Skip to main content

Transplanted: a floral tapestry-woven table carpet once at Knole, Kent

H. L Turner

Transplanted: a floral tapestry-woven table carpet once at Knole, Kent

H.L Turner


Relatively few tapestry-woven table carpets, once common objects in even middle-class western European households, survive. The example purchased in 1911 by J. P.Morgan from Knole, England, since lost sight of, was recently recognized hanging now at Filoli, Woodside, California. It pictures fifty-six different flowers influenced by the long series of plant books printed in Antwerp in the later part of the sixteenth century, contained within a border which combines stiffly geometric scrollwork with formal fruit and flower compositions. A central roundel, one of three, probably displayed the first owner’s arms, later excised and replaced by an unidentifiable coat. The tapestry has no immediate parallels, and its reappearance emphasizes the need for further investigation of this long-lived genre. The chronologically contradictory combination of design elements, some familiar from the 1530s others only from around 1600, makes it challenging to establish the tapestry’s date, its weavers or its place of origin.


Wall hangings of woven tapestry formed an important, even essential, decorative element in the furnishing of the houses of the wealthy. They provided colour, projected an image of power, authority and splendour and declared the owner’s ability to indulge in conspicuous consumption. They would have been in evidence in the late fifteenth-century residence at Knole belonging to the archbishops of Canterbury. Tapestries remained a significant embellishment for a century after Thomas Sackville enlarged the property between 1600 and 1610. Knole’s collection was at its largest – some 317 pieces - by the end of the seventeenth century, built up by direct purchase, by acquisitions from the royal collections, particularly those at Whitehall and by the absorption of tapestries belonging to families into which the Sackvilles married, in particular the Cranfields. On display, therefore, was a variety of styles from differing dates and a wealth of subjects, religious, mythological and classical. Soon thereafter decorative fashions changed, paint, gilding and, later, wallpaper replacing tapestry; by 1799 only seventeen rooms were still tapestry hung. Nevertheless, the pieces no longer in use were carefully stored by successive generations of the family. Amongst them was an unusual and remarkably intimate survival, a tapestry-woven table carpet (Fig.1). Eventually sold, along with others, it was shipped to America and its whereabouts lost.1 It recently came to light, in an American National Trust property at Filoli, Woodside, California.

The whole tapestry

Fig.1. The table carpet hanging at Filoli, Woodside, California; woven in wool;
17 feet 5 ½ inches x 5ft 10 inches (178 x 530 cms), warp count 14 per inch 6/7 per cm. © The Trustees of Filoli

The tapestry’s purpose is revealed by its borders in which the elements are vertical only when the border hangs over a table edge. Although there were once many such items, listed in most inventories of England’s larger houses, their survival is rare. This one, woven with flowers in both field and border is a complex example of a once common furnishing item.It executes in wool what so many Dutch artists would paint in oils. Although its position in the house is unknown, it was probably spread somewhere in the state rooms on the upper floors, covering a table very similar to the Long Table now in the entrance hall. The very rarity of survivals in the genre, however, makes it hard either to date or to set the carpet in the context of its stylistic development.

The tapestry
The tapestry measures approximately 5ft 10 inches x 17ft 2½  inches (1.78 x 5.30 m.). On a green ground, the field pictures 112 flowers, representing 36 separate plants; into this are set three roundels, each surrounded by a plain circle, broken by a lion head mask at the cardinal points. Two are themselves decorated with a formal arrangement of fruit, foliage and flowerheads arranged round a candle-like spike (Fig.2); the central roundel encloses a crest and coat of arms, the only replacement weaving in an otherwise well-preserved tapestry. The border, clearly separated from the field, is in another idiom.

Right Hand Roundel

Fig. 2. The right-hand roundel of the table carpet, © The Trustees of Filoli

The number of flower species illustrated is, in tapestry terms, out of the ordinary. As a planting plan, they would provide colour through much of the year. Spring flowers include snowdrop, daffodil, narcissus and pheasant’s eye (Fig.3); moving into summer, one finds cowslip, periwinkle, tulip, peony, lily of the valley, wood anemones, primrose, violets and the flowers on the trees which would later fruit – apple, grape, medlar, orange, plum, quince. Deeper into summer, blooms include amaradulcis, bleeding hearts, cornflower, borage, Canterbury bells, columbine, daisies of two kinds, iris, honeysuckle, foxglove, pansy, love in the mist, marigold, scarlet pimpernel, lungwort, pinks and carnations, roses, summer snowflake and poppies (Fig. 4). August might see the earliest apples, plums, pears, musk melons followed by grapes, quince and medlars. Autumn shows a marked diminution in species, compensated for by the fruit ripening amongst the changing colours of the tree leaves.


