—a new east end in a round-arched and banded medievalizing mode; Ruskin himself collaborated on the window design, or so it is said. There is sufficient Gothic ‘realism’ in the detail here to justify considering this a round-arched variant of the High Victorian Gothic; but it is definitely of Italian inspiration. It seems also to be related to the later _Rundbogenstil_ of this decade in Germany and Austria; nor is it altogether without resemblance to such a contemporary French church as Vaudoyer’s Byzantinesque cathedral of Marseilles.

Several far more important and better publicized interventions in architecture on the part of Ruskin followed immediately. In considerable part because of his personal influence with Oxford friends, the Gothic design of the Irish architects Sir Thomas Deane[221] (1792-1871) and Benjamin Woodward (1815-61) was accepted for the University Museum at Oxford in 1855. Woodward had already proved himself a would-be Ruskinian in detailing their design of 1853 for the Museum of Trinity College, Dublin, in a Venetian (though largely _quattrocento_) way. As the Oxford Museum rose to completion in the next four years, Ruskin was in continuous contact with Woodward, providing himself the design for at least one window as well as encouraging the delegation to the Irish carvers of much of the responsibility for the ornamental decoration—of which only a small part was, in fact, ever executed. The work of the O’Sheas is better appreciated in Dublin, where the decoration both of the Trinity College building and of the Kildare Street Club of 1861 was carried out by them in a very free and yet boldly naturalistic vein.

The most interesting feature of the University Museum—and one that it is surprising to find Ruskin, who hated iron and all it stood for in the nineteenth-century world, involved with—is the court, with its roof of iron and glass (Plate 86B). How different this is, however, from what iron-founders without architectural control were providing at the same time in the Brompton Boilers! Yet it is even more different from Hopper’s or Rickman’s iron Gothic of fifty years earlier (Plate 60B). For all the elaboration of the ornament, which is very metallic in character but also very aware of Early Gothic precedent, what is most notable is the highly articulated character of the structure, as if the architects had asked themselves: ‘How would medieval builders have used structural iron had it been readily available to them?’ Is this, perhaps, the first echo in England of the theories of Viollet-le-Duc, the French architect who was to exercise an international influence equal to Ruskin’s over the next generation? Probably not, as his own enthusiasm for iron began only rather later (see Chapter 16). Whether or not there is specific influence from Viollet-le-Duc here, his great archaeological publication, the _Dictionnaire raisonné_,[222] had begun to appear the year before. Very soon the structural expressiveness of ‘Early French’ detailing, studied by English architects at first hand as well as in the woodcuts of the _Dictionnaire_, began to supplant Italian polychromy as the hallmark of advanced fashion in the higher aesthetic circles.

A more modest Oxford building by Deane & Woodward, the Union Debating Hall of 1856-7, has more vigour on the whole than does the Museum,

## particularly in its characteristically notched brick detailing. It also

has the advantage of murals by the young Pre-Raphaelites. One of these, who had just left Street’s architectural office to turn briefly to painting, was William Morris (1834-96).[223] His ceiling here initiated the most distinguished career of architectural decoration of the second half of the century. Morris as a critical writer was destined, moreover, to be at least as influential on later architecture as Ruskin or Viollet-le-Duc.

Of the same date, 1856, is perhaps the most successful of Butterfield’s extant churches, that at Baldersby St James near Beverley in Yorkshire, with its contiguous group of vicarage, schools, and cottages. All of stone externally, the polychromy here is rather a sort of ‘poly-texture’ most effectively handled in the banding of the tall pyramidal spire above the plain square tower (Plate 87). Internally a delicate harmony of pink and grey-blue bricks, with accents of creamy stone, replaces the acid chords of All Saints’ in London, a harmony rivalled in the Welsh church of St Augustine’s at Penarth near Cardiff built a decade later in 1866. At the same time, Teulon at St Andrew’s in Coin Street off Stamford Street south of the Thames in London was using the boldest of brick-and-stone banding externally and, inside, elaborate patterns of light-coloured brickwork. Moreover, the rather Germanic planning of this church, demolished since the Second World War, was highly unorthodox by ecclesiological standards. Already it was evident that within the High Victorian Gothic there were to be two streams, one High Church in its patronage and led by architects of considerable learning and sophistication like Butterfield and Street, another more characteristically Low Church and often quite secular; this was generally coarser and more philistine, not to say outright illiterate.

Yet not all the best work of the High Church architects was ecclesiastical. By 1857 J. L. Pearson (1817-97) had already built some respectable if not very interesting churches distinguished chiefly by their very fine spires; but his first work of positive High Victorian character was Quar Wood, a country house he built in Gloucestershire in that year. The skilful asymmetrical massing around the stair tower here, the plastic variety provided by several different types of steep roofs, the crisp precision of the detailing, all combine to produce a modest mansion that is as different in effect from Teulon’s mountainous Tortworth as both are characteristic of the beginnings of the High Victorian Gothic.

Two houses begun soon after Quar Wood, both within the broad frame of reference of the maturing High Victorian Gothic, could hardly differ more from one another. In remodelling Eatington Park in Warwickshire in 1858 John Prichard (1818-86) attempted to mask an underlying Georgian mansion with a profusion of bold innovations in the detailing. Stone polychromy, applied sculpture, bold plastic membering of wall, roof, and chimneys, all are used here more abundantly than ever before. The Red House at Bexley Heath in Kent, on the other hand, which Philip Webb (1831-1915), who had been a fellow pupil with Morris in Street’s office, built for Morris in 1859-60, is notable for its extreme simplicity. So also is the house now known as Benfleet Hall that he built in 1861 at Cobham in Surrey for Spencer Stanhope, another of the young artists who had collaborated on the murals of the Oxford Union. This has a rather better plan than the Red House.

These houses have no external polychromy, only plain red brick beautifully laid; there is no sculptured detail at all; and the few breaks in the loose massing of the walls and roof are closely related to the informal ease of the rather novel plans. Only the high roofs of red tile are similar to those of Pearson’s Quar Wood. But in the plain, very ‘real’, detailing and the segmental-headed white-painted window-sash of an early eighteenth-century sort, set under pointed relieving arches, the relationship is close to the secular work of somewhat older men—to Butterfield’s vicarages of the forties (Plate 122B) and more notably to his clergy house and school at All Saints’, Margaret Street (Plate 86A). Webb had himself worked on some of the latest of the rather similar vicarages and schools that Street had been building for a decade. His first big country house, Arisaig, built of local stone in the remote Scottish Highlands forty miles beyond Fort William in Inverness-shire beginning in 1863, may properly be considered High Victorian Gothic also (Figure 23). It is especially interesting, like Benfleet Hall, for its plan (see Chapter 15).

Down to about 1860 the development of the High Victorian Gothic was on the whole convergent. Henceforth, ambitious young architects tried harder to have personal modes of their own like Butterfield; yet, conversely, many formed loose stylistic alliances in which individual expression became merged in some sort of group expression. The boldest and the most unruly were no longer likely to be of the High Church party, but rather of the Low. St Simon Zelotes of 1859 in Moore Street in London by Joseph Peacock (1821-93) hardly compares with the work of Butterfield and Street in distinction; but its internal polychromy of white and black brick outbids that of their best London churches, also built at the end of this decade.

Butterfield’s St Alban’s in Baldwin’s Gardens off Holborn in London, erected 1858-61, is all rebuilt now. But something of its splendidly tall proportions, if not the rich brick and tile marquetry of the wall over the chancel arch, can still be apprehended. The contrast in quality with Peacock’s work was once amazing. Street’s St James the Less in Thorndike Street off the Vauxhall Bridge Road in London also of 1858-61, is less fine but still much superior to Peacock’s work (Plate 94B). The tall square tower, set apart like a campanile, has a curiously gawky roof based on French models and the interior is somewhat cavernous. But in the richness of its red and black brick patterns, used both inside and out, and in the naturalistic carving of the nave capitals this church of Street’s rivals Butterfield’s All Saints’ and St Alban’s and is, unlike the latter, still completely intact.

Various younger men of Webb’s generation were beginning to make important contributions in church design also. G. F. Bodley (1827-1907), trained in his kinsman Scott’s office rather than in Street’s, built St Michael’s, Brighton, in 1859-62. This must have been very striking for the boldness of its scale and for the vigour of its structural expression before it was overshadowed by the tall later nave beside it added by William Burges (1827-81).[224] But it is not the parody of ‘Early French’ detailing in the square archivolts and spreading capitals of the nave arcade, so soon to be abjured by Bodley, that is significant here but the fact that this was the first church to receive an over-all decorative treatment, including stained glass, at the hands of Morris and his associates, who included the painters Ford Madox Brown and Edward Burne-Jones.

There is still finer glass of this period designed by Burne-Jones in the east window of Waltham Abbey in Essex, where the rear wall was rebuilt in the heaviest ‘Early French’ taste by Burges in 1860-1. As a painter Burne-Jones is hardly to be compared with Ingres; yet as a designer of stained glass the superiority of such early windows of his as these at Waltham Abbey to the ones by Ingres at Dreux and at Neuilly is amazing. It is not the least claim to distinction of the High Victorian Gothic that it nurtured this brilliant revival of decorative art led by Morris. Many churches of the sixties and seventies are worth visiting solely for their windows by Morris, Brown, and Burne-Jones to which there are apparently no worthy Continental parallels.

A quite different sort of contemporary church is White’s Holy Saviour, Aberdeen Park, in London, of 1859. Externally this is quiet and rather shapeless; but inside the red brick of the exterior gives way to a subtle harmony of patterned brickwork in beiges, browns, and mauves—assisted in the chancel by some additional decorative painting—that is unequalled in High Victorian polychromy. Also rather different from standard High Church Anglican work of the day is the Catholic church of St Peter in Leamington of 1861-5 (Plate 89A) by Henry Clutton (1819-93). He had won the competition for Lille Cathedral in France in 1855 with a design prepared in collaboration with Burges, but was not allowed to supervise the construction because he was a Protestant; English Roman Catholics were not so bigoted. Internally the characteristic articulation of Puginian planning was given up; nave and apse form one continuous vessel, almost basilican in effect, under a barrel roof that ends in a half dome. Unfortunately, the painted decoration of the walls and the ceiling here has all been destroyed; the effect must once have been much less barren than it is today. Externally, plain red brick is most happily combined with stone trim treated with great simplicity and yet with extreme subtlety. The inspiration is Early French, perhaps influenced by Viollet-le-Duc,[225] although Clutton knew old French work at first hand; but the smooth concavities and the delicately varied chamfers are handled with the greatest originality and justness of scaling. The fine tower, at once sturdy in its detailing and svelte in its shape, has lost the original pyramidal roof.

Not unworthy of the church, and vastly superior to Clutton’s rather dull country houses, is the contiguous rectory here, a rectangle in plan with the long gable broken only by elegantly chamfered pairs of brick chimneys (Plate 89A). The expanses of plain brick wall are regularly but not symmetrically pierced by coupled windows divided by colonnette mullions of stone. In simplicity of massing this rectory surpassed the Red House and Webb’s other—and in some ways better—early house for Spencer Stanhope, Benfleet Hall. In their simple dignity such things contrast sharply with the more ambitious secular work of the day, by this time reaching peaks of elaboration almost exceeding Prichard’s Eatington Park.

Teulon’s Elvethan Park in Hampshire of 1861, for example, is perhaps the wildest of all High Victorian Gothic houses; this mansion is so complex in composition and so varied in its detailing that it quite defies description. Polychromy runs riot, forms of the most various but undefinable Gothic provenience merge into one another, and the result seems almost to illustrate that original mode of design which Thomas Harris (1830-1900)[226] had just christened ‘Victorian’ in describing a project he published in 1860 for a terrace of houses at Harrow.

However, several churches of the mid sixties rivalled Elvethan Hall, if not Harris’s ‘Victorian Terrace’. There was, for example, Teulon’s own St Thomas’s, Wrotham Road, of 1864, piling up to its heavy central tower among the railway yards of Camden Town in London; and there was also his much more peculiar St Paul’s, Avenue Road, also of 1864, in the approaches to Hampstead. This was purged early of its original internal decoration but it long remained externally an almost unrecognizable variant of the standard Victorian Gothic church. Both have been demolished since the war. At St Mary’s in the London suburb of Ealing, built in 1866-73, Teulon used iron columns for the nave arcade; a still wilder Low Church architect, Bassett Keeling (1836-86), did the same in two London churches, St Mark’s in St Mark’s Road, Notting Dale, and St George’s on Campden Hill (where they have since been replaced), both begun in 1864. Nor were Teulon and Keeling by any means the only architects to revive the use of iron columns in the sixties; even Burges introduced them once in a church, St Faith’s at Stoke Newington, now largely demolished, and also in his Speech Room at Harrow School of 1872.

Of a quite different order is another London church, St Martin’s in Vicars Road, Gospel Oak, also begun in 1864. This is by E. B. Lamb (1805-69), an architect who had already begun to show rather High Victorian tendencies in the thirties. There is no polychromy here, and the inspiration from the past is neither Italian nor French but the still heterodox English Perpendicular. The massive plasticity of Lamb’s personal mode, with much large-scale chamfering and a consistent use of segmental-pointed arches in several orders, is happier where it was exploited more simply on the nearby rectory. The interior of his church, which has a sort of central plan with wide transepts and only a slightly prolonged nave, is a forest of timber-work ingeniously bracketed and intersected in a fashion peculiar to Lamb. Only perhaps in an international context, in relation to the contemporary American ‘Stick Style’, is this sort of structural articulation intelligible (see

## Chapter 15). But the solid, compactly planned, and simply detailed

rectory has virtues not unworthy of comparison with Clutton’s at Leamington, if not perhaps with Webb’s more delicately scaled and functionally articulated early houses.

Two churches by Street, St John’s at Torquay of 1861-71[227] and St Philip and St James’s at Oxford, which was completed in 1862, are more standard products of the early sixties. The former is notable for the very rich marble polychromy in the chancel and the full complement of windows by Morris and Burne-Jones; the latter is more ‘Early French’ with a tall tower rising in front of the polygonal apse and a curiously unorthodox but effectively ‘real’ way of running the nave arches into the east wall with no imposts at all. This device was repeated at All Saints’, Clifton, now a ruin, where the variety of colours of the fine local stones—orange and blue Pennant and cream Bath—permitted a more truly structural polychromy than usual and one of remarkable tonal harmony and elegance. All Saints’ was begun in 1863.

Both Burges and Pearson erected distinguished churches at this time, Burges in Ireland, Pearson in London. St Finbar’s Church of Ireland Cathedral in Cork, designed in 1863 for a competition and built in 1865-76, is of unusual size for a British church of this period and, what is more unusual for a nineteenth-century cathedral, it was completed without serious modification of the original project. Provided with a fine open site and a full complement of towers, two flanking the west front and a taller one over the crossing, this rivals in elaboration the big Continental Gothic churches of the period (see

## Chapter 11). Moreover, the detailing is of a distinctly French

twelfth-century order with very few eclectic or Italianate touches, thus recalling the winning design for Lille Cathedral that he had prepared with Clutton in 1855. Yet the contrast with contemporary Continental Gothic—especially with Lille Cathedral as finally executed by others—is almost as great as in the case of the rather more original English churches of this period by Butterfield or Street.

In the interior of St Finbar’s Burges developed the theme of articulation, a theme more characteristically Early English than ‘Early French’, with remarkable plastic vigour, while the handsome wooden roof, so rare a feature in medieval France, lends to the whole an unmistakably Victorian air. Less subtle, less aesthetic, than other churches of the sixties by younger men, St Finbar’s has the sort of athletic strength that is characteristic of much High Victorian Gothic, expressed in unusually literate, not to say archaeological, terms.

Burges’s church opened the road again towards a more ‘correct’ imitation of the medieval High Gothic, a road along which Pearson soon proceeded more rapidly and more doggedly than he. Yet Pearson’s own South London church of 1863-5, St Peter’s in Kennington Lane, Vauxhall, is more typically High Victorian than St Finbar’s. The carved capitals and the heavy scale of the stone detail are rather ‘Early French’. But walls and vaults are of London stock brick and there is some polychromy of the quieter, less Butterfieldian, sort resembling a little White’s at St Saviour’s. The continuity of the chancel and rounded apse with the nave echoes the ‘unified space’ of Clutton’s Leamington interior. Puginian articulation of plan and mass was henceforth somewhat out of date.

The Albert Memorial[228] in Hyde Park in London is a monument generally—and not unjustly—considered the perfect symbol of this High Victorian period, more perfect than the Houses of Parliament (in the early sixties at last approaching completion) were of the previous Early Victorian period. In 1861 Queen Victoria’s beloved husband, the Prince Consort, died. In the competition for a national memorial to rise in Hyde Park near the site of the Crystal Palace, held the next year, G. G. Scott almost inevitably won first place. Construction of the Albert Memorial began in 1863 and took nearly ten years. By the time it was completed in 1872 critics of advanced taste were already condemning it, yet it represents precisely what Scott most liked to do and what he undoubtedly did best—in his own words, this his ‘most prominent work’ represented his ‘highest and most enthusiastic efforts’. It is, moreover, an epitome of the aspirations[229] that were most widely held when it was designed (Plate 90).

The contrast between this elaborate shrine and Scott’s modest and essentially archaeological Martyrs’ Memorial of 1841 at Oxford is very great—what a long distance the English Gothic Revival had travelled in a score of years! Among Early Victorian memorials the Prince Consort’s cenotaph is rather more like Kemp’s Scott Monument in Edinburgh (Plate 51) than like the Oxford one. But where Kemp’s is soft and monochrome, this is hard and almost kaleidoscopically polychromatic. Scott’s theme is still that of the fourteenth-century English Eleanor Crosses, as is certainly appropriate for a monument to a Royal spouse; but the inspiration came in the main from relatively small reliquaries and other medieval works executed in metal and embellished with enamels and semi-precious stones.

The Martyrs’ Memorial was purely English, the specific precedents for the Albert Memorial mostly Continental: Italian, French, German, and Flemish. The materials are cold and shining, polished granites, marbles, and serpentines of various colours; and much of the detail is executed in gun-metal left plain or gilded. A profusion of white marble sculpture at various scales leads up to the seated bronze figure of the Prince by J. H. Foley, finally installed in 1876, over which is a vaulted canopy of brilliantly coloured glass mosaic. Enamels, cabochons of marble or serpentine, and intricately crisp detail of the most metallic character carry out Scott’s basic idea of a ciborium enlarged, like Bernini’s in St Peter’s, to fully architectural scale.

Beside the Albert Memorial most of Scott’s other work of this period lacks interest. His churches, particularly, are likely to be dull and respectable, reflecting the new eclectic tastes of the day only in a rather inconspicuous way. His Exeter College Chapel at Oxford of 1856-8 is a sort of Sainte-Chapelle; St John’s College Chapel at Cambridge of 1863-9 is equally monumental but somewhat less French in character and also more original in its proportions. His secular work at Oxford and Cambridge is also dull, lacking the Ruskinian touches that give a certain vitality to the Meadow Buildings built for Christ Church in 1863 by Sir Thomas Deane and his son Thomas Newenham Deane (1828-99).

Far finer, however, is their Kildare Street Club in Dublin, facing the Trinity College Museum across an expanse of lawn; for this continues the best Ruskinian tradition of the work that they did earlier with Woodward.[230]

A very striking example of the Gothic of the early sixties in England, superior to anything at Oxford or Cambridge, is the Merchant Seamen’s Orphan Asylum of 1861 by G. Somers Clark (1825-82), now the Wanstead Hospital, in a suburb north-east of London. This is actually more what is supposed to be ‘Ruskinian’, because of its Venetian detailing, than the very original Dublin clubhouse with its consistent theme of segmental arches and its bold naturalistic carving; but, like that, the Wanstead building is generically High Victorian in the asymmetrical massing, the strong colours of the black-banded red brickwork, and the surprising richness of the decoration Clark lavished on a utilitarian structure.

In the early sixties several younger men, most of them trained in Street’s office, were already turning away from the stridency of the work of the High Victorian leaders towards a simpler and suaver mode. Webb’s houses of this period have been mentioned, and will be again (see

## Chapter 15). Here the plain row of small London shops that he built at

91-101 Worship Street, Finsbury, in 1861 might be described. In them the material is not even red brick, but London stocks excellently laid. Almost nothing is overtly Gothic, yet a sense of medieval craftsmanship controls the handling of both the wide shop-windows below and the sash-windows in the upper storeys. Above all, the general composition is quiet and regular, more like Clutton’s Leamington rectory than the asymmetrical articulation that is characteristic of Webb’s own houses of these years.

A similar quietness controls the design of the wing that W. Eden Nesfield (1835-88), son of Barry’s collaborator on Italian gardens, William A. Nesfield (1793-1881), and a pupil not of Street but of Burn and Salvin, was adding to the Earl of Craven’s seat, Combe Abbey in Warwickshire, beginning in 1863. This was Nesfield’s earliest work. Despite his own studies of French Gothic,[231] which he had published the previous year with a dedication to Lord Craven, and the tracings he is supposed to have made from the illustrations of Gothic detail in Viollet-le-Duc’s _Dictionnaire_, the arches at Combe Abbey are round, not pointed, and the major architectural theme is the English late medieval ‘window-wall’ of many lights divided by stone mullions and transoms.

In a completely new house, Cloverley Hall, that Nesfield began in 1865 together with his partner Richard Norman Shaw (1831-1912), the great window-bays and the other ranges of stone-mullioned windows in the beautifully laid salmon-pink brick walls were even more the principal theme of the design. But in the decorations, delicate in scale and elegant in craftsmanship, a new sort of eclecticism made its appearance. Basically the house derives from those manor houses of the sixteenth century that were uninfluenced by Renaissance ideas; but in the detailing of Cloverley there were Japanese motifs, notably the sunflower disks that Nesfield called his ‘pies’, reflecting the new interest in oriental art that such painters as Whistler and Rossetti were taking. Except for its relatively early date, Cloverley Hall has no place in a discussion of High Victorian Gothic, for it is characteristically Late Victorian (see Chapter 15).

Nesfield’s partner Shaw, however, built in the sixties two churches that were still High Victorian in style, one in Yorkshire, the other at Lyons in France. Holy Trinity at Bingley of 1866-7 is one of the finest examples of the ‘Early French’ phase of the Victorian Gothic (Plate 94A). Externally it builds up to a very tall central tower, superbly proportioned and very simply detailed, that more than rivals in quality Street’s at Oxford. Internally the fine random-ashlar stonework—there is no polychromy—the very bold and structural detailing of the square archivolts and the simply carved capitals illustrate even better than does Webb’s domestic work in brick the new and more sophisticated attitude towards the building crafts. The principles involved go back to Pugin; but now for the first time in Webb’s and Nesfield’s and Shaw’s work of the sixties one senses a real respect, at once intelligent and intuitive, for the differing nature of different materials. Such a respect would continue to give special virtue to the work of the most distinguished English and American architects of the late nineteenth and early twentieth centuries (see Chapters 12, 13, 15, and 19).

The Lyons church, which Shaw began in 1868, is perhaps the finest of the many Victorian churches built on the Continent for local English colonies, but very different indeed from that at Bingley. A city church set between tall blocks of flats, this is also very tall in its proportions and has a more urban character than that of the Yorkshire church. French freestone does not lend itself to the particular type of semi-rustic craftsmanship that was now rising to favour with the younger English architects; hence the Lyons church is less significant than the Bingley one in that respect. But Shaw was not primarily a church architect, nor did he long remain a High Victorian (see Chapter 12).

More characteristic of the various new directions that the Victorian Gothic was taking in the mid sixties, directions that soon also led quite away from the High Victorian, are two new churches both designed well before Shaw’s at Bingley and Lyons were begun. At All Saints’ in Jesus Lane, Cambridge, begun in 1863, the spikiness of the Italianizing Victorian Gothic and the rugged structuralism of the ‘Early French’—rarely carried farther than in Bodley’s own early work—gave way to something much more English in inspiration. There is, for example, a very deep chancel and only one aisle, not to speak of a battlemented tower at one side, out of which rises a small stone spire. In fact, Bodley returned here to the fourteenth-century Decorated models preferred by Pugin, some so ‘late’ as to suggest the still forbidden Perpendicular.

Bodley now made even more use of the decorative talents of Morris and his associates than at St Michael’s, Brighton. His St Martin’s-on-the-Cliff, Scarborough, completed in 1863, is a finer church than either St Michael’s or All Saints’. Falling between them in style as well as in date, this has less historical importance, but it also was richly decorated by the Morris firm. At All Saints’ painted polychromy, but of a rather subtle order much superior to most of that of the forties, entirely replaced permanent polychrome. The brocade patterns stencilled on the walls seem almost to be designs of Pugin strengthened in their outlines and their colours by Morris. Although Bodley’s mature career as one of the two principal Late Victorian church architects did not really get under way until 1870, Victorian Gothic was evidently coming full circle at All Saints’, and the High Victorian phase was nearly over.

The other important new church of this period, St Saviour’s, Penn Street, in the Hoxton district of the East End of London, was begun in 1865 by James Brooks (1825-1901). Unfortunately this was very badly damaged in the blitz, and has since been demolished. St Saviour’s was of brick and included some polychromy like Brooks’s slightly earlier East End church, St Michael’s in Mark Street, Shoreditch, of 1863-5. But what was really significant at St Saviour’s was the unified interior space, ending like Clutton’s Leamington church and Pearson’s Vauxhall church in London in a rounded apse (Plate 89B). Notable also were the Webb-like quietness of the general composition and the straightforward handling of the main structural elements. In another, happily unblitzed, church by Brooks in the East End of London, St Chad’s, Nichols Square, in Haggerston, which was begun in 1867, the same qualities can be seen in a more mature state. Moreover, the rather plain windows and the simple moulded brick trim are echoed at domestic scale on the nearby rectory.

The fine vessel of the interior of St Chad’s, with its simple nave arcade of stone, clean red-brick walls, quietly structural wooden roof over the nave, and brick-vaulted chancel, contrasts strikingly with the hectic elaboration and dramatically vertical proportions of Butterfield’s last London church of any great interest, St Augustine’s, Queen’s Gate, of 1865-71. Two churches of the late sixties outside London, All Saints’ at Babbacombe near Torquay, which was built in 1868-74, and the earlier mentioned St Augustine’s at Penarth, begun in 1866, are much more satisfactory examples of Butterfield’s middle period.

Brooks continued through the seventies to develop the implications of his East End churches with great success. The largest and most notable is that of the Ascension, Lavender Hill, in Battersea, which was begun in 1873 and completed by J. T. Micklethwaite[232] (1843-1906), a former assistant of G. G. Scott, in 1883. The vast lancet-pierced red-brick hull of this church is one of the landmarks of the South London skyline; the interior, which is perhaps a little bare, has nevertheless a monumentality of scale rare in English churches of any period. However, this monumentality is rivalled both inside and out in St Bartholomew’s, Brighton (Plate 93B), completed in 1875 by Edmund E. Scott (?-1895), and considerably later in Brooks’s own London church of All Hallows, Shirlock Street, begun in 1889 and never provided with its intended vaults.

Victorian Gothic, whether Early or High, is primarily an ecclesiastical mode. The leading Neo-Gothic architects were happiest when building churches; their few secular works—if parsonages, colleges, and schools can really in this period be called secular—generally have a churchy tone. But it is characteristic of the High Victorian Gothic as opposed to the Early Victorian Gothic, and _a fortiori_ to Neo-Gothic on the Continent, that it became for some twenty years, from the early fifties to the early seventies, a nearly universal mode.[233] A good many houses have already been cited; and certainly no churches of this period provide finer specimens of High Victorian Gothic than the warehouse at 104 Stokes Croft in Bristol, which was built by E. W. Godwin (1833-86), a friend of Burges, in the early sixties (Plate 113), or the office building of 1864-5 at 60 Mark Lane in London by George Aitchison (1825-1910). The one is an especially subtly polychromed attempt to follow Ruskin’s Italianism, the other more ‘Early French’ in its detail, but both use round-arched arcading throughout their several storeys (see

## Chapter 14).

Godwin in two rather modest town halls, one at Northampton of 1861-4, which is very rich in sculptural detail, the other at Congleton, Cheshire, of 1864-7, which is more severe and ‘Early French’ in character, produced two further High Victorian Gothic[234] works of the highest quality (Plate 92A). Unfortunately by the time the taste of the authorities in the larger English cities caught up in the late sixties with the advanced position of the High Church architectural leaders, those leaders had left that position far behind. As a result, many of the biggest and most conspicuous public edifices are very retardataire. Gothic designs won only low premiums in the Government Offices competition in 1857, although both Street’s and Deane & Woodward’s—on which Ruskin advised—were of considerable distinction. When Alfred Waterhouse (1830-1905) two years later won the competition for the Manchester Assize Courts he elaborated the design of this large public structure along the rather unimaginative lines of Deane & Woodward’s earlier Oxford Museum, then just reaching completion.

At best Waterhouse had a rather heavy hand and an uncertain sort of eclectic taste somewhat like G. G. Scott’s. He lacked the cranky boldness of a Butterfield, the sophistication of a Street, and the sense of craftsmanship of such men as Webb and Godwin who were his own contemporaries. But he did have real capacity as a planner of large and complex buildings, something at which most of the leading church architects had little or no experience. Thus his Manchester Town Hall, begun ten years later than the Assize Courts in 1869, while lacking all the refinement of Godwin’s smaller and earlier ones, is a large-scale exercise in High Victorian Gothic of some interest. But inevitably the High Victorian Gothic was a mode less well suited to this kind of monumental exploitation than the contemporary Second Empire mode as naturalized in England and America. For all the skill of Waterhouse in the organization of plan and general composition and in the bold detailing of materials inside and out, the Manchester Town Hall is a late and inferior work—late, that is, in the phase of style which it represents, though not so late in the highly successful career of its architect. It may properly be compared, and to its own manifest advantage, moreover, with Schmidt’s Rathaus in Vienna.

The other most conspicuous High Victorian Gothic public monument, the Law Courts in London, is the work of Street, an older and far more distinguished architect; but it came very late indeed in Street’s career, so late that he died before it was finished in 1882. Designed originally for a competition held in 1866, many years dragged by during which the site was twice changed—once southward to the river’s edge and then back to the north of the Strand—before it was even begun in 1874. Other work of the late sixties and early seventies by Street indicates how completely his own taste had turned away from this sort of French thirteenth-century Gothic even before the Law Courts were started.

At St Margaret’s in Liverpool, for example, which he designed in 1867, Street reverted to English fourteenth-century models; thus, like Bodley at All Saints’, Cambridge, he seemed to be returning to the particular stylistic ideal with which the ecclesiologists had started out twenty-five years before. In the Guards’ Chapel at the Wellington Barracks in London, however, which was all but completely destroyed in the blitz, he in 1877 remodelled the interior of an engineer-built Grecian edifice with incredible sumptuousness in a sort of Byzantinoid Italian Romanesque, using a stone-and-brick banded barrel vault and a glittering investiture of gold and glass mosaic that quite outshone the comparable work of Continental architects in the _Rundbogenstil_. Then, in remodelling the interior of St Luke’s, West Norwood, near London, built by Francis Bedford (1784-1858) in 1823-5, equally Grecian, he used in 1878-9 round-arched Italian detail. Despite the bold banding in brick and stone, this is certainly not Gothic or Byzantine, but rather recalls the Tuscan Proto-Renaissance, or even the _quattrocento_.

Certain buildings by Deane & Woodward and by Scott at Oxford and Cambridge have already been mentioned; much more exists by Scott, Waterhouse, and various others, very little of it of any distinction, yet sometimes fitting not too uncomfortably into the general scene. The most striking example of Victorian Gothic architecture at Oxford, fortunately on an isolated site opposite the Parks, where it had no neighbours earlier than the Museum, is Butterfield’s Keble College, a complete entity in itself, largely built in 1868-70. With its walls so violently striated with bricks of various colours, Keble would have been a most disturbing increment to any existing college; on the other hand, Butterfield’s quietly stone-banded chapel at Balliol of 1857 is that college’s happiest feature, the rest being largely the work of Waterhouse.

Since Keble was founded by Butterfield’s pious High Church friends for clerical students, the chapel, which was added to the group in 1873-6, understandably dominates the whole. Tall and richly decorated, this has many of Butterfield’s virtues, but it quite lacks the directness and the poignance of his best work of the fifties and early sixties. The hall and library are less monumental than the chapel, fitting more easily into the ranges of sets that surround the two quadrangles. The over-all composition is fairly regular, and there is less coarse or fussy detailing than Scott and Waterhouse used for their ‘Collegiate Gothic’. Moreover, the scale of Keble is modestly domestic and, despite its considerable size, the features are simple and crisp; but in the relatively clean air of Oxford Butterfield’s polychromy has received less of the desirable mellowing than it gets in London. The banded walls certainly lack the harmony that the softer colours of the materials used in his country church interiors generally produced.

By the time Keble was completed—indeed in advanced circles well before it was begun—such polychromatic brashness was out of date. Yet at Rugby School, where Butterfield’s buildings of 1868-72 awkwardly adjoin various earlier nineteenth-century Gothic structures, the polychromy is even louder; moreover, it is still less mellowed by time. Although Butterfield lived on through the rest of the century and continued to build many churches and some schools, this first and boldest of High Victorian Gothic architects was more and more left behind after the mid sixties by the evolving taste of his own High Church milieu.

There are other High Victorian Gothic collegiate groups which are, or would have been if carried to completion, far finer than Keble. Being at less renowned institutions than Oxford, they are less well known. University College on the sea-front at Aberystwyth in Wales is by J. P. Seddon (1827-1906), from 1852 to 1862 a partner of John Prichard. This structure was begun in 1864 to serve as a hotel, incorporating as its most inappropriate nucleus a small Castellated villa built by Nash for Uvedale Price in the 1780s. The failure of the hotel project, the slow and faltering start of the college, and the necessary repair and rebuilding after two fires have left a complex pile of most disparate character, even though it is almost all by Seddon. But certain aspects of the building, the bowed section on the sea-front—originally the hotel bar, later the college chapel!—and the entrance and stair tower on the rear are among the grandest and most boldly plastic fragments produced in this period (Plate 91A). Neither Oxford nor Cambridge has anything of comparable quality.