Fig. 3. Detail, Narcissus. © The Trustees of Filoli

Flowers deeper into summer

Fig 4. Detail, Periwinkle and Carnation.  © The Trustees of Filoli

All the species shown would grow, or could be cultivated, in western European gardens, whether urban or rural. They closely resemble the selection which Francis Bacon thought, at much the same time as the tapestry was woven, suitable for a princely garden.2 Nevertheless it is a curious garden, hard to translate, even mentally, into plants in soil, on trellis, or confined by the clipped box hedges of the fashionable knot gardens. The plants are an interesting mixture of the beneficent and the noxious, some serving purposes now forgotten or ignored. Roughly half are poisonous or contain at least some poisonous substance, some of medical value if carefully administered. Others are medicinally beneficial, without reservation. Many were scented. Some were species which could be trained up ‘carpentry’ to make pleached alleys or walks – roses, honeysuckle, periwinkle. Others are, for us at least, plants which grow at least as easily in the wild. Few are new, or exotic, introductions from the East which would excite the botanical connoisseur or impress the viewer, though one might note the lily; none comes from the Americas, still being explored. The overall composition does not appear to be endowed with sentimental overtones, nor burdened with religious significance or emblematic connotations. There is an idyllic absence of pests. Defying long-established tapestry tradition no caterpillar or snail crawls on the leaves or creeps up the stems. Neither do any butterflies, bees or moths flutter above the open blooms. Except for strawberry, bilberry and the garden pea, produce with culinary use is firmly separated from the flowers, some in the roundels, most in the borders. Everything shown had long been familiar in Holland and Germany; much also grew in English gardens (Appendix 1). So did most of the fruit seen in the tapestry - apple, fig, quince, gourds, bilberry, orange and pomegranate. Only the two last might challenge all but the most enthusiastic gardener.

The field is clearly bounded by tripartite border, some fifteen inches (0.38 m) deep, faded on its left-hand side. On the inner side, some three inches (0.7 m) deep, a running design is made up of alternating quatrefoil and trefoil, possibly mandrake and melilot respectively, and an elongated leaf of imaginary origin. The outer border, of similar depth, mimics a picket fence, flower heads and trefoil garnished. Tiny flower-heads are repeatedly laced between the short stakes like flowers on a trellis or border edging.

The central section, the deepest, was designed, if not always executed, with rigid geometric precision. Its regularity confers a unity, and provides the same stability as a garden fence or trellis bounding the plants growing within its confines. On a red, almost purple ground, it shows alternating baskets and tazza, designed so that they were vertical only when the border hung over a table edge (Fig. 5), unlike a hanging where 

To hang over the table edge

Fig. 5. The border elements. © The Trustees of Filoli

elements were vertical in each border. They are linked by clover-leaf strapwork; at each centre is a flower head, paired as opposites, rue and borage. Produce is artistically, as well as practically, piled high on each container, interspersed with flower heads. Occasionally the gardener’s enemy, a marauding pigeon, stands poised besides the heaps, a beady eye fixed on a particularly tempting morsel. At the centre of each border a vase of flowers contains a much looser, almost floppy, arrangement, the freer style seen in later sixteenth-century print creations, for example those of the Wierix brothers.3 When the carpet was spread on the table the containers might well, from a distance, appear to be standing on the table top. That impression is strengthened because each corner is decorated with a large rosette and fruit at each quadrant so that the corner fold would lead the eye around the angle between long and short sides (Fig. 6).

Rosette and fruit corner decoration

Fig. 6. Lower left corner and part of the field. © The Trustees of Filoli

This tapestry is no off the peg purchase and is not an item casually produced. Every detail was carefully planned and drawn before being fitted into a scheme itself designed to avoid repetition. No two examples of the same plant were exactly the same. Whenever a species was repeated its structure, scale and size were varied, as if drawn from different plants. It had reached a different height and had more, or fewer, buds or blossom; its leaves might blend differing hues of green combined in different patterns. The plants shared only one thing. They were never shown, as a tapestry designer sometimes did, with roots as they were executed by artists who supplied drawings for publication in one of the many plant books which rolled off the presses in Antwerp in the second half of the sixteenth century on which some of the tapestry examples were modelled.

Like the later flower paintings, largely produced in Dutch studios, this tapestry carpet was a purchase of considerable value. Its cost is implied by the complexity of the pattern which would have been reflected in the size of the designer’s fee, apparently in preference to choice of more precious materials, for the tapestry is almost entirely woven in wool, rather tight and hard-packed, with a count of 14 warps per inch (6/7 per cm). The silken ladder of threads which secured the warp threads - the galloon - survives on both sides; it is a pattern found in the early seventeenth century in expensive tapestries affordable only by the rich. The only other silk used, of poor quality and now faded, is in the later insertion of the coat of arms. Neither does metal thread brighten the centre of any flower, whether stigma or stamen, or sharpen the veins of a leaf.

Nevertheless, despite the careful preparations of the designer there are some curious features. Although the tapestry was designed to lie flat, the flowers were taken from a source in which the flower was presented to be seen vertically and, apparently, in isolation, even though some branch and twist round, sometimes even into, the adjacent plant. Only rather later than the probable date of the tapestry would it become commonplace for flowers to appear to be strewn across the table top, each lying in isolation from its neighbour.4 Nor, on close inspection, has the design been perfectly executed. The three roundels are slightly uneven in size (25 x 23, 25 x 24, 26 x 23 inches, 0.63 x 0.58, 0.63 x 0.61, 0.66 x 0.58 m., left to right) and are not placed absolutely equidistant from each other or from the borders. Though it might have seemed more appropriate to enclose the contents with a wreath of laurel, the designer chose to use the harsh lines of a lightly embellished oval, broken by lion heads. In the border, the baskets and tazza, designed to sit exactly above each other, have slightly slipped out of true, not of course noticeable when the carpet covered a table and the eye could see only one side.

There is also some evidence of a change of plan after the weaving had begun. On the lefthand side the size of the plants is very much smaller than the same plant’s growth on the right – on the left, therefore, there are many more species in the same space. The birds too, curiously unidentifiable, are strangely positioned. In the left quarter they stand in a straight line. In the other quarters they are positioned in triangles and one of the three is shown coming to rest on a plant, but at right angles to its proposed perch (Figs 7, 8). When the tapestry was on the loom, of course, the birds would have seemed to the weavers’ eye to be correctly placed, for a tapestry pattern is woven sideways between the vertical warp threads. Once cut from the loom, a tapestry is turned through ninety degrees and hangs so that the warp threads run parallel to the floor and the pattern occupies the vertical position.

Birds on the left hand side

Fig. 7. Flower density on the left hand side of the tapestry. © The Trustees of Filoli

Birds on the right hand side
Fig. 8 Flower density on the right hand side of the tapestry.  © The Trustees of Filoli

The tapestry’s history

The tapestry is one of twenty-nine pieces bought by the financier J. P. Morgan in 1911 from Knole, Sevenoaks, Kent, a house in the ownership of the same family, the Sackvilles, for more than four hundred years.5 The story of its almost casual acquisition from an English family in need of money, like so many others, as the result of the depression of the last years of the nineteenth century, is told briefly in Lady Sackville’s diary.6 All left England in order to avoid recently imposed English death duties; eighteen were exhibited in Paris in 1912-13 in a charity exhibition to swell the funds of the Friends of the Louvre before being shipped to the United States.7 Briefly housed in the Metropolitan Museum of Art, New York, this tapestry was withdrawn, with twenty-eight others, in May 1915; most were bought by French and Co, but this one has left no trace in the company’s records.8 By early 1918 it was photographed at Filoli, Woodside, California.

The house it came from
Knole, Kent, built first in the later fifteenth century by the Archbishop of Canterbury, William Bourchier, and was extended by Thomas Sackville who, in 1604, began the building programme which gave the house its present external appearance.9 His descendents still live in part of the much extended, and several times refurbished, house, now gifted to the English National Trust. The Sackville family, earls of Dorset, played a prominent part in English government from the early seventeenth until the mid-nineteenth century. Their collection of tapestries, acquired piecemeal and as personal taste rather than as a coherent programme, was outstanding.