For Trinity College in Hartford, Conn., Burges prepared in 1873 a splendid plan worthy of its fine new site on a high ridge south of the city (Plate 88). Unfortunately only one side of one quadrangle was finished according to his designs; but that is perhaps the most satisfactory of all his works, and the best example anywhere of Victorian Gothic collegiate architecture. The brownstone from nearby Portland, Conn., favourite material all over the eastern states during what Lewis Mumford has called the ‘Brown Decades’, is especially well suited to Burges’s heavy and well-articulated detail. The rough quarry-facing of the random ashlar contrasts tonally with the more smoothly cut trim in a fashion that is polytonal if not polychromatic. The roughness of the stone walls also enhances the massive proportions of the long dormitory range and of the paired towers with their boldly pyramidal roofs. Yet for the classrooms this masonry is articulated into banks of large mullioned windows. Despite the general regularity and even symmetry of the composition, there is plenty of functionally logical variety in the handling of the different sections. Burges was happy in the Scottish-born Hartford architect who supervised the work, G. W. Keller (1842-1935); and Keller revealed his continued debt to Burges in the construction of a Memorial Arch in the park in Hartford which is one of the very few examples of such a Classical monument completely translated into Gothic terms, and not without real interest.

Burges undoubtedly enjoyed more what he did for the Marquess of Bute, beginning in 1865, in restoring Cardiff Castle and Castell Coch in Wales. ‘Restoring’ should be put in quotation marks, for by the time Burges got through with them both were almost as much fake castles as any built in the first half of the century. They lie somewhere between Fonthill Abbey and Peckforton in intention and are considerably more sumptuous internally than either. Although Cardiff Castle, which had been subjected to drastic Georgian remodelling, was gradually re-castellated with considerable consistency, the work there never reached completion. It is chiefly the incredibly rich interiors that are of interest, even if the interest is of a rather theatrical order.

Castell Coch near Llandaff, restored in 1875, interiors of equal fantasy, almost comparable to those of Neuschwanstein; that is, they are more like settings for Wagnerian opera than anything the Middle Ages actually created. But the quality of the imagination and of the execution is of a very much higher order than Ludwig II commanded. Externally Castell Coch is a sober and plausible restoration-reconstruction of a smallish castle, chiefly of archaeological interest, but most romantically sited and solidly built. Beside its integrity the more famous restorations by Viollet-le-Duc at Pierrefonds and Carcassonne appear rather harsh, and obviously modern.

The McConochie house, built in Cardiff for Lord Bute’s estate agent, is one of the best medium-sized stone dwellings of the High Victorian Gothic, superior in almost every way to Burges’s own house at 9 Melbury Road in London. That was built later, in 1875-80, by which time the operatic medievalism of the interiors was quite out of date (see Chapter 12). Here in the Cardiff house the tight asymmetrical composition, the excellent detailing of the handsome stonework, and a generally domestic rather than Castellated air prepared the way for Burges’s fine collegiate work in America.

English architects in the sixties were capable of exploiting a wide range of different aspects of the High Victorian Gothic in almost precisely the same years. Only the size and departmentalized organization of G. G. Scott’s office, the largest of the period and more like the ‘plan-factories’ of the twentieth century (see Chapter 24), can explain how he could be nominally responsible for such a quiet, well-scaled, and advanced church as St Andrew’s, Derby, designed in 1866—some say by Micklethwaite, who was working for him at the time—and also for such a strident, complex, and over-elaborated edifice as the Midland Hotel fronting St Pancras Station. The design for this was prepared in 1865 for a competition held, curiously enough, two years after the shed had been begun by the engineers W. H. Barlow (1812-1902) and R. M. Ordish (1824-86). Such a drastic divorce of engineering and architecture could hardly be expected to produce a co-ordinated edifice, yet both aspects of St Pancras have considerable independent interest. The shed, ingeniously tied below the level of the tracks and rising, for purely coincidental technical reasons, to a flattened point of slightly ‘Gothic’ outline, has the widest span of any in the British Isles and, until the nineties, in the world. It is, therefore, a nineteenth-century spatial achievement of quantitative, if not so much of qualitative, significance. The masonry block at the front is one of the largest High Victorian Gothic structures in the world. It long had ardent admirers, and it has come to have them again, for it epitomizes almost as notably as the Albert Memorial the aspirations of Scott and his generation. The contrast to its neighbour, Lewis Cubitt’s Kings Cross Station, begun some fifteen years earlier, or even to Paddington, where the engineer Brunel and the architect Wyatt collaborated so happily, is striking. The taste of English railway authorities, as of most patrons of architecture, had been revolutionized by the general triumph of the High Victorian Gothic in the late fifties and early sixties. Yet on its completion in the mid seventies St Pancras was even more out of fashion in advanced circles than were Street’s Law Courts, the construction of which only began at that time, so rapidly did taste continue to change in the late sixties and early seventies.

By 1870 church architecture, for example, was in general much chastened. Externally Teulon’s St Stephen’s, The Green, on Rosslyn Hill in Hampstead of 1869-76 is not polychromatic but all of purple-brown brick with some creamy stone trim. It builds up, moreover, somewhat like Shaw’s Bingley church begun a few years earlier, to a tall rectangular crossing tower with rather quiet, more or less ‘Early French’, membering. Inside Teulon achieved in the brickwork a kind of golden harmony of tone resembling that of White’s interior in St Saviour’s, Aberdeen Park, completely eschewing the bold and almost savage patterns of contrastingly coloured bricks he had favoured since the early fifties. In the tremendously tall interior of Edmund Scott’s already mentioned St Bartholomew’s, Brighton—aisleless, chancel-less, and provided with broad, flat internal buttresses—the traces of brick polychromy are hardly noticeable on the walls of a space so grandly proportioned (Plate 93B). The later ciborium here is not by Scott.

Burges in the two Yorkshire churches which he began in 1871 at Skelton and at Studley Royal, both near Ripon, the latter with a very fine rectory near by, still aimed at a rather satiating luxury of both coloured and sculptural decoration in the interiors. But Pearson at St Augustine’s, Kilburn Park Road, in London, initiated at this time a new line of vast plain churches (Plate 93A). That line would culminate in the archaeological correctness of his Truro Cathedral in Cornwall, started in 1880 and finally completed by his son (F. L., 1864-1947) in the present century. His last work, the cathedral of Brisbane, Australia, designed shortly before his death in 1897, was only begun by his son in 1901.

As Pearson’s Kilburn church was built in 1870-80, it should perhaps more properly be considered Late Victorian than High. But Pearson retained here and to the end of his life, particularly in his tall towers and spires, a truly High Victorian love of grand and bold effects. However archaeological he became, and with his passion for rib-vaulting he could from this time on be rather more archaeological in a Franco-English way than Viollet-le-Duc in France or Cuijpers in Holland, his spaces are usually nobly proportioned and his masses crisply composed no matter how ‘correctly’ they are membered. At Truro, where the cathedral rises suddenly out of narrow streets, its granite still almost unweathered, Pearson’s handling of the relationship of the three tall towers carries vigorous plastic conviction; Burges had attempted the same effect at Cork with rather less success when the High Victorian was still at its highest. Brisbane Cathedral is plainer and tougher than Truro despite its very late date.

It would be inappropriate in this chapter to carry the story of Victorian Gothic much further. Scott and Street died in 1878 and 1881 respectively, though Butterfield and Bodley outlived Pearson. Butterfield seems to have frozen for life in the mode of his early maturity, and as a result produced ever feebler work after the mid sixties; Pearson was able to maintain a leading position with a younger generation grown chaster and more archaeological in its standards without forsaking his pursuit of those more abstractly architectonic values which give distinction to his earlier work. It was above all Bodley, however, with his Late Decorated verging on Perpendicular, who set the pace in Anglican church-architecture from this time forward. His personal style, still tentative at All Saints, Cambridge, in the mid sixties, was mature by the time he built St Augustine’s at Pendlebury in Lancashire in 1870-4. Crisp and almost mechanical in its detailing, this tall rectangular mass, buttressed by an internal arcade, is impressive both inside and out (Plate 92B), yet it wholly abjures most of the qualities that had for two decades given special vitality to English Neo-Gothic.

With various modulations what might, rather ambiguously, be called ‘Bodleian Gothic’ remained the favourite of Anglicans in and out of England well into the twentieth century. The continuing admiration for the work of Sir Ninian Comper (1864-1960) in certain milieus suggests that it has not even yet been finally superseded; but much of Comper’s large-scale work dates from before Bodley’s death in 1907. For example, his principal London church, St Cyprian’s in Glentworth Street, was built in 1903. This crisp and clean example of revived Late Gothic, with its elegant gilt font-cover and screen, may wind up this account more appropriately than the vast unfinished cathedral at Liverpool begun by Sir Giles Gilbert Scott (1880-1960), a grandson of the first G. G. Scott, in 1903. But neither is Victorian Gothic; both are rather manifestations of one aspect of twentieth-century ‘traditionalism’ (see

## Chapter 24).

------------------------------------------------------------------------

## CHAPTER 11

LATER NEO-GOTHIC OUTSIDE ENGLAND

THE High Victorian Gothic produced in the United States no such roster of distinguished—or at least prominent and highly characteristic—monuments as in Britain. The period of its florescence was much briefer, and few assured and sophisticated talents came to the fore. If, in the case of Richardson, one such did appear, his maturity came only in the mid seventies, when the High Victorian Gothic was all but over. Why the period was so much shorter in the United States, in effect only the decade 1865-75, is not altogether clear. One reason, undoubtedly, is that the speed of transmission of new architectural ideas from England to America had increased so much by the seventies that the influence of the later English mode which succeeded the High Victorian Gothic around 1870 reached America very promptly indeed (see Chapters 13 and 15). Another quite different reason is that a wave of nationalism in America, parallel to those current in North European countries at the time, encouraged from the mid seventies developments that were more autochthonous. Leadership in commercial and in domestic architecture crossed the Atlantic almost precisely at the moment when, in 1876, the centenary[235] of American political independence was being celebrated.

The phenomenal success in the United States of Ruskin’s treatises, _The Seven Lamps of Architecture_ of 1849 and _The Stones of Venice_[236] of 1851-3, should be emphasized; from 1855 Street’s _Brick and Marble Architecture_ was also available. Yet, despite the warm reception of such relevant writings, few reflections of the High Victorian Gothic can be discerned in American production before 1860. The first is probably the Nott Memorial Library[237] at Union College, Schenectady, N.Y., designed by Edward T. Potter (1831-1904) in 1856 and built in 1858-76. Here the banded arches are pointed and the plan is circular, perhaps in emulation of the Pisa Baptistery to which Ruskin had called attention, but more probably in deference to Ramée’s general plan for the college (see Chapter 1).

The years immediately following the Panic of 1857 and, quite understandably, the Civil War years 1861-5 were relatively unproductive of new buildings, as has already been noted. An edifice far more overtly Ruskinian than Potter’s Library was the National Academy in New York, built by Peter B. Wight (1838-1925) in 1863-5, although apparently first designed as early as 1861. Its Venetian Gothic mode, with pointed arches boldly banded and walls diapered in coloured stones, was still the subject of considerable contemporary controversy as it would hardly have been in England by this date.

Potter and Wight were both young men. Established Gothic Revivalists in America did not swing over as rapidly as in England from the Early Victorian to the High. Upjohn, Potter’s master, was no Butterfield; Renwick when designing St Patrick’s Cathedral in New York in 1859 followed contemporary Continental rather than English models, as has been noted, presumably because his clients were Catholics.

At best the sort of High Church Anglican patronage which sponsored Butterfield’s and Street’s innovations in England was relatively much less important in the United States—or Canada and Australia, for that matter. Enthusiasm for the High Victorian Gothic, although widespread in the later sixties and early seventies, was rarely exclusive as is evidenced by the disparate interests and activities of the members of the prominent and successful firm of Ware & Van Brunt. It has already been noted that when William Robert Ware founded in 1865 the first American architectural school at the Massachusetts Institute of Technology in Boston, he based its instruction on that of the Paris École des Beaux-Arts.[238] His partner Henry Van Brunt (1832-1903) was one of the first to follow Richardson’s lead away from the High Victorian Gothic in the seventies. So little were either of them dyed-in-the-wool Gothicists in these decades.[239]

However, Ware & Van Brunt designed and built in Cambridge, Mass., one of the largest and most conspicuous of mature High Victorian Gothic edifices in America, Memorial Hall[240] at Harvard College, first projected in the late sixties and erected in 1870-8. This somewhat cathedral-like edifice has walls of red brick liberally lashed with black and a massive central tower now denuded by fire of its high roof (Plate 95A). The manner is more than a little Butterfieldian, but the quality is not even up to G. G. Scott.

Before Memorial Hall was designed, a competition held in 1865 for the First Church (Unitarian) in Boston in the new Back Bay residential district had brought out a variety of rather feeble attempts by Boston architects to follow the High Victorian Gothic line. The winning design of Ware & Van Brunt, executed in 1865-7, while not of the wilder Low Church order of Teulon’s or Keeling’s London work of these years, is hardly comparable to Street’s or Butterfield’s, much less to the contemporary production of younger architects such as Brooks, Bodley, or Shaw. Its best feature is the material, the richly mottled and textured local Puddingstone from nearby Roxbury.

The High Victorian Gothic of the sixties and early seventies in the United States was no more restricted to the ecclesiastical field than in England. Despite its churchy look, Memorial Hall served a variety of secular purposes from refectory to concert hall; only the wide transeptal lobby was strictly memorial in purpose. But there was rarely even such relative devotion to the Gothic in this period in the United States as the major works of Ware & Van Brunt display. For example, the untutored Elbridge Boyden (1810-98), best known for introducing the cast-iron commercial front into New England in 1854, could build two buildings for the Polytechnic Institute of Worcester, Mass., in the same year 1866 of which one, the Washburn Machine Shop, is mansarded with crude, vaguely Second Empire, detailing; while the other, Boynton Hall, is in a very provincial sort of High Victorian Gothic. Hunt, product of a Parisian education, designed the Yale Divinity School in New Haven in 1869 in a frenzied, rather Teulonian, Gothic; while in his precisely contemporary Lenox Library in New York, built in 1869-77, he followed closely and with some dignity French, if not specifically Second Empire, models.

It is not really surprising, therefore, that Richardson, returning from Paris and the École des Beaux-Arts at the end of the Civil War and entering a competition for a new Unitarian church to be built at Springfield, Mass., offered a High Victorian Gothic project that seems to derive rather directly from the work of Keeling and other Low Church English practitioners. What _is_ surprising, however, considering the lack of special interest to later eyes in his Unity Church as executed in 1866-8, is the fact that he won the competition! The warm colour and texture of the rock-faced brownstone from nearby Longmeadow laid up in random ashlar, a certain masculine scale in the details, and an attempt at least at a boldly asymmetrical composition evidently struck his contemporaries as very promising, however. (The church was demolished in 1961.)

It was not in the Unity Church, but in Richardson’s second church, Grace Episcopal, in Medford, Mass., happily still extant, of 1867-8, that one recognizes strong personal expression. The more massively pyramidal character of the asymmetrical composition and, above all, the great boulders of which the walls are built, with heavy trim of rough quarry-faced granite, announce an original approach (Plate 91B). Yet this approach was evidently still nurtured on the English High Victorian Gothic models that Richardson knew through the wood engravings in imported periodicals. It is even specific enough here so that one can describe this Medford church as Burgessy rather than Butterfieldian or Street-like; it is certainly no longer Keelingesque like the church in Springfield. Incidentally, when Richardson visited England in 1882 it was the work of Burges, who had just died, that he went out of his way to see—by that time, however, he found it rather disappointing.

If Richardson’s first churches were Gothic, his Western Railway Office at Springfield, built in 1867 for a client associated with the Unity Church commission, was generically Second Empire. Yet this was still more directly derived from current English work that was closely related to that mode, notably the Francis Brothers’ National Discount Building of 1857 in the City of London, than from anything Parisian. His brick and stone Dorsheimer[241] house of 1868 in Delaware Avenue in Buffalo, N.Y., is also Second Empire rather than Victorian Gothic, but very restrainedly so, and hence rather more French in effect. Other work by Richardson dating from the late sixties, such as the B. H. Crowninshield house in Marlborough Street in Boston of 1868-9, was more experimental in design, often recalling wild English work of the early years of the decade. Although built of wood and of very modest size, Richardson’s most interesting house of this period was the one that he built for himself in 1868 at Arrochar on Staten Island near New York.[242] This combines the use of a high mansarded pavilion with a sort of imitation half-timbering related to the contemporary American ‘Stick Style’ (see

## Chapter 14).

In Farnam Hall at Yale College in New Haven (Plate 96A), begun in 1869, the German-trained Russell Sturgis (1836-1909),[243] who had been for a time Wight’s partner, somehow arrived at an almost Webb-like—or at least Brooks-like—simplicity and sophistication of late High Victorian Gothic design, in marked contrast to the stridency of Hunt’s precisely contemporary Divinity School there. This, however, is almost unique. The most characteristic work of the day was produced by such home-trained architects as Ware & Van Brunt, Wight, Edward T. Potter, and his younger brother William A. Potter (1842-1909).[244] Wight’s National Academy in New York has been mentioned. His Mercantile Library in Brooklyn, N.Y., completed in 1869, of red brick with ranges of pointed-arched windows regularly but asymmetrically disposed, is similar—and not inferior—to much of G. G. Scott’s secular work. Edward T. Potter’s Union College Library has also been mentioned. His Harvard Church in Brookline, Mass., of 1873-5 is more conventional for its period. Largely renewed internally after being gutted by a fire in 1931, this shows how effectively such American materials as the popular brownstone from Portland, Conn., and the light-coloured Berea sandstone from Ohio, enlivened by accents of livid green serpentine from Pennsylvania, could produce a polychromy richer and more enduring than the endemic Butterfieldian or Teulonian red brick, with banding of bricks dipped in black tar, that had been in general use for a decade. Along this line Richardson himself followed for a while (see Chapter 13). At the same time William A. Potter, who became very briefly Supervising Architect in Washington in succession to Mullet in 1875, produced a few post offices, such as the one in Pittsfield, Mass., that are characteristic but not very distinguished examples of secular High Victorian Gothic executed in stone. (Both Potters, however, gave up the High Victorian Gothic to accept Richardson’s leadership within the next few years.)

The Boston & Albany Railroad station in Worcester built by Ware & Van Brunt in 1875-7, with its tall and striking tower and its vast segmental-pointed arches at the ends of the shed, provides one of the happiest illustrations of what the rather illiterate approach of even the most highly trained Eastern architects of this period could produce. By working in an almost primitive way, along lines suggested by the half-understood work of the bolder English innovators, something was often achieved of which few Continental architects were capable in this period. In less sophisticated hands, whether of provincial architects or of builders, the results were naturally still cruder, though sometimes equally vital and fresh. In church design,[245] where ecclesiological control of planning was not accepted outside the Episcopal denomination, galleried auditorium schemes with rows of exposed iron columns were often executed with a violence of polychromy and a gawkiness of notched detailing that exceeded Teulon or Keeling at their most extreme. One of the most prominent extant examples is the squarish New Old South Church at Copley Square in the Back Bay district of Boston, built in 1874-5 by Charles A. Cummings (1833-1905) and his partner Sears in 1875-7. Its impressive tower resembling an Italian campanile has now been much reduced in height and chastened in silhouette.

Even more extreme than most churches, but of the highest quality, is the intensely personal work of Frank Furness (1839-1912)[246] in Philadelphia. His building for the Pennsylvania Academy of Fine Arts in Broad Street was erected in 1872-6 in preparation for the Centennial Exhibition. The exterior has a largeness of scale and a vigour in the detailing that would be notable anywhere, and the galleries are top-lit with exceptional efficiency. Still more original and impressive were his banks, even though they lay quite off the main line of development of commercial architecture in this period (see Chapter 14). The most extraordinary of these, and Furness’s masterpiece, was the Provident Institution in Walnut Street, built as late as 1879 (Plate 95B). This was most unfortunately demolished in the Philadelphia urban renewal campaign several years ago, but the gigantic and forceful scale of the granite membering alone should have justified its respectful preservation. The interior,[247] entirely lined with patterned tiles, was of rather later character than the façade and eventually much cluttered with later intrusions, but it was equally fine in its own way originally. Later work by Furness is of less interest, and his big Broad Street Station of 1892-4 has also been demolished. No small part of Furness’s historical significance lies in the fact that the young Louis Sullivan picked this office—then known as Furness & Hewitt—to work in for a short period after he left Ware’s school in Boston. As Sullivan’s _Autobiography of an Idea_ testifies, the vitality and originality of Furness meant more to him than what he was taught at the Massachusetts Institute of Technology, or later at the École des Beaux-Arts in Paris (see Chapter 14).

In the realm of house-design the more-or-less Gothic-based ‘Stick Style’ represented a largely autochthonous American development not without considerable significance and interest (see Chapter 15). In public architecture there was little serious achievement even at the hands of English-trained architects such as Calvert Vaux (1824-95) and his partner F. C. Withers (1828-1901)[248] or second-generation Gothicists like Upjohn’s son (Richard M., 1828-1903). The younger Upjohn’s Connecticut State Capitol[249] in Hartford begun in 1873, the only major American example of a High Victorian Gothic public monument of any great pretension or luxury of materials, is singularly vulgar and stylistically ambiguous, with its completely symmetrical massing and its tall central dome, compared to Burges’s contemporary project for Trinity College there.[250] Doubtless G. G. Scott would not have disdained it, even so!

Still more comparable to Scott’s own thwarted ambitions for a High Victorian Gothic governmental architecture, which led him as late as the seventies to enter various Continental competitions, is an earlier group of buildings in the New World outside the United States, the Parliament House (Plate 97A) and associated structures at Ottawa, Canada, designed by Fuller & Jones and Stent & Laver in 1859 and built in 1861-7. F. W. Stent had come out from England some considerable time before this, having last exhibited at the Royal Academy in London in 1846. Thomas Fuller (1822-98), also English, had settled in Toronto in 1856. Of their respective partners, Augustus Laver (1839-98) and Herbert Chilion Jones (1836-1923), less is known. In the course of the work Fuller and Laver joined forces, moving on shortly to the United States, as has been noted.

The main block at Ottawa, which was by the first-named firm, has been rebuilt after a fire in the present century in a considerably chastened vein, except for the big chapterhouse-like library at the rear, which is original. But the variety of form, the gusto of the detail, and the urbanistic scale of this project made of the Dominion Capitol a major monumental group unrivalled for extent and complexity of organization in England.[251] The buildings flanking the vast lawn extending in front of the Parliament House are by Stent & Laver. These are somewhat less exuberant in scale and more provincial in the character of their detailing than the Parliament House was originally.

Most of the Neo-Gothic in Canada up to this time is more properly to be considered Early rather than High Victorian (see Chapter 6). An exception to this, perhaps, is University College in Toronto, designed in 1856 by F. W. Cumberland (1821-81), who had come out from England in 1847. Yet its rich and rather bombastic Norman design is closer to English work of the earlier decades of the century than to the round-arched Ruskinian Gothic of the fifties.

Australia, the other major British Dominion, had nothing comparable to Canada to offer in this period. Wardell’s English, Scottish, and Australian Bank in Melbourne is a passable example of secular High Victorian Gothic but no more than that. St John Evangelist’s, which he built at Toorak south of Melbourne in 1860-73, is handsomer but very simple—still almost Puginian, indeed—and all of monochrome ashlar. The enormous Catholic cathedral of Melbourne, St Patrick’s, which Wardell began in 1860, is more Continental in character, with two west towers like Renwick’s St Patrick’s in New York and also a tall crossing tower completed only in 1939. The Catholic cathedral of Adelaide, St Francis Xavier’s, begun in 1870 and still without its intended western spires, reputedly goes back to a design prepared by Pugin before his death in 1852. But even the later design of his son E. W. Pugin, on which the executed work was actually based, must have been much modified over the years by W. H. Bagot (b. 1880), H. H. Jory (b. 1880), and Lewis Laybourne-Smith (b. 1888), who successively supervised the job. It is certainly no happier an example of High Victorian Gothic than Wardell’s Catholic cathedral in Melbourne.

The Anglican cathedral in Melbourne, St Paul’s, having been begun in 1850 from designs by Butterfield, ought to be finer. But Butterfield had made the drawings as early as 1847, before even he was a High Victorian, and the laggard execution of the church by Joseph Reed evidently entailed much modification of the original designs. Moreover, the spires by John Barr date only from 1934. For the very late Anglican cathedral at Brisbane, St John’s, perhaps the finest of the lot, which was begun in 1901 by F. L. Pearson from earlier designs by his father J. L. Pearson as has already been mentioned, Butterfield had also prepared designs in 1884.

The architecture of the Dominions remained Colonial in spirit, as these notes on a few Australian churches indicate, well into the present century. First the able Frank Wills, moreover, the English-born architect of Montreal Cathedral, and then Fuller & Laver were drawn away from Canada to the United States, where opportunities were greater. Despite the great interest of the Government Buildings at Ottawa, it was in the United States rather than the British Dominions that the High Victorian Gothic proved a stimulus to such highly original achievement as Furness’s in the seventies.

The High Victorian Gothic episode in American architecture balanced almost precisely the Second Empire episode. Both were disowned, even by many of their most successful protagonists, by the mid seventies. It was the Gothic, however, that prepared the way for the more original developments of the last quarter of the century; as has already been stated, those who had practised chiefly in the Second Empire mode continued to take their lead from Paris. Yet there are paradoxes in the situation which must not be ignored. Richardson, the most creative new force in the seventies and eighties, continually urged young aspirants to an architectural career to study at the École des Beaux-Arts as he had done. Charles F. McKim (1847-1909), Richardson’s first really able assistant, was Paris-trained; partly because of that training, it was he who became in the mid eighties the leader of the reaction against the Richardsonian. Sullivan, the first truly great modern architect not alone of America but of the whole western world, was also in part Paris-trained, even though he was always highly critical of the doctrine of the École and much stimulated by Furness. Finally, it was even more the later writings of the French Viollet-le-Duc than those of the English Ruskin that encouraged bold and imaginative thinking about architecture in America in the seventies and eighties when his _Entretiens_ became available in translation and were first widely read.[252]

Were this a history of architectural thought rather than of architecture—that is of what was actually _built_ in the nineteenth and twentieth centuries—Viollet-le-Duc would play a much larger part. But his production,[253] while not negligible, is curiously ambiguous. His many ‘restorations’ are no contribution to nineteenth-century architecture; rather they represent a serious diminution of authenticity in the great monuments of the past subjected to his ministrations. These include most notably Notre-Dame in Paris, the refurbishing of which he continued alone after the death of Lassus in 1857, and the Château de Pierrefonds, Oise, the rebuilding of which began the next year and continued down to his death in 1879; but the whole list is very long indeed, including Carcassonne, Vézelay, and Saint-Denis, to mention only some of the best known things.

Viollet-le-Duc’s new parish church for the suburb of St-Denis, Saint-Denys-de-l’Estrée in the Boulevard Jules Guesde, built in 1864-7, has considerable interest, however. Unlike most English High Victorian Gothic churches, it is vaulted throughout; but the vaulting does not have that look of a student exercise which characterizes Lassus’s at Saint-Jean-de-Belleville in Paris of the previous decade. The broad square bays of the nave are well lighted by groups of lancets in the clerestory, and there is a sturdy sort of articulation of the elements not unlike that in the early work of Burges (Plate 98). Externally the rather complex plan, with a large rectangular Lady Chapel projecting behind the altar, produces a gawky and confused composition; but the detailing is simple and virile as in the interior. A massive western tower rises over the entrance porch, culminating in a tall slated roof rather than a stone spire. But the plate tracery of the large west window over the porch and the lancets of the stage above are stony enough and have a quite Street-like scale and vigour of form. It is perhaps unfortunate that Viollet-le-Duc built so few new churches; certainly most other French Neo-Gothic work is very inferior to this, as such a large and prominent church as Saint-Epvre at Nancy, begun in 1863 by M.-P. Morey (1805-78), a pupil of Leclerc, well illustrates.

In secular work Viollet-le-Duc was too often content to follow the current Second Empire mode with a good deal of the eclecticism, but little of the plastic boldness, of the English and the Americans. Such more or less Gothic blocks of flats as those that he built in the late fifties and sixties in the Rue de Condorcet and at 15 Rue de Douai in Paris are somewhat more comparable to the secular High Victorian Gothic in England (Plate 101A). These are certainly praiseworthy for the urbanistic politeness with which they fit between more conventional Second Empire neighbours despite their distinctly ‘Victorian’ detail,[254] but there is little originality of conception. On paper Viollet-le-Duc later showed great boldness, however, in certain projects proposing the use of metal structural elements that he published with the second volume of the _Entretiens_ (see Chapter 16).

In the late fifties and sixties the vigour of the ‘Early French’ detailing of certain English architects and a related logic of structural expression then called ‘real’ was often derived in part from a study of Viollet-le-Duc’s _Dictionnaire_. But Shaw’s book of _Continental Sketches_ of 1858 and Nesfield’s similar book of 1862 make evident how intense and how idiosyncratic was their own first-hand study of medieval work across the channel. Certainly the ‘Early French’ detail of the English leaders is generally of higher quality than even Viollet-le-Duc’s best at Saint-Denys-de-l’Estrée.

If there was very little Gothic work done in the third quarter of the century in France comparable in quality or in interest to that of the Anglo-Saxon countries, yet there was a general movement there away from the somewhat mincing attitudes of the forties and early fifties. Just as the Medieval Revival in America, considered in a broad sense, came to its climax in the mature work of Richardson (see Chapter 13)—which is much more Romanesque than Gothic in so far as it leans at all on the past—in France the Romanesquoid work of Vaudremer represents the highest achievement of the period in a non-Renaissance mode (Plate 72A). The same may even be said up to a point of most of the other countries of Europe. Yet the Germanic _Rundbogenstil_ of the third quarter of the century was, for all the size, prominence, and elaboration of such public monuments as Waesemann’s Berlin City Hall or Hansen’s Vienna Waffenmuseum and the real excellence of Herholdt’s Danish work, already a sinking rather than a rising mode.

In Germany and Austria more Neo-Gothic edifices, both secular and ecclesiastical, were built after 1850 than before; several of them have already been mentioned. These are, however, rather examples of contemporary eclecticism than of a concerted movement. In addition to his school and his Rathaus, however, Schmidt built in Vienna some eight Gothic churches ranging in date from the Lazaristenkirche of 1860-2 to the Severinkirche of 1877-8. Most of them are brick-vaulted hall-churches—that is, of the characteristic medieval German plan and section, with aisles of the same height as the nave. However, the largest and most interesting, the Fünfhaus Parish Church of 1868-75, is centrally planned. This is an aisled octagon rising to a ribbed dome with hexagonal chapels grouped around the irregularly polygonal apse (Plate 99B). The spatial complexity of the interior is of real interest, and the walls are painted to suggest polychromatic brickwork of almost English brashness. Two front towers flanking the gabled entrance bay are set close against the dome to provide a very Baroque sort of composition—this is really, therefore, a sort of Sant’ Agnese in Agone or Karlskirche carried out with a G. G. Scott vocabulary of Neo-Gothic elements.

In Hungary the eighties saw a very belated manifestation of secular Neo-Gothic. The Parliament House, begun in 1883 by Imre Steindl (1839-1902) and completed in 1902, was surely inspired by Barry’s in London begun nearly a half-century earlier, but in character it is (not surprisingly) more like Schmidt’s Vienna Rathaus. Thus did outlying countries in the later decades of the century continue to take up modes long obsolescent in the major architectural centres.[255]

The Gothic of C. F. Arnold (1823-90) at Dresden, as seen in his secular Kreuzschule of 1864-5 or the two-towered Sophienkirche of the same years, is inferior to Schmidt’s, both in command of the idiom and in architectonic organization, as indeed is most such German work of these decades. The Johanniskirche in Dresden of 1874-8 by G. L. Möckel (1838-1915), however, has a rather fine tower set in the transeptal position so much favoured in Victorian England. This is bold in scale and carefully detailed in a literate twelfth-century—not to say ‘Early French’—way much as Burges or Pearson might have designed it in England. More characteristic of German work of these decades is the Munich Rathaus, built in 1867-74 by G. J. von Hauberrisser (1841-1922) and extended by him in 1899-1909. Excessively spiky, this seems almost to have borrowed back from G. G. Scott the more Germanic features of his Broad Sanctuary terrace in London of fifteen years earlier. But the Neo-Gothic of the seventies and eighties in Germany is in general no more aggressive and gawky than the popular Meistersinger mode that revived so turgidly the forms of the Northern Renaissance (see Chapter 10).

Holland, which made almost no significant architectural contribution in the first half of the nineteenth century, now produced in P. J. H. Cuijpers (1827-1921) a sort of Dutch Viollet-le-Duc. In addition to undertaking important restorations, he built many vast new Gothic churches of brick which he exposed once more in reaction against the earlier nineteenth-century practice of stucco-coating. Cuijpers was learned and ambitious, and in such work he could be rather more original than Viollet-le-Duc in France, if less so perhaps than Schmidt in Austria. His Vondelkerk, a church of 1870 near the Vondel Park in Amsterdam, is not centrally planned like Schmidt’s Fünfhaus church in Vienna, but he obtained a somewhat similar spatial effect by making the crossing octagonal. The brickwork of the piers and the vaults is very richly treated but in a fashion as much polytonal as polychromatic. The banding is in bricks of different sizes and textures rather than of different colours, and the result has something of the subtlety of the interior of White’s Aberdeen Park church in London.