One might therefore have hoped that, together with the coat of arms, tracing the origins and acquisition of this tapestry would present no special problems; the records of a family preserved through the generations in a continuously occupied house would surely make some mention. However, there are no purchase records and inventories are notoriously laconic. A series of six survives, dating from the forfeitures of the Civil War (1646) to the later decades of the nineteenth century. Though the inventories reveal the extent of the collection, they are not detailed enough to do more than indicate the themes which decorated the main rooms. Identified tapestries were largely figurative, showing scenes from the Bible or from Classical mythology; many more pieces were simply designated as ‘hangings’. It is not therefore easy to visualize how this piece fitted in; there is no clue to its use at Knole. A single reference, too imprecise to be regarded as a definitive reference to this tapestry only, occurs in 1690 mentioning “a fine tapestry carpet with a coat of arms”, then stored in a Wardrobe. The last recorded tapestry acquisitions are the pieces ‘de-listed’ from the inventories of the royal palace of Whitehall in 1694 and those from Copt Hall, Essex, a family property sold in 1702 whose contents were transferred to Knole.10 It seems probable therefore that the carpet had come to Knole by the later seventeenth century. After that, tapestry was going out of fashion – as the two later inventories, of 1765 and 1799, show.11

The other seemingly obvious clue to the carpet’s origins should lie in the coat of arms. But the coat of arms is not contemporary with the original weaving; it has been replaced at a later date (Fig 9). It is unlikely that the

Replacement central section
Fig. 9. Middle section of the tapestry, showing the coat of arms. © The Trustees of Filoli

replacement, much faded, copied the original coat; it is more probable that the tapestry passed to a new owner. The arms have defied even expert attempts at identification, but because the older shield was literally snipped out of the remaining original weave, arms did make up the roundel’s original content.12 The crest, a demi stag Argent attired and unguled Or, the closed visor and the red and white mantling, terminating in gold tassels are all typical of seventeenth century English heraldic practice, suggesting a gentleman, rather than a noble, owner. No link between the crest, helm and mantling, let alone the arms, with the marriage alliances of any member of the Sackville, Germain or, later, the Sackville-West families is traceable.

Identifying the origins and date of the tapestry
Not only are the obvious sources for identification lacking, but, although tapestry-woven table carpets have a long history, they are a little studied genre. Like most, this example lacks a town mark or a weaver’s signature, an immediate handicap to any attempt to name either a workshop or a weaving centre. Though English inventories show that tapestry table carpets were a common household item, long in demand, survivals are few and there is no immediately obvious parallel. Stylistically, this carpet’s distant antecedents lie in the regimented arrangement of fifteenth century millefleurs tapestries, where whole sections of pattern might be repeated across the width of the hanging.13 The stiff blooms of this fashion later became slightly less rigid and appear in more relaxed placing as, for example, in the tapestry dated around 1550 bearing the arms of Paolo Giovio in the Victoria and Albert Museum, London.14 The form, if not the presentation, remained popular, though it mutated, ending finally with the more florid, but once again more rigid, geometrically-disposed compositions of mid- or late-seventeenth century Dutch table carpets, for example those in the Rijksmuseum, Amsterdam.15 The plants in the Filoli carpet, however, are on a larger scale in a freer arrangement, suggesting a date between two distinct styles. The chronological gap in design of surviving carpets might perhaps be filled by examination of tapestry hangings for which designers drew on the mid-sixteenth century portrayal of individual flower species, more widely spaced and sometimes shown as a plant complete with roots, as for example in a section displayed at Hamburg and another in a private collection.16 The earliest printed works to illustrate plants and largely replacing the older herbals were those compiled by Fuchs (1542) and Mattioli (1544).17 Their continuum lies in the plant books of the 1570s-80’s produced by the Plantin press at Antwerp, for example of Rembert Dodeons (1563, 1583), of the Frenchman, Matthieu de L’Obel (1576) and the drawings of Carolus Clusius. In the interests of economy, the same woodblocks were used in successive publications, sometimes even used to illustrate two separate species. It is not therefore possible to trace the source of each tapestry flower to a particular publication, yet the depiction of the tapestry’s flowers, in vertical arrangement, reflected the shapes imposed on plants by the requirements of wood-blocks illustrating those books. The tapestry’s style, however, marked a big change from the single specimens placed in the foreground of large numbers of tapestries woven in Brussels in the first two decades of the sixteenth century and, in some ways, diminished the botanical lexicon seen there.18

An apparent parallel to the style of the Filoli carpet, of similar dimensions, might lie in the floral tapestry at Sudeley Castle, Gloucestershire since 1845 (Fig10).

Floral Tapestry at Sudeley Castle

Fig.10. The hanging  at Sudeley Castle, Winchcombe, Gloucestershire, England, woven in wool, silk and some metal thread; 15ft 71/2 inches x 6 feet 1 inches (1855  x 4761/2 cms); warp count 18/19 per inch, 7/8 per cm. © Sudeley Castle.

Its borders, however, different at top and bottom make visual sense only when the piece hangs vertically, suggesting it is probably not a carpet but a wall hanging. It is set with roundels, three on either side and one below the centre depicting the Virtues. An eighth roundel, centrally positioned and the largest, no longer has its original content, almost certainly the coat of arms of the couple whose unidentifiable initials remain – B/I*F. The flowers, eminently identifiable, are arranged, of necessity because of the other design elements, in sinuous rather than rigidly vertical columns; few are repeated and there are fewer species than in the Filoli carpet. The large blossoms are worked in great detail, rarely having fewer than four colours; blossom on the same stem is often rendered in varying proportions of the same colours. Stalks and leaves are depicted in more automatic fashion, the leaves especially rarely more detailed than half yellow, half green with the vein clearly delineated. Where reduction in size is demanded, for example in the upper register between roundel and border, a smaller flower may be chosen, but the attention to detail remains the same. Unlike Filoli’s carpet the tapestry is woven with the occasional use of more costly, metal thread emphasizing the centres of flowers. Amongst them are butterflies, caterpillars, dragonflies, snails, twenty-six species of recognizable birds, a squirrel, an evil-faced cat and even a lion and a unicorn (Figs. 11,12). Its colouring closely resembles Filoli’s; so too does the pattern of the inner border, but this is a common design in the south Netherlands work and not a sufficiently distinctive indicator of place.

Fig 11. Detail, unicorn and flowers, from the hanging  at Sudeley Castle, Winchcombe, Gloucestershire, England. © Sudeley Castle.
Fig 12. Detail, peacock and flowers, from the hanging  at Sudeley Castle, Winchcombe, Gloucestershire, England. © Sudeley Castle.

The origin of the Sudeley hanging is unknown. Popular descriptions usually claim it as a product of the weaving venture outlined in the will of the English gentleman William Sheldon (1570), who allowed use of his manor house at Barcheston, Warwickshire, for the weaving of tapestry, arras and cloth fabrics.19 However, the author of the attempt made in 1928 to identify tapestry-woven products originating there expressed considerable doubt that the Sudeley tapestry should rank amongst those products and clearly excluded it. Though regarded by its author as tentative, his finer distinctions have been ignored and his endeavours interpreted as definitive. His work has recently been questioned on the grounds that the tapestries used as the starting point for the definition of stylistic criteria could not then, and cannot now, be certainly associated with looms at Barcheston. Moreover, no tapestry called Sheldon is mentioned in any document testifying to its origin, nor is a sixteenth-century provenance known for most pieces so called. Nowhere is there any unambiguous indication that Sheldon’s perfectly reasonable plan was attended by the success with which it has been credited. Unparalleled amongst the tapestries called Sheldon and never definitively admitted to that corpus, the Sudeley carpet might possibly be the work of the hitherto unknown émigré workshops in London around 1600, but is rather more likely to be a product of Antwerp, of southern Holland or further north still.