A larger and later Amsterdam church by Cuijpers, the Maria Magdalenakerk in the Zaanstraat of 1887, is considerably more impressive, both inside and out. Occupying one of those narrow triangular sites so often assigned to important urban churches in this period, the exterior builds up grandly to the rather severe crossing tower at the rear. Inside, Cuijpers made the most of the difficulties of the site also. The east end is conventionally Gothic in plan, and the choir here is brick-vaulted, as is the Vondelkerk throughout. But the taller nave, covered with a wooden roof of ogival section, is much more effective spatially because of the way it is widened by triangular elements at the front where the aisles are cut off owing to the narrowing of the site (Plate 101B). The later painted decorations in this church are harmonious in tone with the brickwork, and the whole has a breadth of attack comparable to some of the best English churches of the seventies, such as Pearson’s in Kilburn or Edmund Scott’s St Bartholomew’s, Brighton, without resembling any of them very much.

Curiously enough for so dedicated a church-builder, Cuijpers’s secular work is more conspicuous, and hence better known, than are his churches. The two largest and most prominent nineteenth-century buildings of Amsterdam are both by him. In these, the Rijksmuseum built in 1877-85 (Plate 101C) and the Central Station of 1881-9, he moved away from the emulation of thirteenth- or fourteenth-century ecclesiastical Gothic towards a more elastic sixteenth-century sort of design, rather similar to the English mode of these decades known as ‘Pont Street Dutch’ (see

## Chapter 12).

The similarity to the Northern Renaissance mode of this period in Germany is nearly as great, as also to such somewhat later Scandinavian buildings as Clason’s Northern Museum in Stockholm and Nyrop’s Town Hall in Copenhagen (Plate 173A). But Cuijpers’s touch is lighter than that of the Germans, and his precedent rather more Late Gothic than Mannerist, while his two chief works precede those that they most resemble in Sweden and Denmark by a decade or more. In both cases the frank incorporation of iron-and-glass elements is notable, a vast shed at the station and two almost equally vast covered courts in the museum. Above all, being the Gothic Revivalist he was, Cuijpers saw to it that the craftsmanship was excellent throughout; while his handling of scale, though ambiguous as in much work of these decades everywhere, is surprisingly successful. Both are very large buildings, placed in isolation where they can be seen from a distance and with carefully studied silhouettes varied by towers and other skyline features; yet the membering is delicate and almost domestic, quite as in the rather comparable English work of George (Plate 104B) or Collcutt (see Chapter 12).

In Italy projects of restoration led, as elsewhere, to the designing of certain fairly ambitious new façades in Gothic to complete medieval churches. The most conspicuous is that of the cathedral of Florence. After many abortive earlier moves, this was finally begun by Emilio de Fabris (1808-83) in 1866, when Florence became briefly the capital of Italy, and completed only in 1887. The earlier and less successful façade of Santa Croce in Florence had been carried out in 1857-63 by Niccoló Matas (1798-1872). It is characteristic of the international architectural scene in these decades that neither of these carefully archaeological compositions in polychrome Italian Gothic comes alive in the way that Italianate High Victorian Gothic often did in the hands of English architects, or even American ones, in the fifties and sixties.

Churches were built for Anglicans in most of the principal cities of Europe in the mid nineteenth century, usually by English architects and always in Victorian Gothic. Sometimes, as in the case of the Crimean Memorial Church by Street[256] at Istanbul and Shaw’s English Church at Lyons, these were by the most distinguished English designers of the day, but more often they were by hacks who lived abroad and specialized in such work. Among the ‘English churches’ of this period that provided good samples of the High Victorian Gothic for foreigners—many were still to all intents and purposes Early Victorian—are two by Street[257] in Rome, one for the English community, the other not ‘English’ at all in fact but built for American Episcopalians. The former, All Saints’, in the Via del Babuino, with a much later tower not by Street, provides internally a moderately successful example of his later work, although it is unimpressive and largely invisible externally. It was begun in 1880, a year before Street’s death, and opened in 1885.

Far finer is St Paul’s, the American church, prominently located among the contemporary banks and blocks of flats of the Via Nazionale and built in 1873-6. Boldly banded in brick and stone and with a tall square campanile at the front corner, this is indeed a richer and more striking example of an Italian Gothic basilica than the Middle Ages ever produced in Rome (Plate 100). The interior, with a rich apse mosaic by Burne-Jones on a glittering gold ground, has an originality and a coherence that is quite lacking in such Italian churches as were redecorated in the later nineteenth century. Late though this is in Street’s _œuvre_, it remains one of his best works.

If the English High Victorian Gothic was to some extent an article of export—and, of course, this account has hardly touched on the vast outlying areas of the British Empire, notably including India, to which it was exported in the greatest quantity—it was nevertheless largely without real influence outside the United States and the British Dominions. In the world picture, it was the British architectural critics of this period, Ruskin and Morris, who would have a vital influence, but that influence came for the most part rather later, around 1890 (see Chapter 16). Cuijpers, however, was a reader of Ruskin from the fifties.

Still to be discussed is the early work of one great architect, also reputedly a reader of Ruskin, whose career began in the seventies with a sharp revulsion from the Second Empire mode towards the Neo-Gothic. The Spanish (or more precisely Catalan) architect Antoni Gaudí i Cornet (1852-1926) was one of the most intensely personal talents that either the nineteenth or the twentieth century has produced. His style hardly matured before the nineties, and what are generally considered his typical works must be discussed later in connexion with the Art Nouveau (see Chapter 16). But what he had accomplished already in the seventies and eighties can be better appreciated here in relation to the contemporary work of those decades in other countries.

Gaudí’s earliest work was at the Parc de la Ciutadella in Barcelona, laid out in 1872, where he assisted the master of works Eduardo Fontseré, while still a student, in various projects for its embellishment. The elaborate Cascade there, incorporating an Aquarium, on which he worked in 1877-82 derives in the main from Espérandieu’s at the Palais Longchamps in Marseilles. But some of the detail, both plastic and incised, has a flavour more comparable to that of the wildest and most eclectic English and American Second Empire work of the previous decade than to anything French.

The first commission for which Gaudí was wholly responsible is the house of Don Manuel Vicens at 24-26 Carrer de les Carolines in Barcelona. This was erected in 1878-80, immediately upon his graduation from the local Escuela Superior de Arquitectura, and in it no trace of Second Empire influence, French or international, remains. A large suburban villa built of rubble masonry liberally banded with polychrome tiles, the Casa Vicens passes beyond the extravagances of a Teulon or a Lamb in the sixties into a world of fantasy that only one or two High Victorian designers such as the Scottish Frederick T. Pilkington (1832-98) ever entered. Yet Gaudí’s general inspiration came definitely from the medieval past. In Spain that past included the semi-Islamic Mudéjar, however, and much of the detailing which appears most original to non-Spanish eyes is, in fact, dependent on local precedents of one sort or another. For example, the floral tiles are merely what the Iberian world knows as _azulejos_ and has continued to use down to the present time, especially in Portugal and Brazil (see Chapter 25).

In all the flamboyance of the decoration of the Casa Vicens, the most personal note is in the ironwork. This is naturalistic in theme and bold in scale; it also includes curious linear elements that wave and bend in a way which is more than a little premonitory of the Art Nouveau of the nineties (see Chapter 16). The entrance grille is a masterpiece of decorative art of this period, rivalled only by some of Morris’s contemporary stained glass.

The very utilitarian industrial warehouse for La Obrera Mataronense of 1878-82 at Mataró, with its great arched principals of laminated wood, should be mentioned to balance the Casa Vicens. Here Gaudí’s prowess as an imaginative constructor—almost a straight engineer—was very evident, as also the fact that the unfamiliar forms he continually used—the shape of the arches here was parabolic not semicircular or pointed—were not a matter of personal crankiness but selected for statical reasons: Gothic in theory, that is, like some of Soufflot’s vaulting, though not very Gothic in appearance.

In 1884, however, Gaudí was made director of works for a large new Gothic church in Barcelona, and from this time forward a considerable part of his activity, extending down through his restoration of the cathedral of Palma on the island of Mallorca in 1900-14, was that of a Gothic Revivalist, if an increasingly unconventional one. Towards such a career his own intense religiosity inclined him quite as much as was the case with Pugin and reputedly also with Cuijpers—Viollet-le-Duc, by exception, was strongly anti-clerical. Unlike Pugin’s or Cuijpers’s, however, Gaudí’s career as an ecclesiastical architect was rather unproductive. Yet from the first he designed and executed church furnishings and, while still a student in 1875-7, he assisted the architect Francesc de Paula del Villar i Carmona (1845-1922) on a project for adding a porch to the monastery church of Montsarrat.

In 1881 Villar was made architect of the proposed Expiatory Temple of the Holy Family (Sagrada Familia),[258] for which a large square site had been obtained between the Carrers de Mallorca, de Marina, de Provença, and de Sardenya in an outlying part of Barcelona, and the construction of the crypt of a great cruciform Gothic church was started in 1882. Two years later Gaudí took over charge of the work, as has been said, completing the crypt by 1891 almost entirely according to Villar’s original and quite conventionally thirteenth-fourteenth-century Gothic design. There followed the construction of the outer walls only of the chevet; these were finished by 1893. The further history of the church will be considered later; for Gaudí’s style underwent extraordinary changes in the nineties as he designed and built one transept façade of the church and its towers—which is about all that exists above ground even today (see Chapter 17).

Contemporaneously with Gaudí’s construction of the crypt and the chevet walls of the Sagrada Familia came four secular works, two of them also quite Neo-Gothic in character and two others of very great originality. The Bishop’s Palace at Astorga of 1887-93 and the Fernández-Arbós house, known as the Casa de los Botines, in the Plaza de San Marcelo at León of 1892-4 might well be mistaken for provincial High Victorian Gothic done in England or America twenty or thirty years earlier. But the city mansion of Don Eusebio Güell at 3-5 Carrer Nou de la Rambla (now Conde del Asalto) in Barcelona, built in 1885-9, is an edifice of the greatest distinction, rivalled for quality in its period only by the very finest late work of Richardson in America (see Chapter 13). The Teresian College at 41 Carrer de Ganduxer in Barcelona is also quite remarkable in its simpler way.

Far suaver than his earlier Casa Vicens, the Palau Güell is quite as strikingly novel all the same. At the base yawn a pair of parabolic arches, their tops filled above a plain reticulated grille with sinuous seaweed-like ornament of the most extravagant virtuosity (Plate 96B). The ‘Dragon Gate’ of the Finca Güell of 1887 in the Avenida Pedralbes is still stranger, with a nightmare quality which those of the house in town happily lack. On either side of the entrance arches and in the projecting first storey the façade of the Palau Güell is no more than a rather plain rectangular grid of stone mullions and transoms. In scale this grid is more like Parris’s Boston granite fronts of the twenties than like English window-walls, but it is detailed in a cranky medievalizing way that is more comparable to Webb’s handling of stonework (Figure 17). The rear façade towards the court includes in the middle a broad bay-window with curved corners protected by sunscreens as original but less fantastic than the grilles at the entrance. The most extraordinary features of the exterior, however, are the chimney-pots rising in profusion above the flat roof like an exhibition of abstract sculpture and entirely covered with a mosaic of irregular fragments of glass, rubble, or coloured tiles. In them the extravagance of his earlier houses was continued, and such terminal features remained characteristic of all his later secular work.

[Illustration:

Figure 17. Antoni Gaudí: project for Palau Güell, Barcelona, 1885, elevation ]

The interiors of the Palau Güell are extremely sumptuous. There is much use of marble arcades of parabolic arches carried on round columns, both arches and columns being detailed with the greatest mathematical elegance and simplicity, yet with considerable variety. Some of the ceilings are of marble slabs carried by visible iron beams, but in the principal apartments there are incredibly elaborate confections of woodwork in the Moorish tradition.

The College of Santa Teresa de Jesús, built in 1889-94 immediately after the Palau Güell, is naturally much more modest than that great merchant’s palace, which continues the line of those that late medieval and Renaissance magnates often built. Rubble walls banded and stripped with brickwork are pierced alternately with ranges of narrow windows and with small square ventilators closed with quatrefoil grilles. The widely spaced windows are capped with steep parabolic ‘arches’ formed by cantilevering inward successive brick courses. The third storey is all of brickwork panelled with blind ‘arches’ between the windows and carried up into large, flat, triangular finials along the skyline. Less ingratiating than the Palau Güell with its luxurious use of fine materials inside and out, this college building is equally regular in composition and no more Gothic in appearance to a non-Spanish eye; in fact, however, it leans even more heavily on Mozarab and Mudéjar precedent than does the Casa Vicens. A certain amount of relatively plain wrought-iron grillework recalls that at the entrances of the earlier houses.

Only perhaps in England and America did the line of descent from the Gothic Revival lead so far away from the standard medievalism of the mid century in the seventies and eighties. But these early works of Gaudí represent only a part—to most critics the less important half—of his production. For strangeness they can be matched in work of equal consequence within this period only by Sullivan’s earliest commercial façades in Chicago (see Chapter 14). Teulon and Harris had reformed by the seventies; Lamb and Pilkington were forgotten. In character Gaudí’s work of the seventies and eighties could hardly be more different from the mature style of the English Shaw. Yet Shaw, at his occasional best, could compete with Gaudí in the quality of his achievement; while his influence, both at home and in the United States, was of very considerable historical importance, as Gaudí’s was not, even in Spain (see Chapters 12, 13, and 15).

For all that Gaudí was actually represented at the Paris Exhibition of 1878—by a glovemaker’s vitrine!—and later by pavilions designed for the Compañía Transatlántica in the Naval Exhibition of 1887 at Cadiz and in the Barcelona International Exhibition of the following year, his work was hardly known at all except to his compatriots before the nineties. In the mid twentieth century, however, his reputation is still rising, as the flood of new publications of the last decade makes evident. The reasons for this will be suggested later, since they apply chiefly to the work that he did after 1900 (see Chapters 16 and 20).

In the European picture as a whole a less notable shift of direction occurred around 1870 than in England and America. There was naturally continuity in the Vienna of Francis Joseph, since the Imperial government called the tune in Austrian architecture and the King-Emperor’s reign went on without a break—indeed, it lasted for another generation and more. What is surprising is that the end of the Second Empire and the beginning of the Third Republic brought so little change in France. There was, of course, a short hiatus in production like that which followed the fall of the first Napoleon. As around 1820, however, so around 1875 the story picks up again almost as if there had been no break at all. Gradually interest in exposed metal construction, in decline since the fifties, revived; by the time of the Paris Exhibition of 1889 French feats of metal construction, not so much the Galerie des Machines as the Eiffel Tower, became the talk of the world (see Chapter 16).

In the fugue-like composition of nineteenth-century architectural history different themes have differing durations. The English theme of High Victorian Gothic, picked up in any case only by the Anglo-Saxon sections of the orchestra, came effectively to an end with the early seventies; the Second Empire theme, whether it be considered in a specialized sense or in a broader one, was picked up at least selectively by the whole western world and not least boldly by the Anglo-American section; moreover, it continued in most countries, with some modulation, for at least a decade longer than the High Victorian Gothic. Yet both in England and America, the important new themes of the seventies and eighties were rooted not in the Second Empire but in the Victorian Gothic, even though they represent something much more original than mere modulations of that earlier theme.

The third quarter of the nineteenth century is notable for the stylistic diversity of its production. In principle there may, perhaps, be no more difference between Visconti’s and Lefuel’s New Louvre and a Butterfield church than between Nash’s Blaise Hamlet and his terraces around Regent’s Park, to cite merely work by one early nineteenth-century architect. Yet thanks to the fugal character of the general historical development, which meant that new modes were added to the architectural repertory—as they had been at least since the twenties—more rapidly than old modes were dropped, the over-all picture became extremely complicated after 1850. It belies the most valid and idiosyncratic achievements of this period, however, to stress too much its apparently limitless eclecticism.[259] The account given in the last four chapters undoubtedly exaggerates the importance of certain modes, if that importance be measured statistically in terms of quantity of production. Qualitative considerations have led to a drastic selectivity, emphasizing relatively limited but vital aspects of architectural production at the expense of others that were far more ubiquitous but generally very dull. With different criteria of selection, using different standards of architectural quality—attainment of archaeological plausibility, say; or success or failure in the incorporation of new technical developments; or realization of programmatic aims—several very different pictures could be, and indeed frequently have been,[260] given of the architecture of the western world in these decades.

At the expense of emphasizing architectural developments peculiar to the Anglo-Saxon world in this same, possibly unbalanced, fashion the next

## chapter is organized around the career, after 1870, of Norman Shaw,

whose early work in the High Victorian Gothic has already received some attention. The chapter following that centres on the achievement of the American architect Richardson, whose somewhat parallel beginnings have also been described in this chapter.

------------------------------------------------------------------------

## CHAPTER 12

NORMAN SHAW AND HIS CONTEMPORARIES

IN England and America there followed immediately upon the ‘High Styles’ of the fifties and sixties phases of stylistic development that cannot readily be matched in the other countries of the western world. This is true both of the quality of the achievement and also of its significance for what came after. Beginning just before 1870 in England and but little later in the United States, these two phases developed in far from identical ways. In both cases their conventional names, ‘Queen Anne’ and ‘Romanesque Revival’, are misnomers. It was a long time before the Queen Anne of the seventies actually became a revival of early eighteenth-century architecture in the same sense as the Greek, Gothic, or Renaissance Revivals. The supposed Romanesque Revival in America of this period was not very archaeological either. It is therefore less inaccurate to label these modes by the names of their principal protagonists: ‘Shavian’ for Richard Norman Shaw (even though that proper adjective refers more familiarly to George Bernard Shaw) and ‘Richardsonian’ for Henry Hobson Richardson. Shaw, however, shares responsibility for the effectiveness of the mutation away from the High Victorian with other men, notably his early partner Nesfield, Webb, Godwin, and J. J. Stevenson.[261] Of all this group, Shaw was unquestionably the most successful, the most typical, and the most influential, though not the most original.

Except for Pugin, no architect since Robert Adam had so much effect on English—and for that matter also on American—production. Moreover, his influence lasted for some thirty-five years, rather longer than did Adam’s. Yet it is not possible to define the Shavian mode clearly as it is the Adamesque or the Puginian. An architectural Picasso, Shaw had many divergent manners which he developed successively, but of which none—except the High Victorian Gothic—was ever entirely dropped. Each of these manners, down to the very end of his long practice, found in turn a following. His latest and most conspicuous work, the Piccadilly Hotel, built in London in 1905-8 between Piccadilly and the Regent Street Quadrant (Plate 107), is more characteristic of the Edwardian Age of the opening twentieth century than his early church at Bingley is of the High Victorian. Outside church architecture the intervening Late Victorian can hardly be defined better than in terms of his various manners, and even in church architecture he had a real contribution to make, if a lesser reputation than Pearson or Bodley.

Yet Shaw cannot be rated with Soane or Schinkel as a nineteenth-century architect of absolutely the first rank; nor yet with his American contemporary Richardson, even though Richardson’s career came to an end a score of years before his. Shaw’s work reflects all too clearly, despite his own vast and sanguine assurance, the general uncertainties of the years after 1870. Webb, though less successful and famous, eventually had more influence, not so much on English architecture in general as on the more creative and original men of the next generation. The later history of European architecture would be much the same—if not that of American architecture—had Shaw never existed; but the modern architecture that first came into being after 1900 in various countries of Europe owed something directly, and even more indirectly, to Webb. In this way Richardson also has more significance than Shaw, despite his lack of influence abroad, for Sullivan and Wright in America both learned much from him.

Norman Shaw was born in Edinburgh in 1831. Brought early to London, he was taken on in his early teens by Burn, the Edinburgh architect then settled in London, who had so great a success designing Jacobethan and Scottish Baronial mansions for the high aristocracy in the forties and fifties. Shaw also studied at the Royal Academy, winning in 1853 their Silver Medal, and in the next year their Gold Medal, with the award of a Travelling Studentship that took him to Germany, Italy, and France. The project which won him the first medal was a surprising production for its period, and quite without relation to his own High Victorian Gothic work of the next decade that has been described earlier (see Chapter 11). A vast design for a college with central domed block and side pavilions loaded with giant orders, this project is more Vanbrugh-like than Second Empire. In some sense Shaw’s career was to come full circle stylistically; but even in the Gaiety Theatre in the Strand in London of 1902-3 and the still later Piccadilly Hotel he would hardly be as whole-heartedly Neo-Baroque again.

In 1858 Shaw published, as has been mentioned before, what is perhaps the most attractive of High Victorian Gothic source-books, _Architectural Sketches from the Continent_, based on his European studies; doubtless on the strength of this book he became at this time, or shortly after, Street’s principal assistant—chief draughtsman, one might call it—in succession to Webb.[262] There he remained for four years, leaving in 1862 to form a partnership with Nesfield, whom he had first known in the early fifties in Burn’s office. As has already been noted, Nesfield was the son of Barry’s collaborator in garden design for all his major country house commissions. Younger than Shaw, Nesfield had gone to Burn’s office in 1850 a year or two after leaving Eton, and in 1853 had moved to the office of his uncle Anthony Salvin, another successful builder of aristocratic country houses. Nesfield, in this year 1862, issued a book rather like Shaw’s of four years earlier as has been mentioned in connexion with his work for Lord Craven at Combe Abbey. Other aristocrats with whom he had connexions through his father soon began to employ him on more modest jobs.

Building lodges and other accessories to great country estates, and in 1864 one in Regent’s Park where everyone might appreciate his highly personal touch, Nesfield revived in effect the Picturesque Cottage mode of half a century earlier. But the materials he used were more various,[263] including tile-hanging and pargetting, and his designs had a general finesse that was much more craftsmanlike than those of the slapdash Nash and his rivals in this genre (Plate 50A). In Nesfield’s first major work, Cloverley Hall in Shropshire, begun in 1865, several characteristic features appear for which his lodges hardly prepared the way (see Chapter 15). There a tall great hall provided the principal interior, and the areas of mullioned windows in the Tudor tradition were so extensive as to constitute real ‘window-walls’ (Figure 24). His very refined and ingenious ornamentation at Cloverley, some of it of Japanese inspiration, has been mentioned.

Even earlier, in 1862, when Japanese art was just beginning to be an inspiration to advanced painters in Paris and in London and the Japanese Government first sent examples of characteristic work to an international exhibition, Godwin, who was just at that point throwing off the influence of Ruskin, had stripped bare the interiors of his own house in Bristol and decorated them only with a few Japanese prints asymmetrically hung. By 1866 Godwin was designing wallpapers of notably Japanese character for Jeffry & Co. and from 1868 ‘Anglo-Japanese’ furniture for the manufacturer William Watt.[264] But _japonisme_ is only a minor theme of this period,[265] and it hardly influenced Shaw at all.

Half a century earlier the prestige of a ranking novelist, Sir Walter Scott, had helped to launch one of the most popular Picturesque modes, the Scottish Baronial, when he asked Blore to imitate the old Border castles in designing his house at Abbotsford. Now in 1861 Thackeray, a novelist many of whose novels were set, not in the Middle Ages, but in early eighteenth-century England and Virginia, designed for himself a house in Palace Green in London opposite Kensington Palace, much of which is more or less of that particular period. This house echoes the modest red-brick manor houses of the time of Queen Anne on both sides of the Atlantic, but it could hardly be less plausible. At the same time Wellington College by John Shaw (1803-70), which was begun in 1856, was reaching completion in a much richer, almost Second Empire, version of the Wren style of 1700.

The serious adumbration of a Queen Anne mode really began a few years later with a small public commission of Nesfield’s. His lodge at Kew Gardens, designed in 1866 and built in 1867, though simple, is already almost an archaeological exercise in early eighteenth-century[266] brickwork (Figure 18). This Kew lodge he followed up a few years later with a big but remote country house, Kinmel Park near Abergele in Wales, built in 1871-4 though possibly designed a bit earlier. To this we will be returning shortly. Shaw had nothing to do with Kinmel Park, since his partnership with Nesfield came to an end in 1868; that was just after the completion of Cloverley Hall on which he certainly collaborated even if his personal contribution there cannot now be readily distinguished. Already in 1866, before Shaw parted from Nesfield, however, his own career had opened with the designing of the Bingley church (Plate 94A) and of Glen Andred, near Withyham in Sussex, a house of great originality of character (Plate 102B).

[Illustration:

Figure 18. W. Eden Nesfield: Kew Gardens, Lodge, 1867, elevation ]

Glen Andred is little more related to the new Queen Anne mode of the Kew lodge than it is to the Gothic of the Bingley church. It does, however, seem to derive somewhat from earlier Nesfield work, or possibly from Devey. Where the High Victorian Gothic had rejected English precedent in favour of Italian and French models, this first Sussex house of Shaw’s is resolutely regional in character. The tile-hung walls above a red-brick ground storey, the white-painted wooden casements, almost as extensive as the ‘window-walls’ of Cloverley, the loose asymmetrical organization of the massing are all related to a local Sussex and Surrey vernacular of no particular period (Plate 102B). The entrance front is more formal, carefully balanced if not precisely symmetrical, and here the pargetting in the central gable is of Jacobethan character. But the great stair-window and the graceful massing of the tiled roofs, quite in the finest tradition of the Picturesque but handled with a new ease and casualness, are more important elements of Shaw’s first manner, which can be called ‘Shavian Manorial’. The hall across the front between the two projecting wings is modest in size, with the principal living rooms loosely grouped round it. Thus this may be considered an early example of what I have rather clumsily called the ‘agglutinative plan’, but as it was never published the extent of its actual influence must remain uncertain.

There was little logic to Shaw’s regionalism. Already in 1868 he was applying his Sussex vocabulary of materials and forms to the Cookridge Convalescent Hospital at Horsforth near Leeds in stony Yorkshire. In general, however, he kept this manner for work near London, using it even as late as 1894 for a house called The Hallams near Bramley in Surrey. He also introduced tile-hanging on some of his houses in London such as West House, at 118 Campden Hill Road, of 1877 and Walton House in Walton Street of 1885 as well as—rather more appropriately—on the suburban Hampstead house that he built in the same year for Kate Greenaway at 39 Frognal.

Shaw’s first client had been a painter, J. C. Horsley, R.A., for whom he made some alterations in the early sixties and whose son later entered his office. Glen Andred was for another painter, E. W. Cooke, later R.A., and West House was for George Boughton, R.A. Kate Greenaway, better known today than these forgotten academicians, was an illustrator of children’s books much patronized by Ruskin. F. W. Goodall, R.A. (1870), Marcus Stone, R.A. (1876), Luke Fildes, R.A. (1877), Edwin Long, R.A. (1878, and again in 1888), Frank Holl, R.A. (1881), are other successful painters and fellow academicians—Shaw became an A.R.A. himself in 1872 and an R.A. in 1877—for whom he built houses (with the dates of the commissions). All but Goodall’s house at Harrow Weald were either in Melbury Road in Kensington in London or else in Fitzjohn’s Avenue near his own Hampstead house of 1875 at 6 Ellerdale Road. Where the prosperous artists, themselves presumably aping the aristocracy, led, magnates and City men were now quick to follow. The Newcastle steelmaster Sir William Armstrong had Shaw build Cragside near Rothbury in Northumberland for him as early as 1870.

Leyswood, near Withyham in Sussex, begun in 1868 at the same time as the Cookridge Hospital, was one of Shaw’s most influential works (Plate 123). More archaeologically manorial than Glen Andred, it provided a mass of suggestions that English and American architects borrowed again and again over the next twenty years and more. Because of Shaw’s later leadership, it is natural for posterity to note what was new here; contemporaries, used to the wild vagaries of the High Victorian Gothic, saw Leyswood rather as a reaction against the ‘modernism’ of the fifties and earlier sixties. Tile-hung upper storeys and barge-boarded gables, richly half-timbered—the half-timbering a mere sham applied over solid brickwork!—long banks of casements that approach the twentieth-century ‘ribbon-window’ and great mullioned bays providing ‘window-walls’ as extensive as Nesfield’s at Cloverley clothed an interior that was not at all medieval but a more developed example than Glen Andred of the ‘agglutinative plan’ (Figure 19). The main reception rooms were grouped about a central hall, from one side of which rose elaborate stairs arranged in several flights about an open well. Webb had already essayed this sort of planning in a more orderly way at Arisaig begun in 1863 (Figure 23); but it was Shaw’s version, not Webb’s, that was generally imitated (see Chapter 15).

[Illustration:

Figure 19. Norman Shaw: Leyswood, Sussex, 1868, plan ]

Shortly after Leyswood, and following fairly closely its manner although with fewer Late Gothic elements of detail, came the house later called Grim’s Dyke built at Harrow Weald in 1870-2 for F. W. Goodall, afterwards the country house of the composer W. S. Gilbert, and Preen Manor in Shropshire also designed in 1870. Then followed Hopedene, near Holmbury in Surrey, and Boldre Grange, near Lymington in Hampshire, in 1873; Wispers, Midhurst, in Sussex, in 1875; Chigwell Hall in Essex, and Pierrepoint, near Farnham in Surrey, in 1876; Merrist Wood near Guildford in Surrey, and Denham at Totteridge in Hertfordshire, in 1877; and so on down into the nineties.

After their showing each year at the Royal Academy Exhibition Shaw’s brilliant pen-and-ink perspectives of these houses were published photo-lithographically in the professional press; moreover, from 1874 the plans were usually given as well, the first published being that of Hopedene. Not surprisingly these were the most influential of Shaw’s works abroad, providing in the late seventies and early eighties one of the most important sources of the American ‘Shingle Style’ (see Chapter 15). Beside them, moreover, Webb’s more prominent London works of the late sixties, the house for George Howard, later Earl of Carlisle, built in 1868 near Thackeray’s in Palace Green, Kensington, and the small office building at 19 Lincoln’s Inn Fields, also of 1868, appear somewhat cranky and overstudied, still rather too Gothic in detail and lacking the comfortable air of his country-house work. However, the modest London studio-house at 14 Holland Park Road, Kensington, which was designed in 1864 and built in 1865 for Val Prinsep, like Morris and Spencer Stanhope one of the crew of artists who worked on the decoration of the Oxford Union, must have been more like the Red House and Benfleet Hall before it was recurrently enlarged by Webb in the following decades. Another London studio-house for the water-colour painter G. B. Boyce at 35 Glebe Place, Chelsea, which was begun in 1869, is in rather better condition today and quite exemplary in its quiet way despite some changes by Webb and others.

At this point came Nesfield’s Kinmel Park. Shaw and other advanced architects must have been aware of the character of the designs for this house from 1870 or 71, even though it was neither shown at the Royal Academy nor published then, and took some four years to complete. Kinmel is much more complicated stylistically than Nesfield’s Kew lodge of 1866-7, but it offers the next step in the development of the new Queen Anne mode. At first sight it might appear to be related rather to Second Empire work, for the main block on the entrance side is symmetrical, high-roofed, and dominated by a bold central pavilion. Moreover, the detailing of the red-brick façades with their profuse light-coloured stone trim is almost as French of Louis XIII’s time as it is English of Queen Anne’s day. The garden front, which is carefully ordered but not symmetrical, and the service wing to the south, much more loosely composed and with a profusion of small-paned double-hung sash-windows and dormers, are more definitely English and also more original.

Webb had been using such windows and even approaching the Late Stuart vernacular in his houses for a year or two before Kinmel was begun. This was most evident at Trevor Hall (Figure 25), built at Oakleigh Park near Barnet in Hertfordshire in 1868-70, for that modest country house was quite symmetrical in design although almost devoid of any sort of ‘period’ detail, whether Gothic or Late Stuart. To more acclaim, Webb had also been responsible for designing with William Morris a little earlier, in 1866 and in 1867, the Armoury in St James’s Palace and the Refreshment Room in the Victoria and Albert Museum. The former,

## particularly, is a very original masterpiece of nineteenth-century

decoration, hardly at all related to the contemporary High Victorian Gothic, yet reflecting the eighteenth century only as regards the treatment of the wainscoting and the door and window casings (which may be of eighteenth-century date). The Refreshment Room is also very fine and now accessible to the public (Plate 97B).

Just after 1870, while Kinmel was still in construction, the main line of development moved from the country into London. The Education Act of 1870 required the building of innumerable new schools, particularly by the London School Board. Among the architects successful in the first competitions that were held for designs for these schools were E. R. Robson (1835-1917) and J. J. Stevenson (1831-1908); they used a non-Gothic vocabulary in London stock bricks trimmed with red bricks cut or moulded along seventeenth-century vernacular lines.[267] This mode was not unrelated to the more definitely Queen Anne models provided by the Kew lodge and by Kinmel, but the new London schools were more irregular in composition and naturally much more cheaply built. Robson, appointed architect to the London School Board in 1871, soon made this mode the official one for schools in London County and this, of course, before long influenced Board School design nationally.

In 1871 Stevenson, like Shaw a Scot out to make a London reputation, built a new house for himself in what is now Bayswater Road. This he named the Red House, like Morris’s at Bexley Heath of a decade earlier, in order to call attention to the fact that its brickwork was not covered with stucco but exposed like that of the Thackeray and Howard houses in Palace Green. In fact, however, it was built like the Board Schools of brownish stock bricks with red-brick detail elaborately moulded, gauged, and cut in the Late Stuart way. Although Stevenson’s house had little of the real elegance of Kinmel or the natural ease of Shaw’s manors, its novelty and its fairly conspicuous location would have attracted attention in any case. But Stevenson, a very accomplished publicist, saw the advantage of proclaiming for this hybrid mode a name, ‘Queen Anne’, which was evidently no less applicable to Nesfield’s Kew lodge and Kinmel or even to his friend Robson’s schools. Thus was a revival formally launched.

Two new buildings in London by Shaw, begun in 1872 and in 1873, were definitely in the new mode. Only at this point, indeed, does the term Queen Anne begin to make any sense as applied to Shaw’s work. Despite the valid claim to priority that Stevenson made for his Red House in a paper read in 1874 at the Architects’ Conference ‘On the recent reaction of taste in architecture’ in which he claimed the Queen Anne mode was a ‘Re-Renaissance’ (_sic_), and his own relative success from this time on as a fashionable London house-architect, the Queen Anne became Shaw’s from the moment that he first turned his hand to it in 1872. Whether the original idea came to him from Devey or from Nesfield—he had probably worked himself on the drawings for the Kew lodge—or was merely an attempt to outbid a rival Scotsman on the London scene makes no real difference.

New Zealand Chambers, the office building which Shaw erected in 1872-3 in Leadenhall Street in the City, was certainly totally unlike anything the Age of Anne ever saw except for the cut-brick detailing of the pedimented entrance. Boldly projecting red brick piers divided the tall façade into three bays, while between them rose oriel windows broken by ornately sculptured spandrels imitated from the mid-seventeenth-century ones on Sparrow’s House at Ipswich. The small panes and thick white sash-bars of these windows made the scale surprisingly domestic in contrast to the usual boldness of High Victorian commercial work, and the whole composition was effectively tied together by an ornately pargeted cove cornice that ran straight across the top (see Chapter 14). Above this the rather simple range of continuous dormers in the roof was very much in the spirit of the ‘ribbon-window’ bands on his country houses.

So dazzled were contemporaries by the lush exuberance of Shaw’s ornament on the spandrels and the cove that they hardly noticed the way in which the bold articulation of this façade by the brick piers, with the areas between nearly all window, frankly reflecting the internal iron construction, provided most satisfactory lighting for the offices; nor that Shaw, while keeping his scale intimate in all the detailing, was not afraid to stress the verticality of his façade by avoiding emphasis on the storey lines. Only the weaker features of the design—the arbitrary asymmetry of the entrance, the profuse ornamentation, and the underscaling—were generally imitated.

Lowther Lodge, built in 1873-4, a large free-standing mansion in Kensington Gore, still survives—it is now the home of the Royal Geographical Society—as New Zealand Chambers does not. Here the vocabulary of cut and moulded brick is more consistently Late Stuart, although the general composition, with many gables, two tall polygonal bay-windows, quantities of dormers, and tall fluted chimney stacks, is as romantically complex as that of Shaw’s manors in Sussex and Surrey. However, both the front and the rear façades, when studied, will be found to approximate symmetry in their principal portions as does the front of Glen Andred; and the main rooms inside, the hall at the front and the drawing-room behind, are quite symmetrical and have recognizably Early Georgian (rather than specifically Queen Anne) fireplaces and door and window casings, although their grouping is still, so to say, agglutinative.

In a Surrey house of the same date, 1873, like Trevor Hall unhappily demolished, Webb moved rather farther in a similar direction. Joldwynds near Dorking was quite as symmetrical as Trevor Hall but even less Gothic. The vocabulary of tile-hanging on the upper storeys, with weather-boarding in the gables, was as authentically regional as that of Shaw’s nearby houses, but the vaguely eighteenth-century vernacular of the detailing was much simpler than Shaw’s repertory of moulded and cut brickwork at Lowther Lodge.

Nesfield, in designing what is now Barclays Bank in the Market Square of Saffron Walden in Essex, remained more eclectic, staying closer to the manorial mode of Cloverley Hall yet using again various Japanese motifs in the rich decoration. This was built in 1874. Godwin, who had just moved to London with the actress Ellen Terry and was now largely occupied with designing stage sets, developed further in the rooms of their rented house in Taviton Street in 1873-4 the Anglo-Japanese mode of his interiors of ten years earlier in Bristol. In 1874 he also arranged an exhibition of paintings in a similar spirit for his friend the painter Whistler at the new Grosvenor Galleries.[268]

In the mid seventies, however, it was Shaw, not Nesfield or Godwin, who occupied the centre of the architectural stage. In the Convent of the Sisters of Bethany of 1874 in St Clements Road at Boscombe near Bournemouth he disguised his use of concrete, then a relatively new building material, with his familiar Sussex vernacular. He did the same in a slightly later series of designs for cottages made of patented prefabricated concrete slabs.[269] It is worth noting, moreover, that the internal iron skeleton above the bold cantilever on the front of his Old Swan House (Plate 103) of 1876 at 17 Chelsea Embankment in London provides in effect an example of what would later be called ‘skyscraper construction’, since it carries completely the weight of the brickwork of the upper walls; this was a decade before the ‘invention’ of this sort of construction in Chicago (see Chapter 14). Shaw’s interest in technical developments and his enthusiasm for new materials and methods was evidently very great, always provided that he could bend them to his

## particular sort of retroactive pictorial vision. When he built the Jury

House for the Paris Exhibition of 1878 of patent cement bricks, for example, he designed the façade very elegantly in his Late Stuart manner just as if it were of cut and moulded clay bricks. Godwin and Whistler, however, were showing at this same exhibition an Anglo-Japanese room of highly original character in association with Watt the furniture manufacturer.

Shaw’s excellent church of this period at Bournemouth, St Michael’s and All Angels, Poole Hill, begun in 1873, is Late Victorian in the crispness and clarity of its design but less archaeological than those of this date by Bodley. It seems to indicate that he could have made a great reputation as a church builder had he not been absorbed with secular work. But by the seventies secular work once again provided the field of major prestige in England, as it had hardly done since 1840, and so Shaw concentrated on it. Having revolutionized country-house design, he now turned, more definitely than at Lowther Lodge—by its size and open siting more a country house set in the city—to urban and suburban domestic work. In these his conquest was even more complete, at least in England and, as regards the suburbs, in America.

The Old Swan House and its neighbour Cheyne House at the outer end of the Chelsea Embankment, respectively of 1876 and 1875, are both mansions rather than ordinary terrace houses. They also represent a considerably further advance along the road towards a formal eighteenth-century revival than Lowther Lodge. Old Swan House is completely symmetrical, and the upper storeys are also quite regularly fenestrated in the early eighteenth-century way (Plate 103). However, the total effect is still highly Picturesque because of the way these upper storeys are cantilevered forward; from the cantilever depend, moreover, elaborate oriels of much earlier character very similar to those Shaw had introduced at New Zealand Chambers. Such oriels he long continued to employ; they are not only a principal feature of his own house in Hampstead, built in this same year, but also of the much later Holl and Long houses. Cheyne House occupies an irregular curving plot with the entrance in Royal Hospital Road; but Shaw used all his considerable ingenuity to give it symmetrical façades, even though the plan actually has little of the orderliness of that of Lowther Lodge.

If these two Chelsea houses seem to presage an early return to the serenity of Georgian street architecture, Shaw’s J. P. Heseltine house of 1875 at 196 Queen’s Gate in South Kensington unleashed a flood of the most individualistic house-design London had ever seen. Stucco-fronted houses of builders’ Renaissance design were still being erected on contiguous sites when this tall gabled façade rose, totally oblivious of old and new neighbours. Cut brick, moulded brick, terracotta, all of the brightest red, surround very large mullioned windows in a composition that is gratuitously asymmetrical at the base but symmetrical in the upper storeys below the crowning gable. For fifteen years such houses proliferated in the Chelsea, Kensington, and Earls Court districts of western London. The best are by Shaw himself, such as those at 68, 62, and 72 Cadogan Square—the first of 1879, the others of 1882—and those at 8-11 and 15 Chelsea Embankment of 1878-9; but more are by other architects, and the vast majority by builders. In the Chelsea Embankment range River House at No. 3 is by Bodley; Nos 4-6 are by Godwin; and No. 7 is by R. Phéné Spiers (1838-1916), an architect whose Parisian training did not restrain him from following Shaw.

Collingham Gardens of 1881-7 by Sir Ernest George (1839-1922) and his then partner Harold A. Peto (?-1890), a sort of square with variously designed houses, all gabled, opening on to a lawn in the centre, provides a still more complete illustration of what may be called Neo-Picturesque urbanism. Not at all Shavian, the detailing of many of these houses is very similar to that of Cuijpers’s Rijksmuseum and none of it Queen Anne. The contiguous mansions that George & Peto built in 1882 near by in Harrington Gardens, one for W. S. Gilbert at No. 19 (Plate 104B), the other for Sir Ernest Cassel, the banker, are the most elaborate single London examples of their domestic work. The house of the composer of the Savoy Operas approaches very closely the German Meistersinger mode of the period, but the touch is much lighter—intentionally whimsical perhaps?—and both the organization of the whole and the execution of the profuse detail is very superior to what one finds in most contemporary German work (see Chapter 9).

Stevenson’s best and most Shavian houses in London are two that he built in 1878 in partnership with A. J. Adams in Lowther Gardens behind Lowther Lodge; however, those he built at 40-42 Pont Street have a certain interest because the mode that he exploited here is often called ‘Pont Street Dutch’, so ubiquitous is it in this part of Chelsea. This name also emphasizes the characteristic tendency of the late seventies and eighties towards varying the English late seventeenth-century vernacular mode by the introduction of Dutch and Flemish elements of detail, usually executed in terracotta, as George & Peto did in most of the Earls Court houses mentioned above. Thus, by the late seventies, the long-established London tradition of coherent terrace design came to an end. That was, on the whole, a real urbanistic misfortune, however excellent some of the best individual houses by the above-mentioned architects may be.

Shaw’s venture into the suburbs initiated a new domestic tradition of positive value and also a tradition of ‘planning’ that has continued with some modification down to the present, both in England and abroad. At Bedford Park, Turnham Green, then well beyond the western edges of built-up London, Shaw laid out in 1876 and largely designed an early ‘Garden Suburb’ (see Chapter 24), in fact, almost a ‘new town’, similar in some ways to the New Towns of the present post-war period, but without any industries of its own. Small houses, mostly semi-detached, i.e., in pairs, stand in their own gardens, simply and casually built of good red brick with a certain amount of modest Queen Anne detailing. The scheme is very complete, including a church by Shaw that is most ingeniously styled to harmonize with the domesticity of the houses, a club, a tavern, shops, and so forth.[270] Godwin’s assistant Maurice B. Adams (1849-1933) and E. J. May (1853-1941) also worked here, as well as Godwin himself; indeed, some of the best houses are not by Shaw but by Godwin.

With characteristic versatility, while the construction of Bedford Park was proceeding in this simplified version of his middle manner, Picturesque but distinctly anti-Gothic, Shaw was also erecting at Adcote in Shropshire in 1877 a large Tudor manor house in reddish stone. This is notable for its restrained, almost ‘abstract’, detailing and for the tall mullioned window-wall of the hall bay, more than rivalling that of Cloverley Hall. Flete, a still larger house in Devon begun the year after Adcote, is also Tudor. Dawpool in Cheshire, demolished in 1926, was begun in 1882 in much the same mode but was even more extensive and elaborate than Flete. J. F. Doyle (1840-1913) of Liverpool collaborated on this.

The Bedford Park church of 1878, St Michael’s, is more or less Queen Anne, at least not at all Gothic. But at Ilkley in Yorkshire Shaw’s St Margaret’s of the previous year is a remarkably personal essay in the Perpendicular, low and broad and elegantly detailed. In quality this is well above his earlier Bournemouth church and rather more original in its proportions than the standard work of Bodley and his imitators at this time. Somewhat similar, and still more original, is St Swithin’s in Gervis Road in Bournemouth, also of 1877; while All Saints’, Leek, of 1886 carries almost to the point of parody the Shavian stylization of English Late Gothic proportion towards the broad and low—visually, that is; ritualistically they are quite as ‘High’ as Bodley’s.

Next Shaw produced his finest and most creatively conceived church, Holy Trinity, Latimer Road, comparable in quality to his early church at Bingley but wholly different in character. This was built in 1887-9 for the Harrow Mission in a poor district of western London. The interior of Holy Trinity is a single vessel, very broad and moderately low, covered by a flat-pointed wooden ceiling which is tied by vigorous horizontal members of iron cased in wood and heavily buttressed externally (Plate 106A). Behind the chancel, which is no more than a square dais on which the altar is raised, rises an ecclesiastical version of the Shavian window-wall, broad and low like the space it terminates but arched and lightly traceried at the top. The result could hardly be more different from Shaw’s domestic Queen Anne of these years. It is on such things as this church, in which his basic architectural capacities are revealed unconfused by frivolous elaboration of detail, that his claim to high talent, occasionally to genius, must be based.

If Shaw did not cease to design churches while continually extending the range of his secular practice, it is a still more notable testimony to the breadth of his approach that he built in 1879, in Kensington Gore between the Albert Hall and Lowther Lodge—and with a characteristic disregard for both—the first really handsome block of flats erected in London; the first, that is, unless one prefers the Second Empire ones of the late sixties in Grosvenor Gardens. The tall and extensive mass of this block, like that of most of his houses of the period, is extremely picturesque in silhouette because of the very tall and ornate gables that face the Park. But these are quite regularly spaced and the walls below them, with the multitudinous segment-arched, white-sashed windows all evenly phrased in threes, illustrate Shaw’s Queen Anne of the seventies at its most disciplined (Plate 104A).[271]

As has been noted, Shaw was by now the preferred architect of most of his fellow Royal Academicians. Webb had built houses for several of the Pre-Raphaelite painters who were his friends and associates. Less successful and more advanced painters employed Godwin. Small though it is and now much remodelled, the White House in Tite Street round the corner from the Chelsea Embankment, which Godwin built for his friend Whistler in 1878-9, has one of the most original façades of the decade. As its name implies, although all of brick, it was not ‘red’ like Morris’s and Stevenson’s famous houses, but ‘white’ because the walls were so painted,[272] recalling perhaps the white-painted Colonial farmhouses of Whistler’s New England youth. The sparse detail is related in its vaguely eighteenth-century character to the Shavian Queen Anne, but it is much more delicate and linear, indeed almost Late Georgian in inspiration. Most significantly, the composition of the façade as a whole, and even more evidently the asymmetrical placing of the door and windows, owes a great deal to those abstract principles of Japanese art which both Whistler and Godwin had been studying for almost twenty years.

Whistler had to sell his house almost as soon as it was finished in order to pay the costs of his unhappy libel suit against Ruskin, a legal battle in which the Late Victorian and the High Victorian came to violent grips. But Godwin went on to build several more studio houses in Tite Street at Nos 29, 33, and 44 in the next few years and also the Tower House in 1885. Similar, but inferior, is No. 31 by R. W. Edis, which John Singer Sargent later occupied. Also in Tite Street is the commonplace terrace house at No. 16, of which the interiors were decorated by Godwin for Oscar Wilde,[273] the greatest aesthete of them all. Wilde’s influential ideas in this field, carried to America on a lecture tour in 1881-2, were largely derived from Godwin, it may be noted.

When Shaw turned again to commercial work it was to design in 1881 the offices for the bankers Baring Brothers at 8 Bishopsgate in the City of London. This small building was as discreet, as orderly, and almost as domestic as Cheyne House. But the next year, so chameleon-like was his development, he gave the more conspicuous Alliance Assurance Building at the corner of St James’s Street and Pall Mall opposite St James’s Palace broad, low, banded arches of brick and stone below and elaborated the vertical articulation of the upper storeys with profuse sculptural ornament.[274] Very tall and scallopy gables provide a Neo-Picturesque effect only too comparable to the most vulgar ‘Pont Street Dutch’ houses designed by his rivals or even to contemporary Northern Renaissance work on the Continent. To emphasize his variousness further, there is diagonally across the street a later edifice for the same clients, built in collaboration with his pupil Ernest Newton (1856-1922) in 1903, so quietly academic in the Neo-Georgian taste of the early twentieth century that one can hardly believe it is also Shaw’s.

His next important secular works after the first Alliance building, both begun in 1887 like the Latimer Road church, contrast with each other almost as markedly as they do with that. Characteristic of the essentially private patronage—patronage from successful artists, patronage from business, patronage from the professional classes—responsible for the best English architecture of this period is the fact that Shaw’s first public commission came only at this advanced stage of his career. London’s Metropolitan Police Offices in New Scotland Yard, of which the original block was built in 1887-8 and the second block to the south added in 1890, have a splendid site on the Thames Embankment. Remembering, it would seem almost for the first time, his own Scottish birth—or possibly in apposite reference to the familiar name of the London police headquarters—Shaw designed Scotland Yard somewhat like a Scottish castle with corner tourelles and tall curved gables, but using throughout heavy and rather academic later seventeenth-century detailing of a much less regional sort (Plate 106B). Red brick and stone in combination make it also as colouristic as the Alliance building, the solidity of the proportions makes it weighty, and the high gables and tower roofs give it great variety of outline. As a result, the total effect is almost High Victorian in its vigour and its massiveness. Shaw is said to have regretted the need to build a second block; certainly it must have been more impressive when the original block stood alone like an isolated riverside fortress.

Scotland Yard seems to look backward somewhat, at least in relation to that gradual development towards orderliness and restraint of an eighteenth-century sort which can be discerned in Shaw’s work of the seventies despite all its variousness. On the other hand, the house that he built in 1887-8 for Fred White,[275] an American diplomat, at 170 Queen’s Gate, so near to that strikingly aberrant terrace house of the previous decade at No. 196, seems to look forward into the early twentieth century, when the eighteenth-century Georgian would provide the basis for a quite archaeological revival. This plain rectangular block of red brick, orderly and symmetrical on the long façade towards Imperial Institute Road and also on the end towards Queen’s Gate, with three ranges of large sash-windows below an academic cornice, is therefore as much a historical landmark, if not an original creation, as was Glen Andred twenty years before (Plate #105:pl105). The suave and well-scaled ornamentation is concentrated at the doorway in the eighteenth-century manner, and the hip roof is unbroken except by regularly spaced dormers. Yet, curiously enough, the plan is somewhat less completely regular and symmetrical than one might expect from the exterior; for example, the large drawing-room towards Queen’s Gate is L-shaped.

Only the excellence of the craftsmanship here, based not on the Sussex vernacular but on the most sophisticated work of around 1720, the prominence of the tall chimneys, and the wide central dormer with its curved top reveal Shaw’s hand and suggest, perhaps, an early date; otherwise such a house might well have been built forty years or so later by many other architects, English and American (see Chapter 24). However, Webb at Smeaton Manor[276] in Yorkshire, built in 1877-9, had already arrived at an almost identical regularity and formality of design (Plate 102A). Characteristically, however, he did not elaborate the exterior with borrowed eighteenth-century detailing, and the house remains almost undatable on internal evidence, like much of his best work.

Scotland Yard is an all but unique example of an English public building of distinction erected in the eighties. Before continuing with the account of Shaw’s work in the nineties, two prominent features of the London skyline, the most striking additions made since Butterfield’s spire of All Saints’ rose in Margaret Street in the fifties and the Victoria Tower of the Houses of Parliament was completed in the sixties, should be mentioned. Both the Imperial Institute, towering over Shaw’s contiguous Fred White house in South Kensington, which was built in 1887-93 in honour of Queen Victoria’s first jubilee, and the Catholic cathedral of Westminster, not begun until 1894, are especially notable for their very tall dome-topped towers. The cathedral, which was designed by J. F. Bentley (1839-1902), a pupil of Clutton, has also a magnificent domed interior. The Institute, built by T. E. Collcutt (1840-1924), was perhaps of less over-all interest but extremely refined and elegant in its detailing compared to the contemporary work of George & Peto, which it most closely resembles. Curiously enough, the very underscaled membering and even so dainty a trick as the use of single courses of red brick here and there in the stonework does not make the 280-foot tower petty. It may be compared to its own very great advantage with Haller’s contemporary tower, in a somewhat parallel Northern Renaissance vein, on the Hamburg Rathaus. Collcutt’s own earlier tower on the Town Hall at Wakefield in Yorkshire of 1877-80 was less successful than this London landmark, which has happily survived the rest of the building.

Bentley’s tower has a similar silhouette, but is more boldly striated by broad bands of brick and stone. The detail, partly Byzantine, partly Early Renaissance despite his distinguished early career as a Late Victorian Gothic church architect, is, like Collcutt’s, rather underscaled. This goes still further to prove the extent to which this period in England saw all architecture, even that of cathedrals, in domestic terms. However, well before Bentley began his cathedral—it is not even yet completed as regards the internal decoration—Shaw had turned towards considerably more monumental forms at Scotland Yard, and even to quite academic design.

At Bryanston, a large country house in Dorset begun in 1889 for Lord Portman, Shaw modelled the main block on Sir Roger Pratt’s Coleshill House of the mid seventeenth century; the side wings here are quite Gibbsian. This is the earliest example of what the English call ‘Monumental Queen Anne’—to distinguish this sort of work henceforth from the freer and more vernacular Queen Anne of the seventies and eighties—and the Americans ‘Georgian Revival’. Two years later Shaw built Chesters in Northumberland. This mansion is equally academic, if less derivative from particular sources; but it is also highly original in plan and conception. The composition of the incurved façade planes, moreover, is as knowing and as ingenious in its formal way as anything he ever built in a more rambling vein.

Later in the nineties Shaw’s stylistic uncertainty—or, if one wishes to call it so, his versatility—was notably illustrated in two large commercial buildings built in Liverpool. The façade of Parr’s Bank in Castle Street, built in 1898 in collaboration with W. E. Willink (1856-1924) and P. C. Thicknesse (1860-1920), is of the suavest academic order. Its proportions are surer than in any of his other works except Chesters, and yet he striated its light-coloured stonework with bands of green marble in a way few later architects working in this vein would ever have thought of doing. Two years later, in the offices that he built in collaboration with Doyle for Ismay, Imrie & Co., later the White Star Line—for whom he also designed the interiors of the liner _Oceanic_—he provided what was externally almost a copy of Scotland Yard, and yet inside he exposed the riveted metal structural members in a fashion at once frank and highly decorative.

If Shaw had had the opportunity to rebuild Nash’s Regent Street Quadrant completely according to the designs that he prepared in 1905 the loss of the original work might not be so serious. Approaching seventy-five, he turned here to a Piranesian Classicism. The colonnaded section finished in 1908, which forms the northern front of the Piccadilly Hotel, though flanked at both ends by an emasculated version of Shaw’s design carried out in 1923 by his disciple and biographer Sir Reginald Blomfield (1856-1942), rivals in boldness anything English architecture had produced since the days of Vanbrugh and Hawksmore. Even more spectacular, and also incomplete, since the gable at the east end was never built, is the Piccadilly façade of the hotel with its tremendous open colonnade raised high against the sky (Plate 107). The Classical serenity of this feature is characteristically contrasted with the voluted silhouette of the tall gable over the projecting wing at the west end, and the exuberance of the whole puts most other Edwardian Neo-Baroque to shame.

To summarize Shaw’s achievement or even to epitomize his personal style is almost impossible. He was, for example, in no ordinary sense of the word merely an eclectic; yet his modes were very various, more various than those of almost any other nineteenth-century architect of equal rank. After his first borrowings from Nesfield, however, they were all his own—his own, at least, until hordes of other architects in England and America took them up, one or two at a time, often vulgarizing them beyond recognition. He was probably not the most talented English architect of his generation and certainly not the most original. How much he owed to Nesfield at the start it is impossible to estimate, even though at least two of the characteristic Shavian modes seem to have been originally of his invention—if not, indeed, of Devey’s!

Yet ironically Nesfield’s own later work appeared to contemporaries almost like an echo of Shaw’s if it was known at all. He never had any such success as did Shaw, and died relatively young in 1888. Godwin also was somehow never able, after 1870, to repeat the public triumphs that had been his in the competitions of the early sixties. In his later life he turned more and more to designing sets and costumes for the theatre and died in 1886, two years before Nesfield. Webb lived on till 1915, although he retired from practice in 1900; his spirit, moreover, lived on in a quite different way from Shaw’s. It was through emulation of the craftsman-like integrity of Webb’s work that the attitudes, rather than the forms, of Pugin’s earlier Gothic Revival were transmitted to the first modern architects quite as much as through study of the writings of his friend and close associate Morris.

------------------------------------------------------------------------

## CHAPTER 13

H. H. RICHARDSON AND McKIM, MEAD & WHITE

THE story of Shaw’s career is a fascinating one, far more interesting in fact than the general history of English architecture in the last quarter of the nineteenth century. It was a success-drama in four or five acts, of which the last was by no means the least brilliant. Richardson’s career was less eventful, even though, at its peak in the mid eighties, it was at least as successful as Shaw’s. It was also incomplete, since death brought his production to an end at that peak when he was only forty-eight. Yet Richardson’s achievement must be considered greater than Shaw’s, qualitatively if not quantitatively, because his work was better integrated and his development more intelligently directed. Moreover, his influence operated on two levels: on one it was as wide, if more evanescent, than Shaw’s—say, what Shaw’s might have been if _he_ had died at the age of forty-eight, that is, in 1879—on another level it was more like that of Webb, affecting deeply several of the most creative American architects of the next two generations.

Henry Hobson Richardson was born in 1838 near New Orleans in Louisiana. Upon graduation from Harvard in 1858 Richardson, bilingual on account of his Louisiana birth, not unnaturally proceeded to Paris to the École des Beaux-Arts, entering there the atelier of L.-J. André (1819-90), a pupil of Lebas who had become a professor at the École in 1855. But after two years the outbreak of the Civil War in the United States cut off his remittances from home and he had to find work in order to maintain himself. His experience in the office of Théodore Labrouste, notably in working on the designs for the Asile d’Ivry outside Paris, was perhaps of more ultimate value to him than what he learned in André’s atelier and at the École. Several of his earliest works in America, designed immediately after his return from Paris in 1865, have been discussed already (see Chapter 11). It was with the Brattle Square (now First Baptist) Church on Commonwealth Avenue at Clarendon Street in the new Back Bay residential district of Boston, the commission for which he won in a competition held in 1870, that his career seriously began. During the years that this was in construction, 1871-2, he had in his office a young assistant, Charles F. McKim (1847-1909), who had returned from Paris at the outbreak of the Franco-Prussian War in 1870. It may well be that the forceful McKim helped Richardson to crystallize the divergent elements evident in his earlier work into a coherent personal style. The Brattle Square Church somewhat resembles in its round-arched medievalism such a Paris church of the sixties as Vaudremer’s Saint-Pierre-de-Montrouge, which Richardson himself may have seen and admired in the early stages of its construction. But the squarish T-shaped plan, without aisles but with transepts, would have been as unusual in France at this period as in England. The material is the richly textured Roxbury Puddingstone rising in broad plain surfaces to the medium-pitched gables. The detail strikes a sort of balance between the French Romanesquoid and the English High Victorian Gothic, the forms being more French, the execution more English. The varied polychromy of the deep voussoirs of the arches is certainly English, but with a personal note in the great variety of the coloured banding. The corner placing of the tall tower, with its fine frieze by the French sculptor Bartholdi, is English in spirit, but its shape is rather more campanile-like than any English church tower had been since the forties.

A similar stylistic crystallization can be seen in the very extensive plant of the State Hospital at Buffalo, N.Y., a commission also won by Richardson in competition in 1870. This was largely re-designed before construction began in 1872 and was in building throughout the whole decade. It was, functionally, the sort of commission for which Richardson’s French training best prepared him, and the planning is French. The other sources of the design seem to have been mostly English, particularly the projects of Burges.

Two buildings in Springfield, Mass., where Richardson had been working on and off since his return from Paris, are even more significant than the Buffalo asylum for the rather definite evidence they offer as to his chief contemporary sources of inspiration at this point. The spire of the North Congregational Church there—commissioned as early as 1868, but built in 1872-3, after being re-designed in 1871 or 72—is a rather squat pyramid of quarry-faced brownstone with four corner spirelets rising from the same square base, apparently a version of the spire Burges designed for his Skelton church begun in 1871 or that of Street’s St James the Less. The tower of the Hampden County Courthouse of 1871-3 also comes from Burges, in this case from the project that he entered in the London Law Courts competition of 1866. The general composition owes more to the slightly earlier English town halls at Northampton and Congleton by Godwin, who was also Burges’s collaborator on the Law Courts project. But the magnificent scale of the random ashlar walls of quarry-faced Monson granite, their coldness relieved by bright red pointing, is as personal to Richardson as the similar brownstone masonry of the North Church and the Buffalo Hospital.

Richardson’s American Express Building,[277] his first work in Chicago, which was begun in 1872, and his contemporary Andrews house in Newport, R.I., both showed comparable evidence of generic influence from contemporary England (see Chapters 14 and 15:ch15#). In this same year, 1872, Richardson won the competition for Trinity Church[278] in Boston, which was to occupy a conspicuous site on the east side of Copley Square, the principal open space in the new Back Bay district. Preceding by a year the Panic of 1873, which slowed building almost to a standstill, this commission and that for the Buffalo Hospital kept him busy through five lean years. As Trinity rose to completion over the years 1873-7, this big Boston church established Richardson’s reputation as the new leader among American architects (Plate 108A). Even before Trinity was finished others were producing crude imitations of it; and over the next twenty years many prominent American churches,

## particularly in the Middle West, followed in some degree the paradigm

that it provided.

Trinity is in plan an enlarged and modified version of the Brattle Square Church. A deep semicircular chancel provides a fourth arm, and a great square lantern rises over the crossing. The elaborate porch, so archaeologically Provençal Romanesque, was added by Richardson’s successors, Shepley, Rutan & Coolidge, in the nineties, as were also the tops of the western towers; the present decorations of the chancel are much later and by Charles D. Maginnis (1867-1955).

The materials of Trinity are pink Milford granite in quarry-faced random ashlar for the walling and the Longmeadow brownstone that he had first used on the Unity Church in Springfield for the profuse trim. The detail changed in character as the work proceeded; in the earliest portions executed it is heavy and crude, with the foliage carved in a naturalistic High Victorian Gothic vein. But the logic of the round arches that Richardson had been consistently using since he designed the Brattle Square Church in 1870 led him to study Révoil’s _Architecture romane du midi de la France_,[279] and such a characteristic feature as the polychromy on the outside of the apse is specifically Auvergnat. Moreover, the executed lantern was rather closely based on that of the Old Cathedral of Salamanca in Spain—a model that Richardson’s assistant Stanford White (1853-1906), who succeeded McKim in his employ in 1872, seems to have suggested.

Most contemporaries, supposing all worthy nineteenth-century architecture to be necessarily derivative from this or that style of the past, believed that Richardson had initiated a Romanesque Revival here. But Richardson remained really as responsive to contemporary English ideas as he had been earlier. For example, the curious double-curved wooden roof with kingpost trusses derives from published examples of similar roofs built or projected by Burges. Equally symptomatic of English influence is the use of stained glass by Morris and Burne-Jones in the north transept windows. That glass, however, is inferior in richness of tone to the small windows in the west front designed by the American artist John LaFarge. LaFarge was also responsible for the painted decoration on the walls and the roofs.

To take over Fuller & Laver’s New York State Capitol at Albany when already partly built in the way that Richardson and Eidlitz—a foreign-born exponent of Romanesque of the earlier _Rundbogenstil_ sort, it will be recalled—were asked to do in 1875 was a thankless job; but this call for Richardson’s aid illustrates the rapidity with which he achieved a national reputation. More important, both historically and intrinsically, than what he was able to carry out in Albany—chiefly the Senate Chamber—were a second house that he built in Shepard Avenue in Newport, R.I., in 1874-6 and a building in Main Street in Hartford, Conn., of 1875-6 (see Chapters 14 and 15). The Sherman house is the first example of a Shavian manor successfully translated into American materials; the Cheney Block (now Brown-Thompson Store) is not Shavian at all, but very similar to the arcaded façades common in England since the late fifties (Plate 116A).

To the late seventies belong two remarkably fine buildings, still obviously related to slightly earlier English work, but more personal than either the Newport house or the Hartford commercial building. With the Winn Memorial Library in Woburn, Mass., of 1877-8 Richardson initiated a line of small-town public libraries that reached its climax in the Crane Library in Quincy, Mass., of 1880-3 (Plate 110). The high window-bands of the stack wings, a monumental stone version of Shaw’s ‘ribbon-windows’, and the stone-mullioned ‘window-walls’ at the ends are more significant than the round stair-turrets and the cavernous entrance arches—Early Christian from Syria[280] in origin, not Southern French Romanesque, it should be noted—that romanticize their generally compact massing. The highly functional planning is asymmetrical yet very carefully ordered, perhaps the one remaining trace of his Paris training.

In building Sever Hall, a classroom building for Harvard College in Cambridge, Mass., in 1878-80 Richardson abandoned rock-faced granite and brownstone, materials whose common use would, a little later, mark the extent of his influence on other architects, for the red brick of the nearby eighteenth-century buildings in the old Harvard Yard. He even imitated the plain oblong masses of these Georgian edifices under his great red-tiled hip-roof; but the front, with its deep Syrian arch and two tower-like rounded bays, and the rear, with a broader and shallower central bow, are wholly Richardsonian. There is a rather Shavian pediment over the centre of the front, however; while the moulded brick mullions of the banked windows and the very rich cut-brick panels of floral ornament seem to reflect current English work by Stevenson and by Godwin as well as by Shaw. Yet the whole has been amalgamated into a composition quite as orderly as anything the English ‘Annites’ had produced. At the same time Sever Hall is almost as vigorous and manly in scale as his contemporary libraries of granite and brownstone.

Two domestic buildings of 1880, one entirely shingled, the other of rough glacial boulders, are even more personal works; and both,

## particularly the former, represent the American domestic mode of this

period now called the ‘Shingle Style’ (see Chapter 15). The John Bryant house in Cohasset, Mass., of 1880 first illustrated his emancipation from the direct Shavian imitation that had begun with the Sherman house and continued in several projects—probably mostly White’s work in actual fact—that were prepared in the later seventies but never executed. Quite a series of later shingled houses by Richardson followed the Bryant house between 1881 and 1886 (Plate 124B).

The contemporary Ames Gate Lodge[281] in North Easton, Mass., has a sort of antediluvian power in the bold plasticity of its boulder-built walls—a theme exploited once before in Grace Church in Medford, Mass., of 1867 it will be recalled—as remote from the Romanesque as from the Queen Anne. A similarly absolute originality of a more gracious order can be seen in the Fenway Bridge of 1880-1 in Boston; its tawny seam-faced granite walls happily echo the easy naturalistic curves of the landscaping by his friend F. L. Olmsted (1822-1903),[282] of which it is a principal feature.