Chronologically, a closer, but far from exact parallel to the flowers in the roundels on the Filoli carpet might be sought in a tapestry showing the Five Senses, most recently displayed at Denver Art Museum.20 Three figures representing Hearing are placed within a laurel-wreathed roundel at the centre, the other four senses, each placed against a landscape background, each encircled, occupy the corners. Branching flowers, fruit, birds and monkeys are set into a black ground. There any resemblance ends, for the border is complex, woven with amorini amongst swags of fruit, foliage, lion heads and classical medallions on a yellow ground. Measuring 7 ft x 5 ft 3 inches, (2.13 x 1.60 m) it was sold several times in the 1920s always with a different attribution. The most probable was advanced by the tapestry historian Heinrich Goebel who suggested it was of Wismar origin with an early seventeenth-century date.21

By the time a dated, if not marked, example survives, style has changed. A Dutch carpet with the woven date, 1618, now at Kronborg Castle, Denmark, shows flowers and fruit arranged at right angles to each other so that however the tapestry was viewed, some at least of the plants would appear as they do in nature.22 Their layout, rigidly spaced, succeeds in leaving the impression that the flowers were strewn, literally, across the table. Its border, mimicking a scalloped edge, is astonishingly simple.Much closer in conception to the later Dutch carpets and with only a tenuous resemblance to the Filoli carpet’s scale and selection of plants it should, for the purposes of dating the latter, be regarded as a terminus post quem.

Just as it is hard to attribute the Filoli carpet to a specific weaving centre, it is also difficult to date, not only because of the paucity of surviving examples and the absence of a near parallel, but also because the Filoli example contains design elements familiar in tapestries dateable from the 1550s to the 1620s. Medallions are common, seen not only in Paolo Giovio’s tapestry but in the slightly later table carpets commissioned by the English Lewkenor and Luttrell families of the early 1560s. In a novel, and costly, variation on the more common vellum-inscribed genealogy, the Lewkenors chose to display their lineage through their coats of arms set amidst swirling foliage; the Luttrells placed theirs against a geometrical pattern in yellow, blue and green very similar to designs found on painted walls, but also to the popular geometrical Turkish carpets.23 Coats of arms either within simple circles or enclosed by wreaths of flowers remained a favourite motif with north German families into the second decade of the seventeenth century.24  The lobed frame from which the border scrollwork is inspired emerges only around 1600, seen in sets of cushions whose exact place of origin is unknown, but likely to be Flanders (Fig.13).25

Meeting of Isaac and Rebecca

Fig. 13. The meeting of Isaac and Rebekah from Scenes of the Lives of Abraham and Isaac, ca. 1600. Wool, silk, with silver-gilt thread (21 warps per inch, 9 per cm.) H 19¾ x W 20 inches (50.2 x 50.8 cm.), The Metropolitan Museum of Art, New York, Gift of George Blumenthal, 1941 (41.100.57a), Image © The Metropolitan Museum of Art.

The most specific stylistic parallel, the lion heads, are seen in various positions and in a range of smaller pieces. One, showing Jacob at the well, contains both lion heads on a lobed frame and the tazzas of Filoli’s border (fig 14). The set most

Jacob at the Well

Fig 14.  Jacob at the Well, H 22 x W 38 inches (0.56 m x 0.96 m);  22 warps per inch (9 per cm);  wool, silk and silver-gilt thread;  Flanders c.1600.  Museum number T.184-1925.  ©Victoria and Albert Museum, London

probably dates to around 1600 and should probably be interpreted as Flemish work, while Filoli’s trefoil-decorated inner border bounding the field can be paralleled countless times in work from the south Netherlands in the later years of the century. 26  As we have seen, the origins of the design in the field lie with Brussels weavers of the 1560s-80s, the formal arrangements in the borders from Netherlandish prints subsequently adapted for weaving, yet the softer, strongly contrasting style used to show plants within the roundels points to rather later taste and date, seen in the little documented small pieces emanating from workshops in north Holland or north Germany. Far more work on products which might have been woven in such workshops is needed before a definitive conclusion can be reached. But for the present, and by elimination rather than from positive evidence, it seems likely that since the design elements suggest a date around 1600-1620, the tapestry might well have originated in a workshop perhaps in Rostock, Lubeck or Wismar, set up by émigré weavers driven from their homelands in the more traditional tapestry weaving areas.27

Might this carpet be considered as an example of work from the ateliers Dr Brosens recently suggested should be looked for in Flemish territories, the traditional heartland of tapestry weaving before the dispersal of much of the workforce through a combination of factors – wars and straitened circumstances at home and growing demand for tapestry which made establishment of ateliers elsewhere an attractive commercial proposition? 28 As we have seen, the Filoli carpet is related only in subject matter, not in style, to the literally later flowering examples regarded as products of Dutch workshops through a combination of inventory evidence, paintings and parallels in damask of Dutch origin.29 The suggestion that such workshops did not create or monopolize production, but might have evolved patterns begun earlier in Flemish ateliers has been put forward on the basis also of inventory evidence. Though neither the evidence published for Delft and Gouda in the late sixteenth and early seventeenth centuries, nor the longer inventory series from Antwerp suggest intensive activity, this may be misleading.30 Earlier English inventory evidence is copious and obligingly detailed for the first half of the sixteenth century, but lapses into reticence – where only two more words would have revealed so much – in the later years.