1881 saw the initiation of a more monumental building for Harvard, Austin Hall,[283] then the Law School, which was completed in 1883. Rich Auvergnat polychromy and a great deal of rather Byzantinesque carved ornament somewhat confuse the direct structural expressiveness of the thoroughly articulated masonry walls; as a result Austin Hall provided a multitude of decorative clichés for imitators to abuse. Much more modest and also much more significant was the station at Auburndale, Mass., also of 1881, built for the Boston & Albany Railroad. This was the first and the finest of a series of small suburban stations notable for the simplicity of their design and for the compositional skill with which the open elements, carried on sturdy but gracefully shaped wooden supports, were related to the solid masonry blocks of granite and brownstone beneath sweeping roofs of tile or slate. If Shaw was called on in the nineties to design the interiors of an ocean liner for the White Star Line, Richardson had already provided in 1884 a railway carriage for the Boston & Albany. This was neither Romanesque nor Queen Anne in inspiration, but had domestically scaled interiors lined with small square oaken panels and no carved ornament of any sort.

Stations, libraries, and houses form the bulk of Richardson’s production from 1882 until his death. But two much larger buildings, which he himself judged to be his master works, were also fortunately initiated, one in 1884 and the other in 1885, well before his last illness began, though both had to be finished by his successors Shepley, Rutan & Coolidge after his death. The Allegheny County Buildings[284] in Pittsburgh, Penna., consist of a vast quadrangular courthouse dominated by a very tall tower that rises in the centre of the front and a gaol across the street to the rear. Except for the courtyard walls, interesting for the variety and the openness of their ranges of granite arcading, the courthouse offers on the whole only a sort of summary of his talents; the detail, above all, is afflicted with an archaeological dryness that must be due to the increasing dependence of his assistants on published documents of medieval carving. The courthouse provided, however, the model for many large public buildings in the next few years. Among these, the City Hall in Minneapolis, Minn., begun by the local firm of Long & Kees in 1887, is not unworthy of comparison with the original, particularly as regards the tower. That of Toronto in Canada, built by E. J. Lennox in 1890-9, is less interesting but even more monumental; it also signalizes the supersession of English by American influence in Canadian architecture at this point, as does the almost equally Richardsonian Windsor Station in Montreal begun by the American architect Bruce Price in 1888.

The Pittsburgh Jail is a masterpiece of the most personal order, Piranesian in scale, nobly expressive of its gloomy purpose, and as superb an example of granite masonry as exists in the world (Plate 108B). It epitomizes Richardson’s genius where the courthouse merely summarizes his talents.

Richardson’s highest achievement, however, was in the field of private building not in that of the public monument. By a happy coincidence his ultimate masterpiece rose in Chicago where, at this very moment, technical advances in construction were being made that would soon bring to a climax the whole story of nineteenth-century commercial architecture (see Chapter 14). Chicago retains Richardson’s last great masonry house, that of 1885-7 for J. J. Glessner, almost as perfect a domestic paradigm of granite construction as the Pittsburgh Jail. To her shame, however, Richardson’s Marshall Field Wholesale Store, built during the same years, was torn down a generation ago to provide a car park.

The Field store occupied an entire block with a dignity and a grandeur no other commercial structure had ever attained before (Plate 116B). Internally it was of iron-skeleton construction; externally the arcaded masonry walls represented a development from those of the Cheney Building of ten years earlier (Plate 116A). Segmental arches covered the broad low openings in the massive ground storey, all built of great ashlar blocks of rock-faced red Missouri granite. The next three storeys, built of brownstone, were combined under a single range of broad arches, yet also articulated within these arched openings by stone mullions and transoms. Above this stage the rhythm doubled, with the windows of the next two storeys joined vertically under narrower arches. The scale of the quarry-faced ashlar was graded down as the walls rose, quite as were the window sizes, and the non-supporting spandrels were filled with small square blocks. The full thickness of the bearing masonry walls was revealed at all the openings. Finally there came a trabeated attic of somewhat Schinkel-like character over which appeared almost the only carved detail on the building, a boldly crocketed cornice. That was ‘Early French’, i.e., of twelfth-century Gothic rather than Romanesque or Byzantine inspiration.

The result was a monument as bold and almost as Piranesian in its scale and its forcefulness as the Pittsburgh Jail; but the walls were also as open, as continuously fenestrated, as those of the court of the Pittsburgh Courthouse. The logical and expressive design of commercial buildings with walls of bearing masonry could hardly be carried further. But in the very year that the Field Store was finished Holabird & Roche, in designing the Tacoma Building, also in Chicago, first showed how the exterior of such edifices might express instead a newly developed sort of construction that allowed the internal metal skeleton to carry the external cladding of masonry (see Chapter 14).

In one last commercial building, much more obscurely located and built of far less sumptuous materials, which was started just before Richardson’s death—it was only commissioned after his last illness had begun—he carried the logic of the design of the Field Store one step farther. It was almost as if he had already sensed, like Holabird & Roche, the implications of the Home Insurance Building in Chicago of 1883-5 by their former employer William Le Baron Jenney, in which the new sort of construction was first used but not at all expressed. On Richardson’s Ames Building in Harrison Avenue in Boston a tall arcade rose almost the full height of the wall beneath a machicolated attic; the depth of the reveals around the sash at the sides of the brick piers was minimized; and above the ground storey the spandrels were of metal panels set almost flush with both piers and sash.

When Richardson died in 1886 the evidence of his great late works indicates that his powers were at their highest. His office, moreover, had never been busier. How Richardson might have developed further it is impossible to say. In the hands of his imitators the Richardsonian mode did not grow in any very creative way during the decade or more that it continued a favourite for churches, public buildings, and even houses built of masonry. Those who had been closest to Richardson when his style was maturing, McKim and White, rarely imitated him; even before his death, in fact, they had already set under way a reaction against the Richardsonian. Their buildings and not his provide the real American analogue to the later work of Shaw in England. Moreover, their leadership succeeded his in many professional circles from coast to coast almost before he was dead.

Leaving aside the modes inherited from the sixties, in any case transmuted almost beyond recognition by the early eighties if not yet entirely superseded, there were at the time of Richardson’s death three main currents in American architecture as against the four or five more or less Shavian modes then popular in England. One was the Richardsonian.[285] This was practised with some success by various Boston firms such as Peabody & Stearns and Van Brunt & Howe. It had been carried to Kansas City, Missouri, by Van Brunt, moreover, and it was being developed with some originality by other Middle Westerners such as George D. Mason (1856-1948) in Detroit, D. H. Burnham (1846-1912) and his partner J. W. Root (1850-91), H. I. Cobb (1859-1931) and his partner Frost, and several other firms in Chicago. The very able designer Harvey Ellis (1852-1904),[286] working for L. S. Buffington (1848?-1931) in Minneapolis, should also be mentioned. Another current was represented by the development leading towards the Chicago skyscrapers of the nineties, in Richardson’s last years more in the hands of technicians than of architects (see Chapter 14).

The third, and for the next few years the most expansive, current was what can already be called the Academic Reaction. This was parallel to, yet already pushing well ahead of, Shaw’s somewhat coy approach to a programmatic revival of eighteenth-century forms; and McKim, Mead & White were its acknowledged leaders.[287] During the years that White was working for Richardson he seems to have been devotedly Shavian. Certain unexecuted house projects from the Richardson office which White signed, done for the Cheney family of Manchester, Conn., the clients for Richardson’s Cheney Block in Hartford, make this particularly evident. When White replaced Bigelow in the firm of McKim, Mead & Bigelow, on his return from the European trip that he took after leaving Richardson in 1878, he found McKim designing Shavian houses with a considerably less sure decorative touch than his own. The McKim, Mead & White country houses that followed, however, such as that for H. Victor Newcomb in Elberon, N.J., of 1880-1 (Plate 125A), that for Isaac Bell, Jr, in Newport, R.I., of 1881-2 (Plate 126), and that for Cyrus McCormick in Richfield Springs, N.Y., of the same years, represent in several ways a real advance over Richardson’s Sherman house.[288] Such an advance is equally to be observed in various houses built around Boston in these years by W. R. Emerson (1833-1918) and by Arthur Little (1852-1925), the very earliest of which doubtless influenced Richardson when he designed the Bryant house (see Chapter 15).

For McKim, Mead & White’s Tiffany house in New York of 1882-3, all of tawny ‘Roman’ brick with much moulded brick detail, the inspiration was largely Shavian also; only the rock-faced stone base and the broad low entrance arch were at all Richardsonian. In the New York house that they began the next year, however—really a group of houses arranged in a U around an open court across Madison Avenue from the rear of St Patrick’s Cathedral—for the railway magnate Henry Villard an entirely different, even quite opposed, spirit appears (Plate 109B). The Villard houses, although on Villard’s insistence still built of brownstone rather than of light-coloured limestone, are as much a High Renaissance Italian _palazzo_ as anything Barry or his contemporaries on the Continent ever designed in the preceding sixty years. Reputedly Joseph M. Wells (1853-90), an assistant in the McKim, Mead & White office who later refused membership in the firm, was responsible for the decision to follow Roman models of around 1500, most notably the Cancelleria Palace, as that was known to him—he had never been abroad—through the plates of Letarouilly’s _Édifices de Rome moderne_.

This type of design represented a conscious reaction against the Neo-Picturesque, whether Richardsonian, Shavian, or _François I_, a return to formal order of the most drastic sort. It represented also a return to close archaeological imitation of a style from the past such as had ended in America, on the whole, with the decline of the Greek Revival a generation earlier. Curiously enough this turn was also something of a declaration of independence from Europe, since the American Academic Reaction as initiated in the design of the Villard houses seems to have had no contemporary sources abroad. However much Shaw’s Queen Anne had, for about a decade, been moving towards an equivalent formality—of a more eighteenth-century sort—Shaw had neither gone as yet so far in this direction nor did he ever turn to the High Renaissance for his models. Continental parallels in the eighties are not hard to find in the work of such men as Balat in Belgium, Koch in Italy, and Wagner in Austria; but their current production was probably not known in the United States, whose foreign relations in architecture had always been largely restricted to England, France, and Germany.

This American return to order was at first more significant for its absolute aspect than for its archaeological bent. Although McKim, Mead & White used a Renaissance arcade at the base of their Goelet Building erected in Broadway at 20th Street in New York in 1885-6, the upper storeys of this modest skyscraper offer a very free, and at the same time a highly regularized, expression of the hive of offices behind, and even of the metal grid of the internal skeleton. Certain houses by McKim, Mead & White in New York of these years were even freer from the imitation of specific Italian precedents; while their Wm. G. Low house of as late as 1886-7, on the seashore south of Bristol, R.I., is a masterpiece of the ‘Shingle Style’ despite the tightness and formality of its plan (see Chapter 15). Carefully ordered under its single broad gable, which even subsumes the veranda at the southern end, the Low house is yet quite without reminiscent detail or, indeed, much of any detail at all (Plate 127). In a group of small houses at Tuxedo Park, not at all academic in their exterior treatment, Bruce Price (1845-1903) was reorganizing the open plan of the Americanized Queen Anne in a schematically symmetrical way at just this time also (Plate 125B; Figure 28).

The possibility of a revival of the American Colonial and Post-Colonial in all their successive phases from the medievalism of the seventeenth-century origins to what can be called the ‘Carpenters’ Adam’ of 1800 had been in the air ever since the early seventies, when McKim had added a Neo-Colonial room to a real Colonial house in Newport, R.I. In the local Colonial architecture Americans found obvious parallels to the seventeenth- and eighteenth-century precedent that Shaw was exploiting in England.[289] The ‘Shingle Style’ employed various features and treatments—such as the all-over covering of shingles itself—that recall American work of the periods before 1800. But because of the continued strength of inherited Picturesque ideals there was no programmatic imitation of formal eighteenth-century house design before the mid eighties. Even such a highly orderly example as Little’s Shingleside House at Swampscott, Mass., of 1880-1 was still quite un-archaeological. Interestingly enough, this seems to have been about the first up-to-date American house to be published in a foreign magazine[290] since the _Allgemeine Bauzeitung_ in 1846 presented examples of Greek Revival terrace-houses in New York.

Following on the completion of the Bramantesque Villard houses in New York in 1885, McKim, Mead & White built in Newport, R.I., in 1885-6 the H. A. C. Taylor house, lately destroyed, which was as Neo-Georgian, in its American Colonial way, as the Fred White house Shaw began in London two years later. For this the American architects adopted the symmetrical Anglo-Palladian plan of the mid eighteenth century and capped the resultant rectangular mass with the special gable-over-hip roof of Colonial Newport. Elaborately embellished with Palladian windows and with much carved detail of a generically Georgian order, the Taylor house provided a new formula of design for domestic work that soon superseded almost completely the ‘Shingle Style’. From the Taylor house stems that mature Colonial Revival which was to last longer in the end in America than had the Greek Revival.

Down to the early nineties, however, McKim, Mead & White were rarely so programmatic in their Neo-Colonial work, and their principal public building of the late eighties, the Boston Public Library, was entirely Italianate (Plate 111). In 1887 they were commissioned to build this major monument on the west side of Copley Square. There it was to face the Trinity Church that had initiated the Richardsonian wave more than a decade earlier—a monument in whose designing, moreover, both McKim and White had actually participated. The Library as built in 1888-92 was a major challenge to the Richardsonian, at least as contemporaries then generally understood and employed what they thought was Richardson’s mode. The contrast it offers to the church opposite is almost as great as to the prominent but low-grade High Victorian Gothic structures that flanked the new site to north and south, the New Old South Church by Cummings & Sears of the mid seventies, still standing across Boylston Street, and the contemporaneous Museum of Fine Arts by John H. Sturgis (?-1888) and Charles Brigham (?-1925) which long occupied the south side of the square.

Trinity is dark and rich in colour, a complex pile rising massively to its large central lantern. Moreover, it was flanked at the left on the Boylston Street side, where Richardson took Picturesque advantage of the corner cut off his site by Huntington Avenue, with an asymmetrically organized and domestically scaled parish house. The Library is light coloured and monochromatic, all of a smooth-cut Milford granite ashlar originally almost white and even today much lighter than the rock-faced pink Milford granite of Trinity. It is, moreover, a simple quadrangular mass, capped by a pantiled[291] hip-roof of moderate height; the scale throughout is monumental and the detail sparse but eminently suave. Yet if the contrast with Richardson’s Trinity of 1873-7 is so great—and even greater with the ponderous vernacular Richardsonian as that was long illustrated south of the Library in the all-brownstone S. S. Pierce Store just built by S. Edwin Tobey in 1887—the continuity with Richardson’s work of the mid eighties is equally notable.

For example, none of Richardson’s own late work was polychromatic. Three of his more prominent edifices, the Allegheny County Buildings in Pittsburgh and the Glessner and MacVeagh houses in Chicago, were all of light-coloured granite, while the Warder house in Washington is of smooth-cut limestone such as Wells had wished to use for the Villard houses. Above all, the quadrangular block of the Boston Library with its regular arcuated fenestration parallels rather closely the design of Richardson’s just completed masterpiece, the Marshall Field Store. Thus, in fact, Richardson’s former assistants, for all the Renaissance precedent of their detailing—and the courtyard of tawny Roman brick is almost more Bramantesque in treatment than the Villard houses—were to a very notable extent only proceeding farther in a direction that he himself had already taken.

Since most contemporaries, in their innocence, thought the Richardsonian merely a Romanesque Revival, it is understandable that they saw in such things as the Villard houses and the Boston Public Library an alternative—and anti-Richardsonian—Renaissance Revival. Nor can it be denied that the handling of the exterior of the Library derives from the sides of Alberti’s Tempio Malatestiano in Rimini almost as directly as the arcade in the court is copied from that of the Cancelleria Palace in Rome.[292]

The stair-hall, the reading-room, and even the minor corridors reveal clearly their Letarouillian origins when they are studied in the architects’ drawings, drawings which imitate the very style of draughtsmanship of Letarouilly’s plates. The stair-hall, executed in yellow Siena marble, has walls decorated allegorically by the French painter Puvis de Chavannes, generally considered the greatest muralist of the age; the delivery room has an entirely different sort of illustrative Shakespearean frieze painted by Edwin A. Abbey; the hall in the top storey contains John Singer Sargent’s most ambitious murals. The associated sculpture by Augustus St Gaudens and others is less interesting; but these notable decorative increments from the hands of painters and sculptors of considerable reputation help to explain why for a generation this building was thought to have initiated a real ‘American Renaissance’ in which all the arts participated. Of this ‘Renaissance’ an international exhibition represented the moment of early triumph.

When, in 1891, it was decided to hold in Chicago the first American international exhibition in recognition of the 400th anniversary of the discovery of America by Columbus, the initial architectural responsibility lay with the Chicago firm of Burnham & Root. They were working at that very moment on two of the most remarkable of early Chicago skyscrapers, the Reliance Building (Plate 115B) begun in 1890, which eventually offered the frankest expression of the new all-skeleton construction, and the Monadnock Building begun the next year, which was the last very tall building to have exterior walls of bearing masonry (see Chapter 14). The more representational Chicago skyscrapers of this period by Burnham & Root, the Women’s Temple and the Masonic Building, were of generically Richardsonian character; and Richardsonian influence was never stronger and more general in Chicago than in the five years following his death. But the principal buildings of the World’s Columbian Exposition,[293] as they rose in 1892-3, proved to be neither Richardsonian nor at all expressive of metal construction in the way of those at the Paris Exhibitions of 1878 and 1889 (see Chapter 16).

Burnham in 1891 called in various leading Eastern architects to assist him in designing the World’s Fair, as the Chicago exhibition was usually called. Then in that same year his partner Root, the designer of the pair, died. So it came about that the Easterners, not so much the ageing Hunt, dean of the profession, as the energetic and executive McKim, called the tune; McKim even provided Burnham with a new designer in the person of Charles B. Atwood (1849-95) to replace Root. The Fair, with the landscape architect Olmsted to collaborate on the planning, came out a great ‘White City’, the most complete new urbanistic concept[294] to be realized since the replanning of Paris and of Vienna in the third quarter of the century (Figure 20).

The metal-and-glass construction of the regular ranges of vast exhibition buildings was almost entirely hidden by the elaborately columniated façades of white plaster that were reflected, dream-like, in Olmsted’s formal lagoons. The architects’ inspiration was generically academic, not specifically Italianate or Classical, and only one or two small State pavilions followed Colonial Revival models. The dominant scale was very large indeed, and the façades of the various buildings, although by many different architects both Eastern and Western, were surprisingly harmonious. The young men back from the École in Paris must have worked overtime to bring up to McKim’s increasingly academic standards the projects of various well-established architects who had been doing more or less Richardsonian work for the last decade.

[Illustration:

Figure 20. D. H. Burnham and F. L. Olmsted: Chicago, World’s Fair, 1893, plan ]

Despite the major importance of the Shavian influence in America around 1880, after the designing of the Villard houses in 1883 American architects moved far more rapidly than Shaw himself along the path towards abstract order and stylistic discipline. The H. A. C. Taylor house introduced, in an American version, the formal eighteenth-century revival—whether one calls it ‘Monumental Queen Anne’ or ‘Neo-Georgian’—before Shaw began his house for Fred White. It is even perhaps significant that this was done for an American client. The World’s Fair of the early nineties brought to the fore a more Classical and ordered sort of Neo-Academicism than Shaw ever reached. By the standards of the next generation, for example, Atwood’s Fine Arts Building at Chicago (Plate 109A), though based on a Prix de Rome project of 1857, was more advanced than Shaw’s Piccadilly Hotel of 1905-8 (Plate 107). The Paris Exhibition of 1889 was notable for its great feats of metal construction, Eiffel’s Tower (Plate 130A) and Contamin’s Galerie des Machines (see Chapter 16). But the façades of the Grand Palais built for the Paris Exhibition of 1900, executed permanently in stone, seem merely a solider realization of the plaster ‘dream-city’ that Burnham and McKim had conjured up on the Chicago lake-front earlier in the decade.

Whether or not there was really influence from Chicago on Paris in the late nineties, there can be no question that the influence of the Fair in America was very great indeed. While the buildings of the Fair were rising in 1892 the young Frank Lloyd Wright built his Blossom house in Chicago in rather obvious emulation of McKim, Mead & White’s Taylor house (see Chapter 15). The following year he submitted in competition a completely academic project for a Museum and Library in Milwaukee. Moreover, this project, based on Perrault’s east front of the Louvre, was more suave in its academicism than the buildings that Richardson’s successors, Shepley, Rutan & Coolidge, who had already gone over like almost everyone else to the McKim camp, were erecting that year for the Chicago Public Library and for the Chicago Art Institute on Michigan Avenue.

It is the great historical paradox of this period in Chicago that at the very time the academic triumph of the Fair was being prepared, nineteenth-century commercial architecture was also reaching its climax there. Even before Richardson died, his tradition had split in the mid eighties. One side of it, that related to his own French training and his dependence on various styles of the past, limited though that was, as also his growing concern with architectonic order, went forward under the leadership of McKim (see Chapter 24). The other side, derived from his sense of materials, at once intelligent and intuitive, and his interest in functional expression—the qualities that were most notable in his shingled houses and his commercial buildings—provided the platform from which first Sullivan and then Wright in the late eighties and the nineties advanced to the creation of the first modern architecture (see Chapters 14 and 15).

If the importance of Richardson and, indeed, that of Shaw—as regards the development of domestic architecture—are to be fully appreciated the stories of the general development of the commercial building and of the dwelling-house in England and America down to 1900 must be known. Of the two, that of commercial architecture is the simpler and also the more dramatic. The culmination of this story in the American skyscrapers of the nineties has been recognized, from the time when so many foreign visitors came to Chicago in 1893 on account of the Fair, as one of the major and most characteristic architectural achievements of the whole period with which this volume deals.

------------------------------------------------------------------------

## CHAPTER 14

THE RISE OF COMMERCIAL ARCHITECTURE IN ENGLAND AND AMERICA

THE line of technical development which runs from the cast-iron-framed textile mills of the 1790s in England to the steel-framed skyscrapers of the 1890s in America seems to posterity a simple and obvious one. But, in fact, various lags and cul-de-sacs make the story long and complex. The most significant technical advances in iron construction of the first half of the century were not in the commercial field, and the account in this chapter is by no means merely a repetition and a continuation of the story of iron construction down to 1855 that has been provided earlier (see Chapter 7).

The great difference between the Benyons, Marshall & Bage mill of 1796 at Shrewsbury, which initiated metal-skeleton structure, and Sullivan’s Guaranty Building in Buffalo, N.Y., of a century later is that the English mill is purely and simply a technical feat of construction quite without architectural pretension. If not literally anonymous, the mill was certainly the work of a millwright rather than an architect; the skyscraper, on the other hand, is a prime architectural monument of the long period of a century and a half that this book covers, and the masterpiece of one of the greatest and most creatively original designers that the nineteenth and twentieth centuries have produced (Plate 119). But the skyscrapers of the 1890s do represent also the culmination of developments in the field of construction that began with the English mills of the 1790s, even if those developments are far from being the whole story of nineteenth-century commercial architecture. How office buildings were gradually received into the realm of architecture and, by the end of the nineteenth century, had risen so high in that realm that few productions of the 1890s in other fields of building can compare in quality of design with the great early skyscrapers is perhaps more significant for western culture in general than the purely technical aspect of the story. The weaving together of these two strands makes the full story one of the most interesting and complex in the history of nineteenth-century architecture.

Nineteenth-century commercial building need not be very precisely defined. It includes several slightly different sorts of edifices suitable for the needs of business, all consisting of a succession of identical upper storeys subdivided into offices or storerooms, with or without shops or representational premises below. Highly specialized and very lucrative concerns such as banks and insurance companies, to whom prestige of various sorts increasingly appeared a major desideratum, were the first to seek dignity and architectural display by employing architects of established reputation. Such agencies also desired buildings that were fire-resistant quite as much as did contemporary mill-owners. Already in Soane’s earliest work at the Bank of England he emulated, as has been noted, certain French technical advances that had just been employed by Louis in the Théâtre Français in Paris before these advances were first adopted in an English textile mill (see Chapters 1 and 7). Along Regent Street, around 1820, Nash and others housed less pretentious types of business in structures of mixed character and of less completely fireproof construction. But the premises on the ground floor here generally required very wide shop-windows of the sort that the use of iron supports made possible, even though the upper storeys were still nearly identical with those of domestic terraces.

In Boston in the mid twenties Parris was designing for the streets flanking his Market Hall commercial façades of a much more novel character, using not iron but granite in monolithic posts and lintels to provide a masonry skeleton filled with wide and close-set windows in all the storeys (Plate 112B).[295] In later Boston work of the next two decades in this tradition architects such as Isaiah Rogers and various builders employed iron for internal supports and sometimes also on the exterior at ground-floor level. But the granite ‘skeleton’ front preceded the skeletonized all cast-iron front in America by precisely a quarter of a century.

In England in the forties complete internal skeletons of iron carrying jack arches of brick or tile, hitherto used chiefly in textile mills, were increasingly adopted for superior commercial work, but the characteristic exteriors of commercial buildings[296] remained entirely of bearing masonry construction. However, in one case at least, a small block at 50 Watling Street in London which was probably built before 1844, the iron came through to the outer surface in the continuous window-bands of the upper storeys, even though the corner piers and the sections of wall between the storeys were of solid brickwork.

From C. R. Cockerell, titular Architect of the Bank of England after Soane’s retirement in 1833, and other architects such as Hopper, banks and insurance companies in London and other large cities obtained in the thirties and forties distinguished buildings all of masonry. In one especially fine edifice, erected in 1849-50 purely for use as offices, Bank Chambers behind Cockerell’s monumental Branch Bank of England of 1845-8, in Cook Street in Liverpool, he closely approached the directness of trabeated masonry expression of the contemporary Boston architects and builders (Plate 112A). The fireproof construction was of vaulted masonry throughout, moreover, with iron used only for the skylights over the stair-wells.

For the general character of commercial architecture down to the late fifties, however, A. & G. Williams’s Brunswick Buildings of 1841-2, also in Liverpool, were more significant. In this very large quadrangular block of general offices they followed the _palazzo_ model provided by Barry’s newly completed Reform Club almost as closely as George Alexander had already done in his Bath Savings Bank the year before. The _palazzo_ mode soon became the favourite one for imposing commercial architecture in Britain and, before long, in the United States as well.[297] With its regular rows of good-sized windows and its special prestige of having housed a commercial aristocracy in Renaissance times, this had certain aspects of suitability, both real and symbolical, to the needs of business-men. It also had serious disadvantages which soon led to a gradual modulation away from the earlier formulas of design.

The wide spacing of the windows demanded by correct _palazzo_ precedent was awkward for offices requiring that maximum of natural light which was so readily provided by Parris and others in their granite buildings in Boston and by the unknown designer of 50 Watling Street in London. Therefore windows were soon much enlarged and also set closer together. Sometimes, moreover, as in a large cotton warehouse built in Parker Street in Manchester in 1850 by J. E. Gregan (1813-55), the increasingly heavy frames were applied only to every other opening. Properly, such ‘palaces’ ought not to be more than three storeys high, but the rapidly rising value of good sites in urban business districts made it ever more desirable to carry office buildings to four and five storeys like the terrace houses of the period.

Already in the Sun Assurance Offices in Threadneedle Street in the City of London, designed in 1839 and built in 1841-2, which do not in fact conform at all closely to the standard _palazzo_ formula, Cockerell not only opened the ground floor with an arcade of haunched-segmental arches but also linked his two topmost floors behind an engaged colonnade in order to reduce the apparent height of the façade to three storeys. Across the street in the Royal Exchange Buildings of 1844-5 Edward l’Anson (1812-88) in 1844-5 lifted his whole palace front above a tall glazed arcade and tied the top-storey windows into a sort of frieze as Barry had already done in the second storey of the Reform Club (Plate 35B). In Manchester l’Anson’s cousin Edward Walters (1808-72) in the Silas Schwabe Building of 1845 at 41 Mosley Street linked the windows of the first and second storeys by an applied arcade.

The building with an exterior entirely of cast iron that James Bogardus (1800-74) designed and built for his own use in New York in 1848-50 was well publicized at the time,[298] and is still famous although long since demolished. On the corner of Washington and Murray Streets in New York another Bogardus building, the Laing stores erected in two months in 1849, is still extant (Plate 67B). Although there was never any such general use of cast-iron fronts in Great Britain as in America in the fifties and sixties, it seems probable from contemporary evidence that some architect, probably Owen Jones, built one at 76 Oxford Street in London a year or so before 1851. However that may be, an ironfounder named McConnel provided the structural elements for an office building that still stands[299] in Jamaica Street, Glasgow, in 1855 with an exterior all of cast iron. A curious feature of the design of this structure is the delicate iron membering that forms a series of arcades between the major structural piers. This decorative device, structurally meaningless in iron except for bracing although employed by Paxton at the Crystal Palace, is probably an imitation of the masonry arcading that was, in the mid fifties, gradually modifying the earlier _palazzo_ paradigm quite beyond recognition.

In 1849 Wild used two ranges of Italian Gothic arcades on his St Martin’s Northern Schools in London, and the perspicacious Street remarked in an article on the obvious suitability of the theme for commercial fronts, as has already been noted. In Manchester in 1851 Starkey & Cuffley in a pair of shops employed ranges of three arches on each of the two fronts in the four storeys, binding them in with coupled columns marking the ends of the party walls.

The lifting of the window tax in 1851 encouraged great increases in window area. In jubilant recognition of this H. R. Abraham the next year made all his windows triplets in the first and second storeys of the W. H. Smith Building at 188-192 Strand in London, but without using any arches at all. Two years later, however, in a building for Heal’s furniture store in Tottenham Court Road in London, James M. Lockyer (1824-65) carried a _quattrocento_ arcade all across the first storey.

By this time architects and public alike had become aware of a different High Renaissance formula from Barry’s (see Chapter 4). Beside the Reform Club in Pall Mall Sydney Smirke’s new front of the Carlton Club, designed in 1847, was coming to belated completion in the mid fifties. Moreover, its Sansovinesque arcades were already echoed in the first storey of Parnell & Smith’s Army and Navy Club of 1848-51 across the way. These London models were closely followed by William B. Gingell (1819-1900) in his West of England Bank in Corn Street, Bristol, of 1854 and quite outranked by the great Venetian _palazzo_ that David Rhind (?-1883) erected in 1855 in Prince’s Street in Edinburgh for the Life Association of Scotland.

Possibly the fine warehouse at 12 Temple Street in Bristol with three groups of triplet arches in each of the upper storeys is by Gingell and of this date. There is none of the Sansovinesque lushness of his bank here, but the fine workmanship of the quarry-faced Pennant stone walls laid up in random ashlar, with smooth-cut Bath stone trim and coloured voussoirs banding the arches, bears some resemblance to the Bristol General Hospital he was building in 1853-7, notably in the very bold rustication of the ground-storey arches.

However that may be, two London buildings of 1855 advanced nearly as far towards the all-arcaded front. Hodgson’s Building by Knowles in the Strand at the corner of Chancery Lane had the general character of a _palazzo_, but all the windows were arched, as in buildings of the _Rundbogenstil_; moreover their trim sank into the wall rather than projecting from it, so that the wall sections between were reduced visually to mere piers, even though they had no imposts. The Crown Life Office, in New Bridge Street, Blackfriars, was built in 1855-7 by Ruskin’s friends Deane & Woodward, with whom he was most closely associated precisely in those years. The round-arched medieval arcading of this façade, with the piers hardly narrower than on Knowles’s building yet articulated by bases and imposts, may surely claim Ruskinian sanction. Here, at any rate, was the first important contact between advanced High Victorian Gothic and the commercial world, a contact destined to be very fruitful over the next fifteen years or so. Henceforth even architects of no aesthetic pretension were ready to exploit arcading.

The English development of arcaded masonry façades can be closely matched in America, specifically in Philadelphia.[300] There S. D. Button (1803-97), Napoleon Le Brun (1821-1901), and others in buildings of 1852-3 in Chestnut Street—that at 239-241 by Button is still extant—consistently used arched openings between slim piers; and Notman in 1855 provided for the Jackson Building at 418 Arch Street a façade even more completely articulated by arcading in all its four floors than the Crown Life Office. By this time, moreover, the trabeated design of Bogardus’s first iron fronts had likewise given way to ornate arcading in emulation of masonry fronts.[301]

Iron remained behind the scenes in most of the English arcaded buildings. In Waterhouse’s Fryer & Binyon Warehouse in Manchester of 1856, however, whose upper walls had the polychrome diapering of the Doge’s Palace so much admired by Ruskin, the first storey was opened up by an arcade carried on coupled iron columns. In the Wellington Williams Warehouse of 1858 in Little Britain in London, the obscure City firm of J. Young & Son used arcades in all the five storeys with iron columns to support the outer orders; thus the width of the piers could be considerably reduced, and the effect of over-all articulation was much enhanced as in the Philadelphia buildings.

Deane & Woodward’s very Ruskinian project of 1857 for the new Government Offices, with its endless Italian Gothic arcading, and a small warehouse in Merchant Street in Bristol of 1858 by Godwin gave some impetus to the use of pointed instead of round arches. But on the whole the best designed among the innumerable arcaded façades in England retained the rounded form, however Gothic their other detailing may be. In one of the largest and finest examples of the early sixties, moreover, Kassapian’s Warehouse in Leeds Road, Bradford, perhaps by Lockwood & Mawson, the detailing is academically Roman (Plate 114B).

Different as they are, this Bradford façade and that of Godwin’s contemporary warehouse at 104 Stokes Croft in Bristol, so much more subtly Ruskinian than anything by Deane & Woodward, are the two masterpieces of the genre at its best moment (Plate #113:pl113). Of very high quality also is 60 Mark Lane in the City of London built by George Aitchison in 1864-5. There the existence of a complete iron skeleton, presumably but not certainly present in most of the other examples, is fully documented. Moreover, on the rear the metal comes through to the outer face of the wall much as it did at 50 Watling Street, built some twenty years earlier.

In Philadelphia William Johnston had begun in 1849 the seven-storey Jayne Building in Chestnut Street,[302] introducing a new vertical formula of design for commercial façades. Above a conventional ground floor, narrow granite piers in the forms of clustered colonnettes rise the full height of the building, merging into Venetian Gothic tracery below a terminal parapet. Whether or not Samuel K. Hoxie, the contractor who provided the Quincy granite for this and other Philadelphia buildings, was familiar with the ‘granite-skeleton’ work of Parris, Rogers, and others in Boston is not clear. But in the next few years a good many façades with a similarly vertical and ‘skeletonized’ treatment were built in Philadelphia by J. C. Hoxie and his sometime partner Button. That across the street from the Jayne Building has already been mentioned, since the openings between the piers are covered with segmental arches throughout. Button’s building at 723-727 Chestnut Street of 1853 and his extant Leland Building at 37-39 South Third Street are even more ‘proto-Sullivanian’, so to put it. Louis Sullivan probably saw and admired such things as the Jayne Building and the Leland Building when he was working for Frank Furness in Philadelphia in the seventies; certainly they are very premonitory of his characteristic work of the eighties and even the nineties.

Various other ways of reducing the wall to little more than a masonry cladding of the iron structural members were also in use in England as well as in America by this time. A notable small edifice in the City of London, of uncertain date and authorship but probably by Thomas Hague and of 1855, is at 22 Finch Lane, with another front to the court at the side. On both these façades the two lower storeys are joined together visually by setting back the horizontal spandrel between them, and the moulded stonework of the very narrow piers is of almost metallic scale and crispness.

Still more striking is Oriel Chambers[303] in Water Street in Liverpool, built in 1864-5 by Peter Ellis (fl. 1835-84), and another smaller building by him at 16 Cook Street of a year or two later. On the front façades of these the masonry is scaled down quite as much as at 22 Finch Lane but given a more decorative treatment, in both cases of rather metallic character. At Oriel Chambers, oriels of plate glass held in delicate metal frames are cantilevered out in every bay of all the upper storeys, producing a regular rhythm broken only by the clumsy cresting on the top (Plate 114A). At 16 Cook Street all the stone spandrels are set back, thus emphasizing even more strongly than at Oriel Chambers the continuous vertical lines of the mullions. The over-all pattern is once more somewhat confused, however, by the arches across the top that link the mullions together. The rear walls of both of Ellis’s buildings are even more open in design and directly expressive of the metal skeleton. Towards the narrow court at the side of Oriel Chambers only every third iron pier is clad with masonry; those between rise free behind the glass of the horizontally sashed windows whose upper planes are slanted inward. This is, in effect, an early example of the ‘curtain-wall’ (see

## Chapter 22).

If in some technical respects the Chicago skyscraper of the nineties seems almost to have come to premature birth in Liverpool in the sixties, as in some other respects it had done in the Philadelphia commercial buildings of the fifties, the immediate influence of these buildings by Ellis seems to have been almost nil. Eventually Owen Jones, in a façade at Derby of 1872, and Thomas Ambler, in a corner building at 46-47 Boar Lane in Leeds of 1873, did come to use only iron and glass, omitting all masonry; but more characteristic commercial work of these years is to be seen in such warehouses by unknown hands as the one at 1-2 York Place in Leeds, with an arcade crisply detailed in moulded brick rising through all the upper storeys, somewhat as on the Philadelphia buildings of the fifties, or a larger example in Strait Street in Bristol, with a much heavier arcade subsuming several upper storeys, handsomely executed in stones of different colours and textures and very boldly and simply detailed. Such things, however, very soon seemed to the English not advanced but retardataire as contemporary attention focused on the Queen Anne of Shaw’s New Zealand Chambers of 1872-3.

Richardson’s very un-Shavian American Express Building[304] in Chicago of 1872-3 first brings that Mid-Western metropolis into this story. That had no arcading, but the windows were very closely set, sometimes (it would appear) with only light metal colonnettes as mullions between them. There was also a directness and a ‘realism’ of treatment throughout comparable to that of Richardson’s more monumental work of this date, notably the Hampden County Courthouse and the Buffalo State Hospital, both designed the previous year and at this time still in construction. But Richardson’s dependence on English commercial work of the preceding fifteen years became closer still in his first really fine business building, the Cheney Block (now the Brown-Thompson Department Store) built in Hartford, Conn., in 1875-6 (Plate 116A). Here the wide ground-storey arcade, including a mezzanine, and the narrower arcade above, subsuming several storeys—as on the very proto-Richardsonian warehouse in Strait Street in Bristol—are carried out with typically Richardsonian stoniness in quarry-faced brownstone. But the banded arches introduce a bold note of High Victorian Gothic polychromy, and the carved detail is in the harsh but richly naturalistic vein—also High Victorian Gothic in spirit—of the ornament on the earliest executed portions of Trinity Church in Boston, probably of a year or two before.

Already, in New York, the skyscraper[305] had been born by this date, and leadership in commercial architecture had crossed the Atlantic for good and all. None of the structures dealt with so far in this chapter except the Jayne Building were more than five or six storeys high, since it could not be expected that business clients would climb more than four or five flights of stairs. But the average height of buildings in the financial districts of cities had, even so, almost doubled since the eighteenth century, partly because of the general rise in the number of storeys, partly because of much increased storey heights. Vertical transportation of human beings, which would allow the erection of office buildings considerably more than five storeys high—industrial buildings were often much taller already—became increasingly feasible during the forties and fifties. Hoists for goods were a commonplace of English warehouse design after 1840, and in 1844 the Bunker Hill Monument had a passenger-hoist operated by a steam engine. In New York the Haughwout Store on Broadway had in 1857 the first practical passenger lift or elevator to be installed in an ordinary urban structure. This was of the type developed by Elisha G. Otis. A lift of another sort was introduced in the Fifth Avenue Hotel in New York later that year. Those of 1860 in the Westminster Palace Hotel in London apparently did not function, at least for some years. The Equitable Building, for which Arthur Gilman and Edward Kimball, with George B. Post (1837-1913) as the associated engineer, won the competition in 1868, was the first office building in New York to have a lift from the time of its completion in 1871. Immediately after this lifts were introduced in several other comparable structures, and one- or two-storey mansards were often added to the tops of existing buildings. A great change was thus at hand in New York in the early seventies.

Despite the Panic of 1873, the mid seventies saw the construction of what may properly be considered the first skyscrapers, the nine-storey (260-foot) Tribune Building and the ten-storey (230-foot) Western Union Building. Both were therefore about double the height even of the tallest office structures, such as the five-storey (130-foot) Equitable Building erected during the preceding boom period. These first skyscrapers rose to altitudes reached hitherto in America only by church spires, as general views of the New York skyline around 1875 make evident. Neither Hunt’s New York Tribune Building, extant but since carried many storeys higher, nor Post’s Western Union Telegraph Building, long since demolished, incorporated any other technical innovations;[306] nor was their design at all closely related, like that of Richardson’s Cheney Block in Hartford, to the advanced English commercial work of the previous decade. Paradoxically, the French-trained Hunt’s building is somewhat the more English of the two in character; but, for all the direct expressiveness of the window grouping in triplets in each bay, the detail throughout is coarse and gawky, and the silhouette of the very tall mansard and the asymmetrically placed tower was from the first overbearing. The later addition of many more storeys has made the building even more top-heavy in appearance. The Tribune Building was of interest chiefly for its relatively great height, now unnoticeable among the much taller skyscrapers built around it later. Its almost complete avoidance of any sort of archaeological styling, however, such as the Romanesquoid of Richardson’s Cheney Block or the violently polychromatic and spiky Gothic of Hunt’s own Divinity School at Yale, on which construction was still at this date proceeding, is certainly worth remark also.

The Western Union Building of Post was only nominally French, for its rather heavy-handed Second Empire treatment owed more to earlier English and American designs in this mode than to anything Parisian (Plate 115A). But the exterior was more orderly, if less expressive, than that of Hunt’s skyscraper and the mansards on top piled up as grandly to the centrally placed tower as on the big contemporary Post Office near by. Yet stylistically both Post’s and Hunt’s buildings were out of date almost as soon as they were finished; and after the hiatus caused by the depression of the seventies the locus of the skyscraper story moved westward to Chicago.

Chicago, already the metropolis of the Middle West, had almost no architectural traditions at this time. First developed as a city in the thirties, the need for rapid building in timber had led to the invention or development of what is called ‘balloon-frame’ construction, in which relatively light studs or scantlings, rising wall high, form a cage or crate whose members are fastened together by a liberal use of machine-made nails. Balloon-frame construction, thus, is a typical offshoot of the industrial revolution, becoming feasible only with the mechanization of the saw-mill and of the manufacture of nails. Theoretically, there might be thought to be some analogy between this New World method of carpentry, so different from the heavy framing of the Old World, hitherto always used in America as well, and metal construction. There is no evidence, however, that Chicago took to iron with any greater enthusiasm in the fifties and sixties than did New York or various other cities; indeed, St Louis seems to have had more and finer examples of cast-iron fronts, particularly in the early seventies. As late as that, moreover, the new cities of the American Northwest were obtaining cast-iron fronts prefabricated from Britain, just as San Francisco had obtained many of her warehouses and immigrant dwellings in 1849-50.

At the opening of the seventies a terrific conflagration[307] all but wiped out Chicago. The need for rapid rebuilding drew thither ambitious architects and engineers from all over the East, but the immediate results of their activities were anything but edifying. Architectural leadership was still centred in Boston and New York; in any case, that leadership had rarely been more confused than in the early seventies when even Richardson was only just maturing his personal style. Richardson’s own Chicago building for the American Express Company was doubtless too indeterminate in character to attract a local following; nor did he build again in Chicago until the mid eighties, by which time various versions of the Richardsonian were already reaching Chicago at second or third hand.

If the Chicago architectural scene had any virtues around 1880 they were largely negative ones: no established traditions, no real professional leaders, and ignorance of all architectural styles past or present. Among the architects who had settled in Chicago in the seventies was a Dane, Dankmar Adler (1844-1900). Into his office in 1879, first as chief draughtsman but soon as partner, came the young Bostonian Louis Sullivan. As has been noted before, Sullivan had been trained first in Ware’s school at the Massachusetts Institute of Technology and later, until he revolted against its rigid doctrines, at the École des Beaux-Arts in Paris. Having worked for Frank Furness, wildest of American High Victorians, Sullivan picked Chicago not alone for its evident professional opportunities but also because he liked the idea of working where there were no hampering traditions. (Moreover, his parents had moved there from Boston.)

The earliest building of any real originality designed by Sullivan, the Rothschild Store in Chicago of 1880-1, seems at first a turgid compilation of barbarisms. Examined more closely, however, and compared with the Leiter Building on its right, which was built two years earlier by the engineer-architect William Le Baron Jenney (1832-1907), the two sorts of innovation that Sullivan essayed here can be readily recognized. On the one hand there is the ornament,[308] undefinable in historic terms yet with a kind of similarity—almost certainly accidental—to the Anglo-Japanese detail of Nesfield and Godwin. At this stage in Sullivan’s career the originality of his ornament must be remarked but can hardly be admired. Below his elaborate ornamental cresting, on the other hand, Sullivan handled the main architectonic elements of his façade with considerable novelty and most admirable logic. Although the building is not tall—no skyscraper, that is, even by the modest standards of 1880—Sullivan did not hesitate to follow the lead of the Philadelphia commercial architects of the fifties in emphasizing the vertical. This he accomplished by continuing the mullions that subdivide his bays across the spandrels, somewhat as Ellis had done fifteen years before in his buildings in Liverpool, rather than by using a multiplicity of masonry piers.

Sullivan’s next Chicago building, the Revell Store erected for Martin Ryerson in 1881-3, continued the theme of the Rothschild Store, but extended it over a much larger corner block with considerable chastening of the ornamental treatment at the top. The Troescher Building of 1884, which came next in sequence, is very much finer. Widely-spaced piers of plain brickwork rise the full height of the façade above a slightly Richardsonian ground-storey arcade of rock-faced stone; between them there are no oriels, as on Ellis’s Oriel Chambers or his Ryerson Building[309] of the previous year, but broad horizontal windows separated by recessed spandrels. These spandrels are rather like Ellis’s on his other building at 16 Cook Street, but their actual prototypes are to be found, more probably, in Philadelphia buildings by Button such as the one at 723-727 Chestnut Street. The ornament here, now still further chastened, is largely confined to these spandrels. The curved cresting across the top, however, recalls a little the turgid crown of the Rothschild façade.

Sullivan’s early buildings were not very tall, and they did not advance the technical development of the skyscraper. In these same years, however, other Chicago architects were doing so to notable effect. For the ten-storey Montauk Block of 1882-3, tall, but no taller than the first New York skyscrapers of ten years before, Burnham & Root introduced spread foundations to carry its great weight on the muddy Chicago soil, out of which earlier buildings had, literally, to be hoisted every few years. In design they were content, however, with a range of ten almost identical storeys of plain brick pierced by regularly spaced segmental-arched windows. Obvious as this treatment may seem, it took courage to use it at a time when most architects were still trying to disguise the embarrassing height of buildings only half as tall by grouping their storeys together in twos and threes.

The Home Life Insurance Building begun in 1883 was also only ten storeys tall.[310] But in building it Jenney invented, or at least introduced in Chicago, what is specifically called ‘skyscraper construction’, that is a method of carrying the external masonry cladding on metal shelves bolted to the internal skeleton. Jenney, however, probably thought he was merely tying together his metal skeleton and his brickwork, not carrying the latter entirely, though this was found to be the case when the structure of the building was carefully examined during its demolition. The Home Insurance Building, in any case, looked far more as if its external walls were bearing than do any of Sullivan’s early works. Jenney, moreover, fought shy of the frankness of Burnham & Root’s treatment of the Montauk Block; instead he phrased his storeys in groups, almost as if several buildings of normal three- or four-storey height had been casually piled one on top of the other.

Before the Home Insurance was finished in 1885 two more major commercial monuments were rising in Chicago, Richardson’s Marshall Field Wholesale Store (Plate 116B), last but one of the large buildings erected in Chicago with walls entirely of bearing masonry, and Burnham & Root’s Rookery Building (see Chapter 13). Both were begun in 1885, Richardson’s being finished in 1887 and Burnham & Root’s a year earlier in 1886. The exterior of the eleven-storey Rookery Building is not an example of the stripped ‘functionalism’ that these architects had introduced in their Montauk Block but rather a provincial imitation of the Richardsonian. In the court walls, however, the architects used—and with complete awareness of its implications—the new structural method of Jenney’s Home Insurance Building, carrying the brickwork above the sides of the central glass-roofed lobby entirely on the internal metal[311] skeleton.

With the advent of Richardson in 1885, the main lines of development in commercial architecture, both as regards design and as regards construction, might seem to have been concentrated in Chicago. It is well therefore to note again that McKim, Mead & White in their Goelet Building on Broadway in New York of 1885-6 provided almost as frank an expression of the skyscraper, or tall office building of many identical storeys, at least above their Renaissance ground-floor arcade, as did Burnham & Root in the Montauk Block. Their windows, however, were phrased in triplets like Hunt’s on the Tribune Building and also grouped vertically within tall bay-width panels of moulded brick rising with only one break to the cornice. This was a quite frank solution of the problem, and is hardly to be castigated as ‘traditional’ or even as ‘un-functional’. Moreover, another New York building, Babb, Cook & Willard’s De Vinne Press of 1885 in Lafayette Street, is not altogether unworthy of comparison with the Field store. It lacks the regularity and the grandeur of scale of Richardson’s masterpiece, but George F. Babb used his fine red brick in a belated _Rundbogenstil_ way, and not without some conscious reminiscence, one may presume, of Durand’s exemplars of the beginning of the century.

Richardson’s last commercial work, the Ames Building in Harrison Avenue in Boston of 1886-7, on which the arcade was carried the full height of the building and the reveals much reduced, had no immediate influence in Chicago (see Chapter 13). Sullivan’s first really great work, the Auditorium Building (now Roosevelt College) in Chicago, derived for the most part straight from the Field store, at least as regards the exterior. Designed in 1886 and built in 1887-9, this is a vast and complex edifice, or group of edifices, with a hotel on the Michigan Avenue front, an opera-house entered in the middle of the Congress Street side, and offices along Wabash Avenue at the rear. The walls are all of bearing masonry still. In order to incorporate more storeys than Richardson had ever done, Sullivan carried up his heavy rock-faced granite base through two mezzanine levels and increased the number of floors subsumed by the main arcade which rises from the first storey (Plate 117A). He also used light stone throughout, instead of the red granite and the brownstone of the Field store, with its surfaces all smooth-cut above the mezzanines.

This flattening of the wall-plane was carried even further on the tower which rises above the portal of the opera-house in Congress Street. On that wide arched panels of very slight projection are filled with articulated screens of stone in which the windows are arranged in a continuous grid with no evident storey lines. The eaves gallery at the top of the tower, a stubby colonnade set in a long horizontal panel with a continuous ribbon-window behind—the window in fact of the Adler & Sullivan office—is so like Thomson’s on the front of his Queen’s Park church of the sixties in Glasgow that it is hard to believe Sullivan did not know it. Yet other evidence indicates that he continued to abjure all European influence at this point in his career.

In the interiors, particularly the bar and the banquet hall at the top of the hotel, Sullivan’s ornament changed even more markedly than his exterior design. Here also there is possibly Richardsonian influence, but coming from the Byzantinizing detail worked out by John Galen Howard of the Richardson office for the MacVeagh house of 1885-7 in Chicago rather than from the Field store.

However, one cannot entirely discount the possibility of a contribution in the field of ornament by a brilliant young man of twenty, Frank Lloyd Wright, whom Sullivan and Adler had just taken on as a draughtsman in 1887 and who was soon given charge of the innumerable detail drawings that this vast project required. Nurtured on Owen Jones’s _Grammar of Ornament_,[312] which the Paris-trained Sullivan claimed not to have known, as well as on the writings of Ruskin, Morris, and Viollet-le-Duc, Wright may perhaps have encouraged Sullivan to move away from the bold coarseness of his earlier ornament towards the lush elaboration of intricately plastic surface decoration henceforth characteristic of his work. It is tempting, even, to believe that Jones’s page of Celtic ornament particularly attracted the Irish Sullivan’s fancy.[313]

Together with the Auditorium, though commissioned a year later, there was also rising in Chicago in 1887-9 the Tacoma Building of William Holabird (1854-1923) and Martin Roche (1855-1927), two young architects trained in Jenney’s office. Here the exterior walls on the two fronts were entirely carried by the metal skeleton within, only the rear walls and some of the interior partitions being of bearing masonry like the walls of the Auditorium. Moreover, this fact was made evident in the frank if not particularly distinguished treatment of the two fronts. Vertical ranges of oriels were carried the full height of the building, and there was only a minimal brick and terracotta sheathing of the structural verticals and horizontals. A more or less Richardsonian cornice capped the whole, but the general effect was closer to Ellis’s Oriel Chambers of the sixties in Liverpool or to some of Sullivan’s earlier buildings than to the Field store.

Despite the general swing of Eastern architects towards the Neo-Academic in these years, some who were doing commercial work were not out of step with what was happening in Chicago. For example, there are office buildings and warehouses in Boston and New York of relatively modest height built in the late eighties and early nineties that emulate in brick the arcading of the Field store with almost as much success as Sullivan. Similar things can be seen in many Middle and Far Western cities, but these derive more probably from Sullivan or Burnham & Root than directly from Richardson.

In the Middle West, moreover, McKim, Mead & White were building in 1888-90 two very large business buildings, still with bearing masonry walls, for the New York Life Insurance Company, one in Omaha, Nebraska, and one in Kansas City, Missouri, of effectively identical design. Unlike the already characteristic Chicago ‘slabs’—the quadrangular plan of the Rookery Building is exceptional—these are U-shaped, and each has a tower rising above the main mass at the rear of the court. The treatment of the walls with tall arcading follows as evidently from the Field store as does Sullivan’s at the Auditorium; like that of the contemporary Boston Public Library, however, the fairly simple detailing is of High Renaissance rather than Richardsonian Romanesque character.

Before these towering blocks were finished in the West the new ‘skyscraper construction’ had been introduced in New York by Bradford Lee Gilbert (1853-1911). His Tower Building of 1888-9, as its name implies, was a tower, not a slab, with more or less Richardsonian detailing. It is worth noting that the Tower Building—ten storeys, 119 feet—was _not_ as tall as the first New York skyscrapers built in the early seventies with bearing walls. Indeed, Post’s World or Pulitzer Building of 1889-90 in New York with twenty-six storeys, the tallest built up to then—309 feet—still had bearing walls. Of course, the Eiffel Tower, completed in 1889, exceeded in height by a great deal all the skyscrapers of its day whatever their construction; indeed, it was not overtopped until the Empire State Building in New York rose from the designs of Shreve, Lamb & Harmon in the early 1930s at the end of the second wave of skyscraper building following the First World War.

Post’s Western Union Building of the early seventies was in the Second Empire mode; his World Building was still French, but what can better be called ‘Beaux-Arts’. It is designed like a series of three- or four-storey Renaissance palaces, one on top of the other, and crowned with a large and ornate dome. The next New York skyscrapers all followed the new structural method introduced by Gilbert in the Tower Building; but Post, Price, and the other architects who designed them used an ornate paraphernalia of Renaissance ornamentation with none of the discretion of McKim, Mead & White on their Kansas City and Omaha insurance buildings. Characteristic of the period are Price’s American Surety Building at Broadway and Wall Street, begun in 1894, and his St James Building of 1897-8 at 1133 Broadway, both in New York, and Post’s Park Building in Pittsburgh, completed in 1896. The latter’s Havemeyer Building in New York, completed earlier, in 1892, was still somewhat Richardsonian however.

The maturing of an original sort of skyscraper design around 1890 is a Middle Western, and almost specifically a Chicago, story to which New York architects made no contribution. Boston’s architectural leadership had ended with the death of Richardson; despite the prominence of McKim, Mead & White and their large Eastern following, leadership in this field passed almost at once to Chicago. It was most appropriate that Richardson’s masterpiece, the Field store, should have been built there; the inspiration it provided, as we have already seen in the case of the Auditorium Building, played an important part in the succeeding Middle Western development.

In 1889-90 Jenney built for Levi Z. Leiter a large building on South Clark Street in Chicago now occupied by Sears, Roebuck & Company. In this he not only used the new ‘skyscraper construction’ for the exterior walls but also—with the presumptive aid of his assistant and later partner William Bryce Mundie (1863-1939)—arrived at an expression of its structural character almost as logical as that of the Tacoma Building yet much more monumental. Like most other Chicago designers in these years, Jenney and Mundie were influenced here by the Field store. The uncompromisingly block-like shape of this tremendous building, with its heavy plain entablature and pilaster-like corner piers, is Richardsonian both in its scale and in its simplicity (Plate 117B). The various groupings of stone mullions that clad the main piers and subdivide the bays, lithe and light though they are, were clearly envisaged as Romanesque colonnettes and even carry modest foliate capitals. Despite the dichotomy of the solidly Richardsonian silhouette and the open screen-like treatment of the walls, the effect is coherent and dignified. In this respect the Sears, Roebuck Building is superior to Sullivan’s very Richardsonian[314] Opera House Building in Pueblo, Colorado, of 1890 which was burned in the 1920s. The Walker Warehouse in Chicago of 1888-9 better displayed his great talent.

Three buildings of the early nineties, two in Chicago by Daniel H. Burnham’s firm and one in St Louis by Sullivan, illustrate the wide range of creative possibilities in skyscraper design at this point. The most advanced is surely the Reliance Building, at least in terms of direct structural expression. This was carried up only four storeys in 1890, though extended to its present thirteen storeys by D. H. Burnham & Company in 1894. As completed, this is a refined and perfected version of Holabird & Roche’s Tacoma Building (Plate 115B). The light-coloured terracotta cladding of the vertical members, particularly on the flat oriels, is reduced to a minimum; the terminal member is a thin slab, not a cornice or an entablature; and the only stylistic reminiscence is in the cusped panelling—neither Romanesque nor Renaissance, but slightly Late Gothic in character—of the spandrels. What we see was presumably designed as well as built in 1894.[315]

Burnham & Root’s other significant skyscraper of this particular moment, the sixteen-storey Monadnock Building begun in 1891, the last tall Chicago building with bearing walls of brick, was and still remains more famous than the Reliance; doubtless it is also finer, although much mid-twentieth-century critical opinion has favoured the Sears, Roebuck Building of Jenney & Mundie and the Reliance because they are more advanced technically. The smooth shank of the Monadnock, varied only by the slight projection of the recurrent oriels, has a most subtle and elegant taper or reverse entasis. The final bending outward of the brickwork to provide a cove cornice unifies the whole formal concept with extraordinary effectiveness. Few large buildings have ever achieved such monumental force with such simple means. There is almost literally no detail of any sort, whether derivative or original.

Sullivan’s Wainwright Building of 1890-1 in St Louis, Missouri, in which he and Adler used ‘skyscraper construction’ for the first time, no longer dominates two- and three-storey neighbours as it did when newly built; thus the prominence that the relatively great height gave it in the city picture of the nineties can hardly be realized today. But Sullivan undoubtedly sought to emphasize what seemed to contemporaries, as they do not to posterity, its very tall proportions (Plate 118). Continuous pilaster-like piers of brick, quite like those on his Troescher Building of 1884, clad the vertical elements of the steel skeleton, yet identical brick piers with no major structural members behind them also serve as intervening mullions. But at the base the wide windows of the ground storey and the mezzanine reveal the true width of the actual bays of the steel skeleton as the treatment of the shank of the building does not. The piers are considerably broader than most of those on the Sears, Roebuck Building; but they are also topped, like Mundie’s, with ornament that forms a sort of capital. Moreover, the attic storey above is quite hidden behind a deep band of the richest Sullivanian ornament elsewhere restricted, as on the Troescher Building, to the recessed spandrels. The ‘cornice’ above this frieze-like attic is merely a slab, but a much thicker one than that which caps the Reliance Building. Nothing of Richardson’s direct influence is left; but by now Sullivan had learned from the Field store the basic lessons of scale and order, applying them here in a visually sure but not particularly frank way to the new type of metal-skeleton construction. The plan is U-shaped, like those of the McKim, Mead & White buildings in Kansas City and Omaha, but the court is to the rear, so that the block appears unified from the surrounding streets.

In Sullivan’s next important work, the Schiller Building in Chicago of 1891-2, he adopted—exceptionally for him—a truly tower-like shape. Here the masonry piers that clad the structural steel stanchions are not doubled by identical mullions between; instead these piers are linked by arches below a sort of frieze. The ‘frieze’ is really a very ornately arcaded eaves-gallery, not a flat band as on the Wainwright Building, occupying a whole storey below the thick slab cornice.

Interchange of ideas was continuous in these years between the various Chicago architects’ offices, while the influence of the Academic Revival in the East, dominant in almost all the buildings at the World’s Fair of 1893 save Sullivan’s own Transportation Building, was still negligible in the commercial field. Thus Sullivan’s Stock Exchange Building of 1893-4 in Chicago borrowed its rather clumsy ground storey and mezzanine, with a cavernously Richardsonian arched entrance, from Burnham’s Ashland Block of 1892 and its oriels from the Tacoma or possibly the Reliance Building. These oriels alternate with horizontal openings of the type known as ‘Chicago windows’ sharply cut through the smooth light-coloured terracotta of the wall plane. ‘Chicago windows’, with a wide fixed pane in the centre and narrower sashes that open on either side, were used by most Chicago architects in this decade and the next. A heavy moulded cornice, not just a thick slab, crowns the whole above a colonnaded eaves-gallery somewhat like the one at the top of the Auditorium tower.

What should probably be considered Sullivan’s masterpiece, the Guaranty Building in Buffalo, N.Y., followed in 1894-5 (Plate 119). One of the most significant new themes in the design of this skyscraper, whose premonitory character can only be fully appreciated in relation to the use of _pilotis_ in later modern architecture (see Chapter 22), is already to be found in a project of Sullivan’s of the previous year for the St Louis Trust & Savings Bank. This is the treatment of the ground storey, where the terracotta sheathed piers were isolated from the wall plane by bending back the tops of the shop-windows. The piers are thus nearly free-standing and seem to lift the shaft of the building above them right off the ground. This allows circumambient space to penetrate under the main volume of the building. Thus the fact that the edifice is a hollow cage is very strongly suggested, and the wide shop-windows do not appear to undermine the walls above them as in so much commercial work of the nineteenth century.

There are several reasons, not intrinsic to Sullivan’s design, that explain why the Guaranty remains the most effective of all the early skyscrapers. Since downtown Buffalo has not filled up with buildings of equal or greater height in the way of downtown St Louis and the Chicago Loop, the Guaranty still rises high above most of its modest neighbours, in effect a tower as well as a slab, although actually of U-shaped plan like the Wainwright. In this city, moreover, which has in the last sixty years remained considerably cleaner than Chicago, the colour of the tawny terracotta sheathing has not been so much obscured by grime as on the Stock Exchange Building. These were happy local conditions that Sullivan could not foresee.

The plastic handling of the crown of the Guaranty was perhaps suggested to Sullivan by the effectiveness of the cove at the top of Burnham & Root’s Monadnock Building. Here the crowns of the arched façade bays—two to each structural bay, as the wide spacing of the piers at ground-storey level so clearly reveals—are related to the outward curve of the top of the wall below the terminal slab. The profuse and melodious curvilinear ornament, subsuming the round attic windows, echoes and complements the plastic theme. This is an example, rare even in Sullivan’s most mature work of the mid and late nineties, of the successful integration of architectonic and decorative effects. The treatment of the terracotta cladding throughout the exterior of the Guaranty, moreover, covered all over as it is with lacy geometrical ornament in very low relief, seems to lighten the whole. The cladding is read as a mere protective shell carried by the underlying steel structural members and not as solid brickwork like the piers of the Wainwright Building.

Just as the Wainwright Building may be contrasted on the one hand with the still greater solidity of the Monadnock Building—in that case justified by the bearing-wall construction—and on the other with the openness of the Reliance, so it is of interest to compare the Guaranty with two other big business buildings of 1895 by other Chicago architects. In the Ellicott Square Building, also in Buffalo, Burnham was strongly influenced by his close association with McKim at the World’s Fair. With the assistance of his designer Atwood, whose short life ended this same year, he adopted the elaborate Renaissance membering and the heavy masonry vocabulary of the New York skyscraper architects, although he retained the quadrangular plan and the glass-roofed central court of the Rookery. On the other hand, in Chicago Solon S. Beman (1853-1914) in the Studebaker (now Brunswick) Building came very close to providing an all-glass front, despite the profusion of Late Gothic frippery with which he detailed his very restricted terracotta cladding.

Adler had parted from Sullivan in 1895, but Sullivan’s career as a skyscraper builder continued for a few more years at a very high level. In his next skyscraper, the Condict Building in New York of 1897-9, he reduced very considerably the width of the mullions between the piers so that they became mere colonnettes, and even these are omitted in the first storey. But this highly logical differentiation between pier and mullion, related to the treatment of his Rothschild Store of 1880-1, still gets lost at the top in a flurry of ornamentation almost as turgid in its very different and almost _quattrocento_[316] way as the top of that very early façade. The treatment of the ground storey was originally like that of the Guaranty, but has been modified by later shop-fronts.

The next year Holabird & Roche built three contiguous buildings on Michigan Avenue in Chicago for Harold McCormick (Plate 120). The two southerly ones are excellent examples of the work of the Chicago School; they are a little less extensively glazed than Beman’s Studebaker Building or Holabird & Roche’s own McClurg Building of 1899 but with crisp and simple, if quite conventional, moulded brick detail on the piers and rather plain cornices of wholly academic character. Standard Chicago windows are used throughout. The third façade on the north, that of the Gage Building at 18 South Michigan Avenue, while fronting a structure also by Holabird & Roche, is itself by Sullivan. A different arrangement of the windows, a bolder moulding of the terracotta cladding of the piers—there were no intervening mullions now, any more than on his Troescher Building of 1884—and a strategic spotting of the chicory-like ornament—as well as, originally, a rich picture-frame-like band around the ground-storey shop-window—produce an entirely different effect. This effect is no less expressive of the underlying structure, but it represents a fuller and subtler deployment of architectural resources than Holabird & Roche provided on the façades next door.

The Gage Building was Sullivan’s penultimate major work. With the Carson, Pirie & Scott Department Store his career as an architect of big commercial buildings came to an end. This was designed in 1899 and the original three-bay and nine-storey section on Madison Street built in 1899-1901 for Schlesinger & Mayer; it was completed in 1903-4 for the present owners with the erection of the twelve-storey section that runs along State Street.[317] This building, which was Sullivan’s swan song, has also seemed to many critics his masterpiece (Plate 121). It lacks, however, the unity of the earlier Guaranty Building, having been built in two—indeed actually in three—successive campaigns. Despite the prominence of its site in the Chicago Loop, the store is inevitably overshadowed today by later and taller neighbours; nevertheless, it occupies a very high place in the Sullivanian canon.

There is no vertical emphasis except on the rounded pavilion at the corner, where continuous colonnettes rise the full height between the rather narrow bays; this feature was intended from the first but not built until 1903-4. The wide Chicago windows are crisply cut through the white terracotta sheathing just like the windows between the oriels on the Stock Exchange Building. The underlying grid of the structural steel frame—always more horizontal than vertical in effect, as the Reliance Building so clearly reveals—completely controls the surface pattern of the fenestration. On the Guaranty Building Sullivan emphasized the structural piers at their base by bending back the shop-windows of the ground storey; here it was the topmost storey that he set back, revealing the tops of the piers like little free-standing columns beneath the terminal slab in the spirit of his earlier eaves galleries. This treatment—most unfortunately replaced in 1948 by a flush parapet—increased very notably the effect of volume in much the same way as the parallel treatment at the base of the Guaranty.

At the base here, however, the shop-windows are carried up two storeys and given picture-frame-like surrounds, somewhat as on the Gage Building. In the cast-iron ornamentation of these frames, now much simplified, as also in that of the canopy on the north side and around the entrances in the rounded corner pavilion, Sullivan reached a peak of virtuosity in the lush decoration that has seemed to later critics quite at odds with the severe rectangularity of the façades above. There can be no question, however, that Sullivan considered ornament of the greatest importance in architecture and gave to its invention and elaboration his best thought and energy. It is certainly an interesting coincidence, moreover, rather than a matter of influence either way, that in these very years in Europe the newest architectural mode, the Art Nouveau, also put heavy emphasis on a somewhat similar sort of curvilinear decoration, often in association with exposed metal construction, and most notably on department stores (see Chapters 16, 17).

Sullivan’s ornament never had much influence either at home or abroad. Although Sullivanian skyscrapers of varying size and quality exist in many Middle Western and Far Western cities, most of them built in the first two decades of the new century, only the Rockefeller Building in Cleveland, built in 1903-6 by Knox & Elliot and extended laterally in 1910, really employs ornament, although of a drier and more geometrical order deriving from Owen Jones’s _Grammar_, in anything like Sullivan’s way. On Sullivan’s own late buildings, mostly tiny banks in small Middle Western towns, and in comparable work by his former assistant George G. Elmslie (1871-1952)[318] and William G. Purcell (b. 1880) the ornament tends to get more out of hand than on any of his skyscrapers of the nineties except perhaps the Condict Building. The best of Sullivan’s is the National Farmers’ Bank at Owatonna, Minn., of 1908; but Purcell & Elmslie’s Merchants’ National Bank in Winona, Minn., completed in 1911, might easily be mistaken for Sullivan’s work, for it is of comparable quality.

In the skyscrapers of the late nineties and the first two decades of the twentieth century designed in other Chicago architectural offices, such as D. H. Burnham & Co., Jenney & Mundie, and Holabird & Roche, there was rarely any attempt to vie with Sullivan as an ornamentalist but rather a continuance of the straightforward sort of design of the last-named firm’s Michigan Avenue buildings of 1898-9. A particularly fine and very large example is their Cable Building in Chicago of 1899. In the Fisher Building of 1897, also in Chicago, the Burnham firm more or less repeated the formula of the Reliance Building, but with a profusion of rather archaeological Late Gothic detail, eschewing the New York influence apparent in the Ellicott Square Building of 1895. Jenney & Mundie, rather more than the others, tended to follow the leadership of the New York architects of the day in using academic detail.

On the whole, the Chicago School continued to be vigorous, if not especially creative, down to the First World War, all the way through a period during which New York skyscrapers, still usually conceived as shaped towers rather than as plain slabs, received a succession of different stylistic disguises as they rose higher and higher. The forty-seven-storey (612-foot) Singer Building[319] of 1907 by Ernest Flagg (1857-1947) with its curious bulbous mansard—’Beaux-Arts’ of a quite aberrant sort—was followed by the campanile-like 700-foot Metropolitan Tower in Madison Square of 1909 by Napoleon LeBrun & Sons;[320] and that in turn by the cathedral-like Late Gothic elaboration of the Woolworth Building[321] of 1913 by Cass Gilbert (1859-1934), fifty-two storeys and 792 feet tall, which is still one of the major landmarks of downtown New York (Plate 178). A new flurry of skyscraper building followed in the twenties (see Chapter 24). The story with which this chapter is concerned, however, had reached its climax with the Chicago skyscrapers of the nineties, even though they were soon overshadowed in height and in contemporary esteem by the taller and more spectacular towers of Manhattan. Moreover, most of the big cities of the country, including Chicago, eventually sought to imitate the New York mode. But size is not, even in this period, a measure of quality, and the tallest skyscrapers are not the best, any more than the longest bridges are the most beautiful. So far the results of the revival of skyscraper building in the last fifteen years have rather confirmed this judgement (see Chapter 25).

A difficult question remains to be asked, even if it cannot be very satisfactorily answered: Why was the nineteenth-century development of commercial architecture, from Nash’s Regent Street to Sullivan’s skyscrapers, so completely an Anglo-American achievement? A few reasons may at least be suggested. On the Continent business activity was less concentrated in special urban districts in the nineteenth century, and was hence less likely to develop its own architectural programme. The big new nineteenth-century blocks in cities like Paris and Vienna and Rome generally serve a variety of purposes and almost always consist of residential flats in the upper storeys. In England and in America, on the other hand, most dwellings were still not flats but houses before 1900, and these fled farther and farther from the commercial areas as the nineteenth century progressed. The high property values in the central urban districts of the big Anglo-American cities, rising very rapidly in the second half of the century, encouraged the exploitation of their sites with taller and taller buildings. These values also helped to drive out the earlier inhabitants, leaving such areas as the London City and the Chicago Loop all but deserted after office hours.

Neither the office blocks of London and the big provincial English cities of the fifties and sixties nor, _a fortiori_, the skyscrapers of New York of the seventies and those of Chicago of the nineties can readily be matched elsewhere—except, of course, to some extent in the British Dominions and Colonies. Yet European cities do offer certain nineteenth-century commercial structures that are of real interest. The covered _passages_ and _galeries_, from the modest ones of the early decades of the century in Paris to Mengoni’s great Galleria Vittorio Emanuele II in Milan (Plate 75B) of the sixties, offered an urbanistic device of real significance. This is barely to be appreciated in the various extant English and American examples, such as the still flourishing Burlington Arcade in London or the Arcade in Providence, R.I., which is maintained as a historic monument though all but deserted by commerce.

Related to these structures serving multiple business purposes was the gradual development of the department store, a grouping together of various separate shops under one management and one roof, of which the Galeries du Commerce et de l’Industrie in Paris of 1838 were a relatively early example (Plate 62A). Exploiting like the _galeries_ the possibilities of iron-and-glass roofing, the early Continental examples of the department store had their more modest English and American counterparts such as Owen Jones’s Crystal Palace Bazar of 1858 in London or the Z.C.M.I. in Salt Lake City, founded by the Mormon leader Brigham Young himself and housed in cast iron in 1868.

The most notable later nineteenth-century department stores were in Paris and Berlin. In Paris the still extant Bon Marché of 1876 in the Rue de Sèvres by L. C. Boileau (1837-?), son of the builder of several Second Empire churches of iron, and the engineer Eiffel and the Printemps at the corner of the Rue de Rome and the Boulevard Haussmann of 1881-9 by Paul Sédille (1836-1900) were remarkable in conception if without much distinction of design. However, the Bon Marché is now completely masked externally by a masonry façade of the 1920s, and little of interest remains visible inside the Printemps. Of the portion of the Wertheim Department Store in Berlin built by Alfred Messel (1853-1909) in 1896-9 nothing survives.

Just after 1900, when the metal-and-glass construction of the interiors of department stores came to be generally exposed externally, this line of development came to its climax (Plates 131B and 133). This climax is so closely associated with the decorative and architectural development called Art Nouveau that the later Continental department stores may better be discussed in connexion with that (see Chapters 16, 17). Being of exposed metal, however, not of masonry-sheathed ‘skyscraper construction’ and relatively low, these stores are closer in character to the cast-iron commercial buildings of the third quarter of the century in America and Britain than to the tall Chicago structures of 1890-1910.

Steel construction of the American type, with the internal skeleton carrying a protective cladding of masonry, has gradually spread since the opening of the century to all parts of the world that produce or can afford to buy structural steel. It was, for example, introduced into London by the Anglo-French architects Mewès & Davis in building the Ritz Hotel there in 1905. Yet it remains typically American. In most other countries reinforced concrete rivals or completely takes its place as the characteristic material for building large structures of all sorts. The story of reinforced concrete had its technical beginnings in the mid nineteenth century; but it was not before the nineties that it first began to be exploited on a large scale and for conscious architectural effect. The first important reinforced concrete buildings, French like most of the best department stores of around 1900, will be mentioned later (see Chapter 18).

The whole picture of architecture in the twentieth century, so different from the picture of architecture before 1850, was modified by the developments that culminated in the Chicago skyscrapers. However important this has been for all later architecture both technically and aesthetically, it is important to stress here, as with the mid-century monuments of iron and glass, that the successive stages in the development are not solely, or even primarily, of premonitory and historical interest. From Parris’s granite buildings in Boston of the twenties, through the arcaded English commercial work of the fifties and sixties, to Richardson’s Field store and Sullivan’s skyscrapers in Chicago, St Louis, Buffalo, and New York, enlightened commercial patrons demanded and often received the best architecture of their day. The functional and technical challenges of commercial building seem to have brought out the creative capacities of three generations of architects as no other commissions did so consistently. Compare Parris’s Grecian temple church, St Paul’s in Boston, with his granite ‘skeleton’ fronts beside the Quincy Market (Plate 112B); set Godwin’s Stokes Croft Warehouse beside his town halls (Plates 113 and 92A); measure Richardson’s Field store even against his Pittsburgh Jail (Plates 116B and 108B). Then the strictly _architectural_, as well as the technical and social, significance of the major commercial monuments of the nineteenth century will be evident.

This chapter has summarized what was probably the greatest single innovation in nineteenth-century architecture, the rise of a new type of building to a position of prestige and of achievement comparable to that of churches and palaces in earlier periods. The same cannot be said of domestic architecture. The house was hardly a nineteenth-century invention like the office building. It was, however, modified almost beyond recognition as the century progressed, at the hands of several generations of creative architects. Around 1900 there are few if any churches, for example, to rival Sullivan’s skyscrapers in quality; but there are some houses, especially several by his disciple Wright and by his English contemporary Voysey.

------------------------------------------------------------------------

## CHAPTER 15

THE DEVELOPMENT OF THE DETACHED HOUSE IN ENGLAND AND AMERICA FROM 1800 TO 1900

IN the long story of man’s dwellings from prehistory to the present, the Anglo-American development that took place in the hundred years between the 1790s and the 1890s is of considerable significance, particularly as it provides the immediate background of the twentieth-century house. Architectural history has generally been little concerned, in dealing with periods earlier than the eighteenth century at least, with the habitations of any but the upper classes. The study of rural cottages in various regions of the world has been more a matter for anthropological investigation; the housing of the urban poor, when that was other than the makeshift adaptation of grander structures fallen into decay, remains for most early periods a matter of mystery. We know that ancient Rome had its blocks of middle-class flats of many storeys; although the links are not easy to recover, there was certainly some continuity in Mediterranean lands between that form of urban housing in antiquity and what can be traced from the medieval period down to the nineteenth century. Northern Europe in the late Middle Ages saw rather the development of individual urban dwellings with party walls, ancestors of the terrace-houses that first appeared in England in the seventeenth century.

The detached house of moderate size, so familiar today, the principal type of dwelling to undergo notable development in the nineteenth century in Anglo-Saxon countries, has no such remote Classical origins as the Continental flat or apartment. It made its appearance as the dwelling of the yeoman when economic conditions in late medieval England encouraged the rise of a class between the feudal landowner and the peasant parallel to the skilled artisan class in the towns. The conditions of settlement of the British colonies in America,

## particularly in New England, encouraged the continuation through the

seventeenth century of this type of dwelling almost to the exclusion of any other sort, since towns were then small and large estates rare. Around 1700 in America, though considerably earlier in England, relatively advanced contemporary modes began to have some influence on the design of such houses. With a lag of as much as a quarter of a century, the architectural developments of the home country were generally followed in the colonies; nor did political independence much affect the dependent cultural relationship in this field after the American Revolution.

The effects of the Picturesque point of view on the development of the house in England around 1800 were several (see Chapter 6). On the one hand, the newly fashionable attitude gave prestige to modest detached dwellings, raising the social status of the ‘cottage’ from an agricultural labourer’s hovel to a middle-class habitation or even on occasion a holiday ‘retreat’ for the upper classes—at first by adding the French adjective _orné_ (Plate 122A). At the same time the status of the ‘villa’ tended to be reduced from a large Italianate mansion on its own estate to a moderate-sized house at the edge of town. In much of the prolific architectural literature of the period, the hierarchy of residential building types was Rousseauistically inverted as rustic models, both native and Italian, were proposed for emulation in edifices of fairly considerable size. Thus several modes of informal design that had made their eighteenth-century debut in garden ornaments received more serious attention from architects as they came to be considered suitable for medium-sized dwellings and even sometimes for quite large mansions. As we have already seen, the towered Italian Villa was first introduced as a modest detached house by Nash at Cronkhill in 1802. It was similarly utilized by Schinkel (Plate 14A) and Persius at Potsdam a generation later, although Royalty still preferred to dwell there in Grecian dignity or Castellated pomp (see Chapter 2). Somewhat later, however, the Italian Villa provided (none too happily) a Royal retreat when Prince Albert decided on this mode for Osborne House on the Isle of Wight in the mid forties.

Not all Picturesque modes were equally adaptable to middle-class dwellings. The Indian found its most notable realizations in a large country house, S. P. Cockerell’s Sezincote, and a Royal folly, Nash’s Brighton Pavilion (Plate 48). There were, however, considerably later American examples[322] on a somewhat more modest scale, such as Iranistan at Bridgeport, Conn., built for Barnum in 1847-8, and Longwood, near Natchez in Mississippi, designed by Samuel Sloan in 1860 that have been mentioned earlier. But the Indian mode contributed the veranda, henceforth an integral feature of American domestic architecture, though rare after the Picturesque period in England. Verandas very early lost the Oriental detail, however. In front of Rustic Cottages they were often supported by bark-covered logs, but they could also acquire the formal character of Italian loggias, Tudor arcades, Swiss galleries or, most frequently, Classical porticoes and ‘pilastrades’ when adapted for use with other current modes.[323] In some cases the veranda, carried on occasion to two storeys in height, became the main theme of the exterior, yet was detailed so simply that no modish name properly applies (Plate 122A).

Even the Castellated mode, although used mostly for rather large houses (Plate 49), encouraged loose asymmetrical massing of the sort that is still more characteristic of the towered Italian Villa.

The Picturesque was thoroughly eclectic, in both possible senses of the word, as well as occasionally original. On the one hand, the point of view encouraged the parallel use of diverse modes. In theory, these were to be chosen according to their suitability to various sorts of natural settings, but in practice several were often employed side by side, as in Nash’s Park Villages in London, begun in 1827, and in the contemporary and later development of comparable suburban areas both in England and in America. On the other hand, the combination in one design of features derived from several different modes was allowable, even praiseworthy—low-pitched roofs with very broad eaves borrowed from the Swiss Chalet, towers from both the Castellated Mansion and the Italian Villa, bay-windows from the Tudor Parsonage, and verandas from the Indian were all part of a common repertory exploited rather indiscriminately. Basic to the Picturesque point of view and often determinant of choice of mode and even of individual features was the preoccupation with the natural setting; verandas, loggias, bay-windows and prospect towers were desirable, even necessary, features because they made possible the fuller enjoyment of the circumambient scene.

[Illustration:

Figure 21. T. F. Hunt: house-plan (from _Designs for Parsonage Houses_, 1827) ]

All these features affected house-plans in detail; but domestic planning in general was not as consistently re-organized as might have been expected, if only because the Picturesque point of view was so predominantly visual rather than practical in its usual concerns. Asymmetrical massing allowed, even forced, asymmetrical planning, however, thereby encouraging functional differentiation of the disposition and the sizes of various rooms (Figure 21). Yet very often, behind irregular exteriors, the plans were only slightly dislocated from the formal patterns of the preceding Palladian period. Although the increased articulation of most house-plans allowed the introduction of windows on several sides of many rooms, more significant at this stage was the frequent use of irregular shapes for the larger rooms, their main rectangular spaces complicated by external oriels and by internal ingle-nooks. None of these individual changes can be very precisely dated, at least in the current state of knowledge of the development of the house-plan in this period. Almost all of them were generally familiar in England by 1810. Tudor Parsonages, whether or not occupied by members of the clergy, were likely to be most adeptly planned.[324] In them the well-defined needs of a family of relatively high social status but low income encouraged a more efficient grouping of the rooms and a clearer distinction of separate functions—entrance hall, drawing-room, dining-room, study, kitchen, scullery—than had been common earlier in such medium-sized dwellings.

In the first third of the century the various Picturesque modes of house-design were very widely exploited in England for middle-class habitations in the new suburbs, having generally made their first appearance a decade or so earlier in lodges or other accessories to large private estates. They were also popular at the new seaside resorts, such as Sidmouth in Devon and Bournemouth in Hampshire, where they often housed more exalted clients. At Sidmouth, for example, what is now the Woodlands Hotel was remodelled from a barn into a barge-boarded Cottage Orné by Lord Gwydyr in 1815; the nucleus of the Knowles Hotel there was Lord Despenser’s cottage of a few years earlier; and the Royal Glen Hotel, a modest Castellated house then known as Walbrook Cottage, was built early enough to house Queen Victoria as a baby. Although the prestige of the Picturesque declined rapidly in high aesthetic circles after 1840, the rigorous principles of Pugin and the ecclesiologists had little effect on the operations of suburban builders, who continued for decades to follow the various well-established modes of a generation earlier.

As Latrobe’s ‘Gothick’ Sedgley, built outside Philadelphia in 1798, and various other Neo-Gothic structures in Philadelphia and Boston of the first decade of the new century make evident, the Picturesque came early to the United States. Yet it was hardly before the thirties that the various Cottage and Villa modes began to compete at all with the Greek temple and the formal post-Palladian house modernized by the use of Grecian detail; only with the appearance in 1842 of _Cottage Residences_ by A. J. Downing (1815-52)[325] were they enthusiastically propagated.

Earlier, new developments in the planning of the ubiquitous moderate-sized free-standing houses were not very notable in America. In the 1790s the influence of Adam, and possibly of the French, encouraged some experimentation with variously shaped rooms; but this largely died out as the necessary rectangularity of the Greek temple house, only extended by one or more wings in the largest examples, reimposed the formal Anglo-Palladian plan with central stair-hall and four nearly equal-sized corner rooms. For smaller houses with pedimented fronts, however, a sort of terrace-house plan was increasingly popular, with stair-hall at one side, two principal living rooms one behind the other, and a narrower kitchen wing extending to the rear. A planning innovation that first appeared in America in the 1790s, by no means unknown earlier in England but rare except in terrace-houses, was the opening together of two rooms—front and back parlours—by means of broad sliding doors. This became increasingly common after 1800. Moreover, the temple portico provided the equivalent of a shallow veranda across the front of the house and was sometimes replaced or supplemented by a deeper colonnaded porch at the sides or rear. The veranda, indeed, had reached the southern states fairly early in the eighteenth century, arriving from the East via the West Indies. In its usual two-storeyed form it was easily merged with the monumental colonnades demanded by the Grecian mode (Plate 38B).

Thus, even before a rather belated wave of strong Picturesque influence began to drive out the temple house in the forties, early nineteenth-century American houses had certain definitely post-Colonial characteristics in their plans. Of later house-planning in the United States in the forties and fifties almost everything that has been said about English planning in the preceding decades applies (Figure 22). By this time in England, however, newer planning ideas were being introduced by leading architects in relatively large houses. At Scarisbrick, for example, where the remodelling and extension of the existing Georgian house began in 1837, Pugin revived the medieval great hall (see Chapter 6). A few years later in his own house, The Grange of 1841-3 at Ramsgate,[326] by no means a mansion in size or scale, the more modest two-storey hall incorporates the staircase and also provides, with the galleries above, the central core of communication. Parallel with these examples, which were of Gothic inspiration, Barry at Highclere adapted the glass-roofed central _cortile_ of the Reform Club to domestic use, associating with it the main staircase rising in a contiguous vertical space.

At the hands of High Church architects the parsonage, by definition no mansion but a modest free-standing gentleman’s residence, was also undergoing a characteristic development. No longer Tudor, of course, it was still not forced to be archaeologically decorated in its planning, since there were few if any relevant medieval models to imitate. The doctrine of ‘realism’ condemned the shabby construction and careless use of materials that had too often been characteristic of Picturesque house-building in the previous decades, while the need for economy discouraged the ornamentation common on contemporary churches.

Such a vicarage as that which Butterfield built in 1844-5 to go with his ‘first’ church, St Saviour’s at Coalpitheath, Gloucestershire, is a model of simple masonry construction. In the random ashlar walls are set wide banks of plain mullioned windows, Gothic only in the arching of their heads, where they can serve best to light the various rooms (Plate 122B). The massing also is irregular yet orderly with several high gables, a porch, many tall chimney stacks, and a broad bow-window elaborating the basically rectangular block. But, in the language of the ecclesiologists, ‘the true Picturesque derives from the sternest utility’, and so all these projecting features were such as could be readily justified functionally, like the ritualistic articulation of contemporary churches. The plan of Butterfield’s vicarage has the virtues of those of the Picturesque Tudor Parsonages in the variety of room-sizes and shapes provided and also in the opportunities that the windows offer to enjoy surrounding nature. There is also at Coalpitheath a very modest version of Pugin’s stair-hall at The Grange, not a mere lobby but a central space designed for easy horizontal and vertical communication.

Any serious revival of medieval craftsmanship in masonry was all but impossible in America; in any case it was largely irrelevant in a land where most houses were built of wood. But in reaction to the white-painted clapboards and the smooth Grecian trim of the previous decades, echoing however humbly the marble of Greece, Downing in the early forties proposed and many at his behest adopted variant treatments for the exterior sheathing of Picturesque villas and cottages that were rather more expressive. The distinguished native craftsmanship evident in the more monumental edifices of the Greek Revival executed in fine ashlar of granite or other light-coloured stone, or else in smooth red brick, died out. Such materials had no more appeal than did crisp white-painted wood to a generation indoctrinated with the Picturesque point of view. Yet clapboards remained the usual surfacing material for wooden houses, even if they were now painted, not white, but in the stony hues—grey or beige—that Downing recommended in his books with actual coloured samples.

[Illustration:

Figure 22. A. J. Downing: house-plan (from _Cottage Residences_, 1842) ]

The treatment Downing preferred was board-and-batten.[327] This he made a constituent element of the very original Bracketted mode that he offered as an American alternative to the imported Italian Villa and Tudor Parsonage which he was energetically engaged in nationalizing. Board-and-batten provides a stronger pattern of light and shade, and also the verticalism that appealed increasingly to mid-century taste. This sheathing also offers a sort of symbolic expression of the light ‘balloon-frame’[328] construction that was beginning to come into general use by the fifties, though this method of wooden framing was apparently never known to Downing, since he died in 1852 before it reached the eastern states where he lived and worked.

With their board-and-batten walls, their ample verandas, and their bay-windows, what are still usually called ‘Downing houses’ constitute a largely original American creation in spite of the frequent use of Tudoresque detail on barge-boards and veranda supports and even of elaborately moulded terracotta chimney pots. Yet in their planning the houses designed by Downing and his architect friends Davis and Notman do not advance much beyond the models published in the English books of the previous decades that were their immediate prototypes (Figure 22). The verandas are usually wider and more prominent, however, and the front and rear parlours are likely to open into one another, as sometimes also into a modest central hall.

In America as in England, the Picturesque period came to no sudden end. The recurrent publication of Downing’s books even after the Civil War[329] indicates how long his models remained favourites with American builders and their small-town and suburban clients. However, even before the Civil War a mansarded Second Empire mode was beginning to become popular (see Chapter 9). With the wide acceptance of this and of the High Victorian Gothic there developed a rather sharp split between autochthonous and imported types of house-design, drastically though the imported types were usually Americanized outside the bigger eastern cities. To this situation we must return later.

Something has already been said of the major turn that took place in the development of the English house around 1860 (see Chapters 9 and 12). When seen in relation to the parsonages that his master Street and also Butterfield had been building in the previous fifteen years, Webb’s Red House built in 1859-60 for William Morris is considerably less revolutionary than has sometimes been supposed. Had this been built in Gloucestershire rather than in Kent, it would certainly have been of stone like Butterfield’s Coalpitheath Vicarage; as it is, the entrance porch is no simpler or less Gothic than Butterfield’s. The particular window forms, moreover, can be matched in Butterfield’s Clergy House and School at All Saints’, Margaret Street, and the somewhat rustic ease of composition in his cottages at Baldersby St James. Yet the planning here is highly individual, suited to the special needs of a client who was an artist and a writer, not a parson.

The next house that Webb built, now known as Benfleet Hall, Cobham, begun in 1860 for the painter Spencer Stanhope, has been less publicized, and it never had the rich furnishings that Morris and his associates designed and executed for the Red House. Yet it is perhaps more significant in the general history of the Anglo-American house. There is here, for example, a small stair-hall of the order of Pugin’s at the Grange or Butterfield’s at Coalpitheath around which the other ground-storey rooms are loosely grouped. The particular character of the plan can, in fact, best be matched at Hinderton, a small country house in Cheshire that is hardly more of a mansion than Benfleet, which Waterhouse built in 1859. This house is in Waterhouse’s gawkiest High Victorian Gothic, with none of the simplicity and delicacy of Webb’s early houses. It is rather unlikely that Webb was actually emulating it, but the plan was twice published[330] and hence soon known abroad.

Webb’s Arisaig in Inverness-shire was begun in 1863 (Figure 23). Built of local stone, it is somewhat more conventionally Gothic externally; moreover, it is of country-house size, a mansion rather than a modest artist’s dwelling like the Red House or Benfleet Hall. The plan has two major aspects of interest: the two-storeyed hall, with gallery above, occupies a central position and the principal rooms on both storeys are very efficiently grouped about it within the bounding rectangle of the main block of the house. In other words, Arisaig’s hall seems to derive as much from the Highclere sort of glazed central court as from Pugin’s revival of the medieval great hall.

Cloverley Hall, which was built by Nesfield and Shaw in 1865-8, attracted much favourable contemporary attention largely because of the superb craftsmanship of the brickwork and the originality of the _japoniste_ ornament (see Chapter 12). It is destroyed now except for the extensive service and stable wings and the gate lodge; but the amount and the character of the fenestration, providing in some areas what amounted to window-walls of stone-mullioned and transomed lights, and the character of the plan make it still memorable. It was also the first of the many notable Late Victorian manor houses which both Nesfield and Shaw would build when working alone.

[Illustration:

Figure 23. Philip Webb: Arisaig, Inverness-shire, 1863, plan ]

Like Arisaig, Cloverley was a large country house. The medieval great hall, first rather modestly revived by Pugin at Scarisbrick, here returned at full scale; but it was placed in a corner of the main block—as was occasionally its position in the sixteenth century—so that it might receive light from one end as well as from the side (Figure 24). From the entrance, however, one passed by this hall through the ‘screens’ under a gallery to arrive at a stair-hall, more in the manner of Waterhouse’s and Webb’s, around which the other principal rooms were compactly grouped. There was also here a very skilful play with levels, the hall being lower than the rest of the main floor, and therefore part-way down to the basement—containing a billiard room and so forth—which was entirely above ground at the rear of the house.

[Illustration:

Figure 24. Nesfield & Shaw: Cloverley Hall, Shropshire, 1865-8, plan ]

While Cloverley Hall was still in construction, Shaw had begun his own personal career as a house-builder at Glen Andred in 1866-7 (Plate 102B), where he introduced a more vernacular manner (see Chapter 12). Following this came his Leyswood in 1868-9, a mansion as large as Cloverley Hall and in some of its decorative features more archaeologically Late Medieval. As at Cloverley Hall, the amplitude of the fenestration, however, arranged here in long mullioned bands as well as in tall window-walls, has seemed more significant to posterity than the stylistic detailing[331] (Plate 123). Above all, Leyswood marked a further stage in the development of the ‘agglutinative plan’ (Figure 19), of which the first well-publicized example was Waterhouse’s at Hinderton. Here the great hall and the stair-hall of Cloverley are combined to form a central spatial core of communication, somewhat as at Webb’s Arisaig, but the shape of this is quite irregular and the reception rooms are grouped very loosely about it, more as at Benfleet Hall. Projecting well out of the main block, the dining-room and the drawing-room both receive light from three sides. Moreover, the space of these rooms is articulated, as in certain Picturesque houses of forty and fifty years earlier, by ingle-nooks, oriels, and various other irregularities. Perspectives of Leyswood—not the plan[332]—were published in the supplement to the _Building News_ of 31 March 1871 and made at once a tremendous impression both in England and in America (Plate 123).

[Illustration:

Figure 25. Philip Webb: Barnet, Hertfordshire, Trevor Hall, 1868-70, plan ]

In a house by Webb of the same date as Leyswood, Trevor Hall at Oakleigh Park, Barnet, in Hertfordshire, the arrangement of the rooms about the central hall was much more compact (Figure 25). The whole formed a square and allowed a quite symmetrical treatment of the three principal fronts. This house is now destroyed except for the gate lodge. Less interesting in plan but significant for its very modest size is Webb’s Upwood Gorse, Caterham, Surrey, built for Queen Victoria’s dentist Sir John Tomes also in 1868. The consistency and the simplicity with which the local vernacular of brick below and tile-hanging above is handled in connexion with plenty of white-painted Queen Anne sash-windows regularly but not symmetrically spaced offers a curiously close prototype of the American ‘Shingle Style’, although the initiators of that mode a decade later can hardly have known of this house, since it was never published. It was rather Shaw’s houses of the next decade, of which his drawings were exhibited each year at the Royal Academy and given great prominence in the professional Press, that provided the exemplars which architects generally imitated both at home and abroad; from 1874 on the plans were usually illustrated as well as Shaw’s own very virtuoso pen-drawn[333] perspectives (Plate 123).

Webb’s houses for the painters Val Prinsep and G. B. Boyce in Kensington and Chelsea, of 1865 and 1869 respectively, were the first English ‘studio-houses’—houses, that is, in which the studios, naturally equipped with very large windows, were the principal rooms. These provided a more livable alternative to the great halls that Shaw generally provided in his country houses; but it was the larger artists’ houses of the seventies and eighties which Shaw built for his fellow academicians that received contemporary publicity.

By the mid seventies Shaw was moving in the formal and symmetrical direction initiated by Webb at Trevor Hall and soon carried much further by Nesfield at Kinmel Park as regards both the planning and the external organization of his larger London houses. Lowther Lodge in Kensington Gore of 1873-4 is the first of his domestic commissions that may properly be called Queen Anne rather than Manorial. The even more formally designed Cheyne House and Old Swan House, of 1875 and 1876 respectively, on the Chelsea Embankment followed shortly after (Plate 103); but he long continued to build more loosely composed houses in the country, as has been noted earlier.

Before turning to the results of Shaw’s very notable influence in the United States in the seventies, something should be said of the situation there in the preceding decade. The Second Empire mode had been increasingly popular for houses from the mid fifties and was especially fashionable during the boom period that followed the Civil War. It had no positive contribution to make to the general Anglo-American development in these decades, however. In the domestic field more or less Gothic modes were its significant rivals; first Downing’s wide-veranda-ed version of the Tudor Cottage; then, after 1860, what Vincent Scully has christened the ‘Stick Style’.[334]

On houses in this mode, which is really hardly Gothic at all, a sort of imitation half-timbering panels the exterior walls, suggesting, like Downing’s board-and-batten sheathing, the underlying wooden stud-structure of balloon-frame construction. This construction came to be generally used in the East as well as in the Middle West, where it originated, after it had been explained by William E. Bell in his _Carpentry Made Easy_ in 1858. More striking is the open stickwork of the ubiquitous verandas. This can be seen in an early form on the Olmsted house in East Hartford, Conn., of 1849 by the English architect Gervase Wheeler,[335] who obviously derived it from Picturesque models in England dating back at least to the thirties. In the J. N. H. Griswold house of 1862 in Bellevue Avenue in Newport, R.I., by the French-trained Hunt, now the Newport Art Association, the ‘sticks’ of the wall surface are so sturdy that they may well be the actual framing members.

Very characteristic of the maturity of the mode is the Sturtevant house at nearby Middletown, R.I., built by Dudley Newton (1845?-1907) a decade later in 1872. Here the gawky vigour of the Stick Style, its intense woodenness, and its descent from several different Picturesque modes—not least the Swiss Chalet—is very evident (Plate 124A). Extensive surrounding verandas are of the very essence of the mode; but the internal planning, while informal and often asymmetrical, is rarely very open. Several books by Eugene C. Gardner (1836-1905)[336] of Springfield, Mass., give a sophisticated architect’s rationale of the mode. But the exemplars that G. E. Woodward[337] offered in the sixties are more typical, and were more widely imitated in actual production; for the Stick Style had almost run its course by the time Gardner began to present his excellent house designs. Woodward was no architect, and for the most part the Stick Style should not be considered an architect’s mode. It represented rather a popular attempt, remarkably successful for a few years, to create an American domestic vernacular, suited to the materials in general use and to the current methods of building, comparable to Downing’s earlier Bracketted mode. Like the Second Empire vogue the Stick Style died out, at least in the East, during the general hiatus in building production after the financial Panic of 1873.

By that time Shaw’s influence had begun to reach America.[338] Moreover, the possibilities of agglutinative planning about a great hall had been realized by Richardson well before a Shaw plan—that for Hopedene—was first made available in the _Building News_ in 1874. It is, of course, possible that McKim, in passing through England on his way home from Paris in 1870, had seen (or merely heard of) the character of Webb’s, Nesfield’s, and Shaw’s houses of the sixties and transmitted that information to Richardson.

An undated project of about 1871 by Richardson for a house to be built in Newport, R.I., for Richard Codman includes his first great hall[339] of the Shavian sort; but the Codman plan is already in advance of, or at least rather different from, those of Shaw. This hall, out of which the stairs would rise in an L-shaped at the rear, was to be very large in relation to the other rooms, and thus definitely a principal living area not a mere foyer or centre of circulation. The drawing room and dining room were to open out of the hall through wide doorways so that some sort of spatial continuity would have extended through all the reception rooms of the ground storey. There was to be a large veranda at the rear in the well-established local tradition. The exterior as shown in the elevations is not at all Shavian but rather related to the Stick Style, like Richardson’s own house at Arrochar on Staten Island of 1868.

Richardson’s first executed country house, the F. W. Andrews house of 1872-3 at Newport, R.I., was much more Shavian in plan. Four or five rooms were grouped about a relatively smaller central stair-hall and most of these were articulated by bay-windows and ingle-nooks. But the main block was also surrounded by verandas, features which are rare and always of modest extent on Shaw’s houses. The Andrews house was burned a long time ago, but from the existing elevations it would appear that the external treatment represented a sort of transition between the Stick Style, then at its apogee, and Shaw’s Surrey vernacular translated into American materials. The verandas were still detailed in a Stick Style way, and flat stickwork interrupted the continuity of the wall surfaces; but the clapboarding of the lower walls evidently took the place of the brickwork Shaw used—it was almost certainly painted red—and the wooden shingling of the upper walls was a happy substitute for English tile-hanging. Shingles were, of course, an old though largely forgotten American sheathing material long used especially for roofs.

By the time Richardson came to design his next large house, that for William Watts Sherman on Shepard Avenue in Newport in 1874, the perspectives of several of Shaw’s manors had appeared in the _Building News_ and the plans of two. As a result, probably, of his assistant Stanford White’s Shaw-like skill with the pencil, the Sherman house was notably Shavian externally. Above the ground storey, which is of Richardsonian random-ashlar masonry in pink Milford granite with brownstone trim, the walls and the high roofs are covered with shingles cut in various decorative shapes suggested by those of Shaw’s tile-hanging. Many of the casement windows are grouped to form window-walls in the ground storey and arranged in long horizontal bands above. The half-timbering of the front gable, with painted decoration on the intervening plaster, was taken straight from Shaw’s Grim’s Dyke; the carved ornament on the barge-boards is almost Nesfieldian in its suggestion of _japonisme_. Thus the whole is as perfect a specimen of Shaw’s Manorial mode as anything any architect other than he or Nesfield ever produced in England. The house has since been much enlarged, partly by White in 1881, partly by Newton very much later, but always with due respect for the character of the original design.

The plan has more of the independent virtues of that of the Codman project. The hall provides a principal portion of the living area, and the other main rooms open into it through wide doors; thus there is some flow of space throughout the whole original block. The original library at the rear corner, later replaced by a large ballroom, ended in a Shavian rounded bay with a continuous window band, a feature Wright would copy later. Yet otherwise the house was less articulated than Shaw’s earlier ones, having rather the compactness though none of the symmetry of Webb’s Trevor Hall.

The mid seventies saw many other American reflections of Shaw’s Manorial mode and soon of his Queen Anne also, none of them so successful as the Sherman house. But the deep business recession that followed the Panic of 1873 led to a general mood of repentance after the extravagances, architectural and otherwise, of the post-war boom. From the resultant nostalgia for the simpler ways of the American past there began to develop at this time a great interest in the houses of the Colonial period, an interest that readily merged, however, with the current English preoccupation with the vernacular of the seventeenth and eighteenth centuries. To an extent difficult for posterity to appreciate, the nascent ‘Shingle Style’,[340] which crystallized towards the end of the decade with the revival of building production, was to its protagonists already a sort of Colonial Revival. Although its origins are partly Shavian, it represents above all a reaction, as did Shaw’s Manorial mode in England, against the ‘modernism’ of the High Victorian Gothic and the Second Empire, now grown thoroughly unfashionable except in the West.

Boston was still the architectural metropolis of the United States, and it was around Boston, especially in the work of Emerson and Little, the latter a serious early student of old Colonial work, that this crystallization of the Shingle Style first took place (see Chapter 13). But it was at once taken over and given a somewhat more Shaw-like elaboration by the New York firm of McKim, Mead & White, formed in 1879. From the early eighties, and for over a decade, the Shingle Style was widely practised by architects from coast to coast, and not least happily in the Far West. The characteristic use of shingles as an all-over wall-covering emphasized the continuity of the exterior surface as a skin stretched over the underlying wooden skeleton of studs, in contrast to the way the preceding Stick Style had echoed that skeleton in the external treatment. The shingles properly provide the name for a most characteristically American domestic mode; but it was in planning that American architects made the really original contribution in what was the most significant development of the detached house since the Picturesque period.

[Illustration:

Figure 26. W. R. Emerson: Mount Desert, Maine, house, 1879, plan ]

One of the first mature examples of the Shingle Style, a house built by Emerson on Mount Desert in Maine in 1879, well illustrates the virtuosity of the new planning (Figure 26). Rooms of varied shape and size are loosely grouped about the hall and open freely into one another. The various levels of the different areas are related to the landing levels of the elaborate staircase. Above all, it should be noted that the verandas are not mere adjuncts or afterthoughts, as they were even on Richardson’s Andrews house, but major elements, both space-wise and visually, of the whole composition. Such houses parallel in their three-dimensional complexity the massing of the Italian Villas of the earlier nineteenth-century decades with their loggias, pergolas, and prospect towers, yet they bear little or no visual resemblance to them, since the later houses are always much more sculpturally plastic and less articulated in composition. The windows are generally of double-hung small-paned sashes of a type at once Queen Anne and Colonial, but they are frequently grouped in the Shavian way, as well as being ingeniously placed in order to vary the internal lighting effects, so that the pattern of fenestration is not at all of an eighteenth-century order.

Richardson certainly did not initiate the Shingle Style; but he took it over in 1880 and made it very much his own, using it for all his later country and suburban houses. Dropping all detail, whether Richardsonian Romanesque, Shavian Manorial, Queen Anne, or American Colonial, he retained much of the ease and casualness of Shaw’s best early houses. But there is also a great deal of similarity to the simple massive effects of the old Colonial houses also. Spiritually, so to say, if not so much visually, Richardson’s shingled houses most resemble Webb’s best work; of these Richardson presumably had no knowledge, although it is just possible that he might have seen some when he was in England in 1882, well after the Shingle Style was fully established.

Richardson’s Stoughton house in Brattle Street in Cambridge, Mass., of 1882-3 is perhaps his best shingled one, at least in the relatively untouched form in which it, almost alone, alas, has come down to us (Plate 124B). It certainly shows little evidence of the interest that he is known to have taken in Burges’s and Shaw’s work while he was abroad just before this. The entrance, originally, was through the loggia recessed into the main mass of the house (it is now from Ash Street on the left). The living-hall extends, as in the Sherman house, from front to back and the stairs sweep up in a quarter-circle over the entrance. The drawing room at the corner and the dining room behind the loggia both open into the hall through wide doors; only the small library is isolated from the general flow of space. Externally, the shingled surfaces, broken only by banks of double-hung windows, model the complex mass into a unified composition, the almost submerged stair-tower successfully linking the two gabled wings at right angles to one another by its rounded form. There is no ornament of any sort, and the weathered grey of the shingles is varied only by the dark-green paint of the window sash.

McKim, Mead & White’s houses of the early eighties, several of them equally fine, are usually rather more elaborate in their massing and are likely to be enlivened with much imaginative detail.[341] Some of the detail recalls this or that style of the past, but all of it is thoroughly personalized by White’s delicate hand. One of their best houses is the one for Isaac Bell, Jr, built in 1881-2 in Bellevue Avenue in Newport, R.I. (Plate #126:pl126). This is less unified externally than the Stoughton house but more open in plan (Figure 27). A wide veranda, with very elegant bamboo-like supports, extends around two fronts, expanding into a two-storeyed open pavilion on the right. This pavilion provides a semicircular void to balance the round tower at the rear left corner. The patterns of the original cut shingles on this house, although obviously suggested by English tiling, are much softer and more graceful, almost bringing to mind birds’ plumage.

Inside, the hall is articulated by a wide ingle-nook, rather dark and low, in sharpest contrast to the great flight of stairs beyond down which floods light from the window-wall at the half landing. Twenty-five-foot sliding doors, hung from above, make it possible to open the drawing room through almost its entire length into the hall. The Bell dining room, connecting at its end through French windows with the curved portion of the veranda, has some of the finest of White’s orientalizing detail. This is much more original than that in the new library he decorated at this time in the Sherman house or the dining room he added to Upjohn’s Kingscote, both also in Newport.

McKim, Mead & White’s slightly earlier H. Victor Newcomb house of 1880-1 in Elberon, N.J., is at once clumsier and more Shavian externally than the Bell house; but the spatial treatment of the living-hall is most original and very significant for later developments (Plate 125A). The main rectangular space, of which the shape is emphasized by the ceiling beams and by the abstract geometrical pattern of the floor, seems to flow out in various directions into other rooms and into several bays and nooks; but the actual room-space is sharply defined by a continuous frieze-like member that becomes an open wooden grille above the various openings. There can be little question that the major influence here is from the Japanese[342] interior, but from the Japanese interior understood as architecture. This is not just a superficial matter of Nesfieldian _japonisme_ such as White was employing so much in his ornament in these years. The Kingscote dining room has somewhat similar spatial qualities but more eclectic detailing and richer materials: marble, Tiffany glass tiles, cork panels, stained glass, etc.

[Illustration:

Figure 27. McKim, Mead & White: Newport, R.I., Isaac Bell, Jr, house, 1881-2, plan ]

In 1879 Cyrus McCormick had his Chicago mansion built by the local architect Adolph Cudell (1850-1910) and his partner Blumenthal in the form of a very corrupt Second Empire _hôtel particulier_. It is good evidence of the rapidity with which taste changed at this time that two years later he called on McKim, Mead & White to build for him in Richfield Springs, N.Y., one of the finest and most carefully composed of all their Shingle Style houses. This house is notable not only for the subtly Japanese character of the various sorts of veranda supports but even more for the way the composition is unified under the broad front gable by the long horizontal line of the veranda roof repeating that of the stylobate-like stone wall of the terrace below. It is most unfortunate that this house is now in a state of near-collapse.

Little’s contemporary Shingleside House of 1881 in Swampscott, Mass., has been mentioned already. Soberer than the Bell or the McCormick houses in its rectangular shape and almost total lack of exterior detail, this had a galleried two-storey hall with a window-wall as the principal living area. In the combining of different levels this house recalled a little Cloverley Hall, but it was completely Americanized in scale and in detail without being archaeologically Neo-Colonial.

By the mid eighties J. Lyman Silsbee (1848-1913) had introduced the Shingle Style to Chicago, and other Eastern architects were building good houses of this order in such Western towns as Cheyenne, Idaho; Colorado Springs, Colorado; and Pasadena, California. In Philadelphia Wilson Eyre (1858-1944) developed the mode with a very characteristic personal difference, often eschewing the use of shingles. If his exteriors are rather English in their frequent use of brick and real half-timbering, his plans are most original. The long rooms of varied and irregular shape are strung out on either side of halls from which rise stairs within grilled enclosures of a sort that appeared in England only in the houses of the nineties by Voysey and his contemporaries.

The heyday of the Shingle Style was brief, even though it continued in use well down into the nineties. The Colonial Revival implications, present from the first, soon encouraged more and more comprehensive use of eighteenth-century detail, and this supported the general tendency of the mid eighties in America away from the irregular and towards more formal order (see Chapter 13). Something of this change could be seen in Richardson’s latest houses in masonry such as the Glessner house of 1885-7 at 18th Street and South Prairie Avenue in Chicago, which still stands, and the contemporary Mac Veagh house, long since destroyed, also in Chicago, both of which were almost symmetrical as regards their front façades. The most drastic examples, of course, of this Academic Reaction were such houses as McKim, Mead & White’s Villard group in New York (Plate 109B) and their H. A. C. Taylor house in Newport with its formal Anglo-Palladian plan of central hall and four corner rooms. Despite its even tighter plan, however, their extant W. G. Low house in Bristol, R.I., of 1887—a year later therefore than the demolished Taylor house—can properly be cited again as a masterpiece of the Shingle Style (Plate 127). This illustrates very well how the loose massing of the houses of the early eighties could be organized into a carefully balanced composition without succumbing to any historical mode of design, whether Italian Renaissance or American Colonial.

## Particularly interesting in this connexion are the small houses at

Tuxedo Park, N.Y., which Price designed for Pierre Lorillard in 1885-6, some years before he began to build Renaissance skyscrapers (see Chapter 14). Lorillard’s own house has a rather tight plan of the Neo-Colonial sort; but the exterior with its paired chimneys on the front, a Richardsonian entrance arch between them, and the verandas and terrace treated as voids carefully related to the solid mass behind is still in the earlier tradition (Plate 125B). In such other houses by Price at Tuxedo as those for William Kent and Travis C. Van Buren, the loose open plans of the immediately preceding years were organized into T and X patterns, and the verandas and terraces were even more formally treated as important elements in compositions made up of well-defined voids and solids (Figure 28).

This brings us to the beginning of the career of Frank Lloyd Wright, already introduced as an important coadjutor of Sullivan from 1887 to 1893. Although Wright’s mature career begins only about 1900 (see

## Chapter 19), his apprentice years as a builder of houses provide a very

significant episode that is closely related to the earlier story of the nineteenth-century house in England and America. By the late eighties a full-dress Colonial Revival was under way in the East. But it was the

## particular combination of freedom and order that had been achieved by

Richardson in his latest houses, by McKim, Mead & White in their Low house, and by Price in his Tuxedo houses which was the immediate tradition from which Wright’s domestic architecture grew far more than the work of Sullivan.

[Illustration:

Figure 28. Bruce Price: Tuxedo Park, N.Y., Tower House, 1885-6 ]

Born in 1867, Wright had had some two years in the Engineering School—there was no architectural school—at the University of Wisconsin when he came to Chicago at the age of twenty in 1887. He first found work in the office of Silsbee whom Wright’s uncle Jenkin Lloyd Jones had brought to Chicago a year or two earlier to design All Souls’ Unitarian Church, of which he was minister. The young architect’s first work, nominally a Silsbee commission, was the Hillside Home School built in 1887 for his aunts near Spring Green, Wisconsin. This was a rather provincial specimen of a Shingle Style house and was later demolished by Wright himself.

Shifting over the following year to the Adler & Sullivan office, Wright by 1889 was married and ready to build a house for himself on the strength of a five-year contract with his new employers. This house, at 428 Forest Avenue in Oak Park, Ill., still extant but much pulled about, derives almost entirely from Price’s cottages at Tuxedo except that the plan is much less formal. In the interior, the wide openings between the rooms are not framed by architraves but seem to have been produced by pulling back the walls beneath the continuous frieze. In this treatment, rather Japanese in concept, Wright would seem to have been influenced by White’s handling of the hall of the Newcomb house, even though that is rather Japanese also in some of the detailing and Wright’s is not.

Wright’s next important work is the James Charnley house at 1365 Astor Street in Chicago, built in 1891-2. This was actually a commission of the Adler & Sullivan firm, but one of which he had entire charge. A city house built of tawny Roman brick like that used for the court of the Boston Public Library, this is as formal[343] as anything McKim, Mead & White had yet designed. But there is no High Renaissance or Colonial reminiscence whatever in the external detailing. The Charnley house is rather a conscientious attempt to emulate in a modest three-storey residence the highly original design of Sullivan’s newly completed Wainwright Building in Saint Louis.

Wright was also accepting various private commissions on the side, mostly very small ones, by this time. The George Blossom house of 1892 at 4858 Kenwood Avenue on the south side of Chicago, however, is of more consequence. Externally, this follows rather closely McKim, Mead & White’s Taylor house in the curved Ionic entrance porch and the recurrent Palladian windows, not to speak of the use of yellow-painted clapboards and white-painted trim of simplified academic character. Even the plan is for the most part symmetrically ordered. But behind the formal range of entrance lobby and two small corner rooms at the front the whole centre of the house opens up as a single great living-hall. In this living-hall a wide ingle-nook is lined up on axis with the entrance, the elaborate staircase rises in several flights across one end, and wide openings connect with the library and the dining room. The dining room, which ends in a curved bay with a continuous window-band, is almost a copy of the original library of Richardson’s Sherman house. In another Wright house of 1892, that for A. W. Harlan, also on the south side of Chicago, at 4414 Greenwood Avenue, which Sullivan happened to see, he recognized his assistant’s hand and this brought about the break between the two before Wright’s contract ran out.

When Wright set up for himself in 1893 there were two paths open to him. That he actually considered following the path of Academic Reaction, so heavily publicized by the success of the World’s Fair, is evident from his project of this year for a Library and Museum in Milwaukee (see

## Chapter 13). But when Burnham at this point offered to send Wright to

Paris to study at the École des Beaux-Arts and then to the new American Academy which he and McKim were planning to start in Rome, in preparation for taking him on as designing partner, the young architect turned the opportunity down.

The W. H. Winslow house of 1893 in Auvergne Place in River Forest, Ill., always considered by Wright his ‘first’, shares many qualities with the Blossom and Harlan houses, but is altogether a much more mature and original work (Plate 128A). The front is completely symmetrical and as formal as that of the Charnley house of two years before.[344] Broad and low, of fine Roman brickwork with a rich band of moulded terracotta the full depth of the upper-storey windows below the wide eaves, the general effect of this has usually been considered very Sullivanian. But as Wright himself was responsible for the Adler & Sullivan work that this house most resembles—the Charnley house, certainly; and the Victoria Hotel of 1892 at Chicago Heights, probably—it is more accurate to consider that the Winslow house represents a continuation of his own manner of the previous year or two. The plan is more axial and less open than that of the Blossom house, the still rather Richardsonian dining room with its rounded bay being placed here at the centre of the rear. The staircase, still so prominent in the Shingle Style way at the Blossom house, is here pushed out of sight between walls.

[Illustration:

Figure 29. Frank Lloyd Wright: Chicago, Isidore Heller house, 1897, plan ]

Wright’s next important house, that of 1897 for Isidore Heller at 5132 Woodlawn Avenue on the south side of Chicago, perhaps shows some Japanese[345] influence in the succession of eaves-lines, one above the other. It is the development of the plan, however, that is most significant, as also the effect of the planning on the treatment of the exterior (Figure 29). The two principal living rooms are linked by a stair-hall into which they both open through wide apertures—no more mere doorways than in his own house of 1889, but tall breaks in the continuity of the walls. Although these rooms have ingle-nooks, they are not casual and cosy in the Shingle Style way but carefully ordered; both, indeed, are of regular cruciform shape. This shape, moreover, is given external expression in the plastic articulation of the external massing, an articulation that the multiple eaves echo above.

Two years later, in the Joseph W. Husser house, now destroyed, in Buena Park on the north side of Chicago, Wright’s personal development of domestic planning was carried much farther (Figure 30). Here the main living rooms were all raised to the first storey in order to have a good view of Lake Michigan, and the interior space was continued uninterrupted along the main axis of the house from the dining-room fireplace across the landing and through to the living-room fireplace. But the dining room was also articulated along a cross axis, extending outward into a large polygonal bay facing the lake, somewhat like the more Richardsonian bays of the Blossom and Winslow dining rooms.

[Illustration:

Figure 30. Frank Lloyd Wright: Chicago, J. W. Husser house, 1899, plan ]

Between the two houses just described, in which Wright’s planning developed so rapidly and so boldly towards unified but articulated space, came the River Forest Golf Club in River Forest, Ill. The front wing of this, built in 1898,[346] showed a comparable maturing of his vocabulary of wooden construction. The two Chicago houses were both of brick with rather lush Sullivanian terracotta decoration below the eaves not unlike that on the Schiller Building. At the Golf Club the characteristic feeling of the Shingle Style for rough natural wood surfaces was revived by Wright but made more architectonic in scale. Below continuous window bands protected by his characteristic hovering eaves, the lower walls and the terrace parapets were sheathed with boards and battens, not applied vertically as by Downing, but horizontally. Uncovered terrace, covered veranda, glazed foyer, all were closely related spatial areas, the last two unified by the continuous roof. The only solid element was the broad stone chimney marking the point where the main axis and the subsidiary axis of the low side-wings crossed. In 1901 the building was much enlarged by Wright, but quite in the original spirit (Plate 128B).

In 1900, the last year of the nineteenth century, with which this account of Wright’s beginnings may properly close, he built two houses side by side in Kankakee, Ill. He also designed for the _Ladies Home Journal_ ‘A Home in a Prairie Town’ which was published in February 1901. The larger of the two Kankakee houses, that for B. Harley Bradley at 701 South Harrison Avenue, is a large, loosely cruciform composition with low-pitched gables projecting in blunt points well beyond the ends of the wings. The smaller Hickox house, next door at 687 South Harrison Avenue, has a more advanced plan under similar roofs. Wood stripping suggests the stud structure underneath the stucco of the walls as do also, and rather more directly, the wooden window mullions (Plate 142A). The living room here, flanked by semi-octagonal music and dining rooms, extends across the ‘garden front’ and opens by french doors on to the uncovered terrace (Figure 31). Here the articulated but unified space of the Husser house was reduced in scale and simplified until it provided a quite new concept of domestic planning, later to be widely influential internationally (see Chapter 22). Towards that new concept much of the development of the Anglo-American house since as far back as the 1790s may seem—not too exaggeratedly—to have been tending.

[Illustration:

Figure 31. Frank Lloyd Wright: Kankakee, Ill., Warren Hickox house, 1900, plan ]

The _Ladies Home Journal_ project for a ‘House in a Prairie Town’, from which the term ‘Prairie Houses’ for Wright’s characteristic production of the next decade derives, is larger than the Hickox house, but the living area was intended to be very similarly unified and articulated. In one version Wright even proposed carrying this space up two storeys in the centre, somewhat like one of Shaw’s manorial halls. As on the River Forest Golf Club, the long lines of the low hip roofs shelter very long window-bands—out of Shaw, via Richardson, presumably. Although the _Ladies Home Journal_ house was intended to be stuccoed like the Kankakee houses, the window mullions echo the underlying wooden stud structure. As at the Golf Club, the chimneys would be the only really solid elements, passing up through the crossing volumes defined by the two levels of roof. The lower line of eaves extends, somewhat as on McKim, Mead & White’s McCormick house, over the _porte-cochère_ on one side and over the veranda on the other, a treatment Wright had already tried out somewhat clumsily on the Bradley house.

In considering the significance of these Wright houses of 1900 it must be recognized that even in America they were highly exceptional. Despite the fact that the ‘Prairie house’ project was published in a general magazine of national circulation, its immediate influence was very slight indeed. For all the vigour of the two great Chicago achievements of the nineties, Sullivan’s skyscrapers and Wright’s earliest houses, the main direction of American architecture in 1900 was quite different. So also in the England of these years, where Shaw’s house for Fred White and his Bryanston had introduced by the nineties almost the same sort of Academic Revival as had McKim, Mead & White’s Villard and Taylor houses, the work of Voysey, the English architect most comparable to Wright, was also almost as exceptional. The line of architectural development had already split as sharply as in America, with the difference that the longer-lived Shaw himself had taken the lead in the academic direction that Richardson’s pupils, McKim and White, took in America.

Although Charles Francis Annesley Voysey (1857-1941)[347] was ten years older than Wright, it is understandable with English conditions that his architectural career got under way little earlier. From 1874 to 1880 he worked as a pupil in the office of Seddon; from 1880 until he set up for himself in 1882 he was assistant to Devey.[348] In 1883 Voysey sold his first designs for wallpapers and printed fabrics, but for several more years he did little building. His first house, The Cottage at Bishop’s Itchington in Warwickshire, was built only in 1888; in the next two years various projects of his, increasingly original in character, were published in the _British Architect_; of these the one for a house[349] at Dovercourt of 1890 was the most advanced.

By the late eighties Nesfield and Godwin were both dead and leadership in English architecture, particularly as regards the domestic field, rested more firmly than ever in Shaw’s hands. The forces of innovation in English art were concentrated in the decorative field, thanks in part to Webb’s continuing activities with the Morris firm. But there is some question how well younger men like Voysey really knew Webb’s architectural work; almost none of it was published, and some of the best is hidden in remote parts of Scotland and the North of England. The work of A. H. Mackmurdo (1851-1942) was perhaps somewhat better known, but he was much more active with furniture, chintzes, and wallpapers than with building in the eighties. A project for a ‘House for an Artist’ that he published in his magazine _The Hobby Horse_ in 1888 was of considerable promise, however. In any case Voysey soon rivalled Mackmurdo as a designer of furniture, wallpapers, and chintzes, and quite outclassed him as an architect. Mackmurdo’s most significant influence was probably abroad (see Chapter 16).

The existence of an earlier project dated 1888 for Voysey’s house for J. W. Forster at Bedford Park has led to some confusion. The executed house dates from 1891. Sometimes known as the Grey House, it is very different indeed from its neighbours, by this time some fifteen or more years old, by Godwin, Shaw, and their pupils. For one thing, its walls are covered with roughcast, already used by Voysey on The Cottage at Bishop’s Itchington; for another, it is a three-storey rectangular box, severe and rather formal beneath its low hipped roof, not quaint and irregular like even the simplest of the earlier houses. The casement windows are arranged in bands between stone mullions, regularly but not symmetrically, and the eaves troughs are supported by delicately curved iron brackets. Otherwise there is no external detail.

The plan of the Forster house is also compact and regular, with entrance on the left side and living room across the front. In other words this house represents as much of a reaction against the picturesqueness of the earlier Queen Anne as does Shaw’s Fred White house, yet is quite without eighteenth-century reminiscence.[350]

More interesting and more prominent than the contemporary storey-and-a-half house known as The Studio at 17 St Dunstan’s Road in West Kensington are a pair of terrace-houses, also designed in 1891 but begun only the next year, at 14-16 Hans Road off the Brompton Road in London. Here Voysey dropped the roughcast he had originally proposed and used Webb-like red brickwork with the windows characteristically arranged in bands between plain stone mullions. The elegantly original detailing of the projecting stone porches and the curved line of the parapets at the top are related to his contemporary decorative work and in notable contrast to the almost ‘Monumental Queen Anne’ treatment of Mackmurdo’s slightly later house next door at No. 12.

A moderate-sized country house, Perrycroft, Colwall, near Malvern, begun in 1893, may be considered Voysey’s first mature production, introducing in executed work the personal mode of design for which the Ward project of 1890 had already shown the way, and from which he never moved very far in later years. This is comparable, not to Wright’s ‘first’ house in River Forest of the same date, but to his more advanced work of the end of the decade, the River Forest Golf Club and the Hickox house. Roughcast walls, windows arranged in bands between plain mullions,[351] a regular composition approaching but not quite reaching symmetry, these all follow from the Grey House and the Studio. But, being in the country, the house could spread out more. Moreover, the roofs were raised to a medieval pitch—45 degrees—so that their conspicuously heavy slating is as much a part of Voysey’s simple craftsman-like mode as are the off-white roughcast walls. The planning is closer to Webb’s than to Wright’s, the rooms being less symmetrically shaped and not opening at all into one another in the way of the Ward project.

A rather larger house, begun in 1896 on the Hog’s Back near Guildford in Surrey for the American Julian Sturgis, presumptive original of Santayana’s _Last Puritan_, has a somewhat less balanced composition with a prominent cross gable near one end (Plate 129A). The characteristic stone-mullioned lights of several of the rooms are here so extensive in their grouping as to constitute window-walls of the earlier Shavian sort.

In what is doubtless Voysey’s finest work, Broadleys on Lake Windermere, designed in 1898, the roofs are lower once more, and the window-walls are concentrated in three rounded bays along the lakeside terrace (Plate 129B). Here the hall in the middle is carried up two storeys, quite as Wright proposed to do in one version of his first _Ladies Home Journal_ house (Figure 32). In its horizontality, its concentration of fenestration, and its avoidance of medieval feeling, this house represents the extreme point of innovation and originality in Voysey’s work.

His own house, The Orchard, at Chorley Wood in Hertfordshire, was completed in 1900. Externally this resembles closely his earlier houses, but The Orchard has two cross gables and hence a stronger feeling of symmetry. Towards this the more regular and carefully balanced spacing of the window bands further conduces. In studying the vocabulary of this house, a vocabulary destined to be parodied _ad infinitum_ by architects and then by builders in the next twenty-five years, one can understand his feeling that he was a reformer not an innovator—the last disciple of Pugin, so to say, to whose secular work a line can be traced back via Webb, Street, and Butterfield. In Voysey’s special sense of continuity, which grew on him in later years, lies his great difference from Wright; for Wright was certainly determined, from the time he designed the Winslow house, to be as great an innovator—as much of an architectural creator—as was Sullivan in his skyscrapers. None the less, to look forward a little, such a house by Voysey as that now called Little Court at Pyrford Common in Surrey, built in 1902, is quite worthy of comparison with Wright’s masterpieces of that year (see Chapter 19). It shows little further development beyond his houses of the late nineties, however, except for a certain increase in horizontal emphasis.

[Illustration:

Figure 32. C. F. A. Voysey: Lake Windermere, Broadleys, 1898-9, plan ]

Just before and just after 1900, Voysey’s work was very much better known and more influential in England, and increasingly in other countries,[352] than was Wright’s either at home or abroad at that time. Moreover, many contemporaries in England were building rather similar houses. One of them, M. H. Baillie Scott (1865-1945), who also worked a good deal on the Continent, developed his planning much farther in the direction of Wright-like openness along the lines suggested by Voysey’s project of 1890 for the Ward house. The many houses, both executed and projected, that Baillie Scott published in _Houses and Gardens_ in 1906 made his planning known to the young architects of the Continent (Figure 33). Characteristic is his Blackwell house on Lake Windermere of about 1900 with an enormous two-storey living-hall elaborated spatially by various ingle-nooks and so forth. The plan was published by Muthesius in 1904, and may well have influenced Adolf Loos in Vienna and other Europeans even before his own book appeared (see Chapters 20 and 21). After 1906 Baillie Scott’s work became quite ‘traditional’, and it is hard to believe that the projects published in the later version of his _Houses and Gardens_ in 1933 are by the same man.

[Illustration:

Figure 33. M. H. Baillie Scott: Trevista, _c._ 1905, plan ]

The name of W. R. Lethaby (1857-1931), later the biographer of Webb and an influential writer on architecture, should also be at least mentioned here. When Lethaby left Shaw’s office, where he had been chief assistant, he began his career by building Avon Tyrrell in Hampshire in 1891, a large brick country house closer to Webb’s than to Shaw’s in character. But his main contribution was not in the field of domestic architecture.[353]

Already by the mid nineties, the most successful English house-builder, more than rivalling Voysey in the quantity and occasionally even in the quality of his domestic work, was Sir Edwin L. Lutyens (1869-1944). Beginning like Voysey in the late eighties by building cottages, his first house of real distinction was the one he built for his cousin and frequent collaborator, the garden-designer Gertrude Jekyll, at Munstead Wood near Godalming in 1896. Several other good houses followed shortly, including notably The Orchards, Godalming, in 1898; but this early period of his work really culminates in Deanery Gardens at Sonning in Berkshire of 1901 (Plate 182B). In these houses are preserved all the best of the Shavian Manorial—the great timber-framed bay-window of the two-storeyed hall at Deanery Gardens is exemplary—simplifying and regularizing that mode under the influence of Webb and even approaching Webb’s standards of craftsmanship in the execution.

Like Webb in his later work, Lutyens used almost from the first a good deal of stylistic detail in interiors; he also turned back towards the ‘traditional’ in his exteriors considerably earlier than Baillie Scott when designing such houses as Overstrand Hall in Norfolk and Tigbourne Court at Witley in Surrey, both built in 1899 two years before Deanery Gardens. Lutyens became from about 1906 the leading architect of his generation in England, and his later work will be treated elsewhere (see

## Chapter 24). His increasing material success after the opening years of

the century, rivalling Shaw’s in the previous generation, is to a certain extent the measure, though not the cause, of Voysey’s decline in popularity.

C. R. Ashbee (1863-1942) and George Walton (1867-1933)[354] were other domestic architects active in the nineties and the early years of the new century. The latter belongs to the Glasgow School, of which Mackintosh was the principal figure, and like Mackintosh he was more decorator than architect (see Chapter 17). One house in England, The Leys at Elstree of 1901, may be mentioned here. The interiors are fine examples of the Arts and Crafts mode, as it is sometimes called, more stylized than Voysey’s but less original than Mackintosh’s. The plan is organized symmetrically around a large two-storey hall rivalling Baillie Scott’s of the period in its complex spatial development.

Ashbee was one of the first Europeans to appreciate the significance of Wright, and was appropriately chosen by Wasmuth to write the introduction to his second publication of Wright’s work in 1911 (see

## Chapter 19). Three houses by Ashbee side by side in Cheyne Walk in

London, No. 37 of 1894 and Nos 38-39 of 1904, represent the chronological span of his significant architectural production and illustrate clearly his characteristic progress from the Shavian to an originality at least comparable to Voysey’s. Closely associated with the Arts and Crafts Exhibition Society, Ashbee was like most of these men except Voysey[355] and Lutyens generally more active in the field of decorative art than in building. Right through this period English decorative art exercised a major influence on the Continent (see Chapters 16 and 17). So close is Mackintosh’s tie with the Continent that his schools and even his houses are better discussed in relation to the Art Nouveau.

Of all these English architects who have just been mentioned, Voysey was the most creative in the field of domestic architecture and, except for Lutyens, the most productive down at least through the early years of the twentieth century; after 1910 he built almost nothing at all. Yet Voysey did not die until 1941, by which time a younger generation, to his confusion, had accepted him as a father of a modern architecture that he disapproved as strongly as did Lutyens. In 1940 he returned almost from the grave to receive the Gold Medal of the Royal Institute of British Architects.

From the Picturesque cottages of the opening decades of the nineteenth century to the early masterpieces of Wright and Voysey around 1900 is a far cry, further perhaps in the drastic revision that it represented of so old-established a building type as the dwelling-house than from Parris’s Market Street buildings in Boston of 1824 to Sullivan’s Carson, Pirie & Scott Store in Chicago as completed eighty years later in 1904. Yet in Anglo-American domestic architecture, quite as was the case with commercial architecture, real achievement recurred all through the century.

------------------------------------------------------------------------