This earlier evidence, however, indicates plainly that table carpets with foliage or floral content were a common household item. Moreover, the material was often specified – silk, crewel, linen, Turkey-make and, admittedly more rarely, of tapestry. There are also glimpses of design. Thus entries for the three thousand or so tapestry items inventoried at the death of king Henry VIII (1547), usually by men familiar with the contents of the Wardrobes, are sometimes very explicit.31 Several of his sixty tapestry carpets, out of a total of perhaps two hundred and fifty in a variety of fabrics, were described in detail. One was said to be “of tapestry wrought with the king’s armes and with roses and flower deluces conteyning in length iij yards di and in breadth i yarde di skante”; another was “wrought with blewe knottes roses and flowersdeluces conteyninge in length one yarde quarter skante”. Thirteen others were verdours of brode blomes’, each carpet painstakingly measured. A fourth was described as being“of verdowres with iij roundes of Imagerie in it oon of them is a maundye conteyning in length iiij yards iij quarters and in breadth one yerde three quarters” and another as being “ of fine tapistree ymagerie lined throughout with canvas conteyninge elles flemyshe xxj quarter di”. They were not only straight lengths; there were variations, having “vallaunces aswell on either syde as at both endes with a ladie in the myddes of the said carpett and a unycorne by her side the said vallaunces on the side conteyning in length the pece three yards di and in depth iij quarters di the said vallaunce at the endes conteyning in length iij quarters di and in depth three quarters”, a consciousness of the need for the valances, like the later borders, to look well when hanging over the table edge.

King Henry’s subjects might emulate, but could not outdo, him. In 1523, Dame Alice Hungerford had in ‘the wardrobe chamber in a presse, vj fine carpettes for cobburds, iij gret kerpettes for tables, ij of them of fine arras and the other of verder, vij bastard carpettes for cubbords and tables’.32 The inventory of Sir Henry Guildeford, ‘late Comptroller of the king’s most honourable household’ (1532), noted fourteen carpets in London and, at Leeds Castle, Kent, ‘a carpet of verdour in the closet’ – though, like many of his countrymen, he had a predilection for the geometric Turkey carpet, one of which was specifically described as being for a table. He also had ‘a coberd cloth of cours flowers.’33 The most specific description is an entry amongst the goods of Thomas, late Duke of Norfolk, executed in December 1546. His carpets included ‘a verdure with buildings, beasts and birds and three garlands with skutcheons 5 yards in length, 1 yard q.di in depth’.Slightly later, in 1561 William Herbert, earl of Pembroke, owned amongst several others ‘a longe carpet of broad leaves of arras making and ……… a long carpet with small flowers’.35

None of the inventories provides any clue to places of manufacture. Nevertheless, it is abundantly clear that the floral tapestry table carpet had an existence long before the later seventeenth century. It is scarcely a fashion or a taste which would die out and it seems less and less likely that the later examples are, as so often argued, a result only of the increasing interest in botany and the mass printings of the plant books in Antwerp or of influence from the more delicate florilegia. There is, of course, no evidence for a weaving industry in England to explain the presence of tapestry carpets in the country and so all the examples described above must be considered to be imports, at a time before any of the towns in Holland which produced the Dutch carpets had an industry, which itself grew out of the dispersal of skilled weavers. Increasing, but still sketchy, evidence for a tapestry industry in England begins only in the later part of the sixteenth century. Most clearly centred in London, where men from towns as far apart as Emden and Enghien settled, it may possibly also have been practised, at least on an occasional basis, in the towns which had invited Flemish émigré weavers to settle and, in teaching their skills including arras and tapestry weaving, reinvigorate declining urban centres. Amongst these were Sandwich, Maidstone, Colchester, Halstead and Norwich, together with the enterprises of individuals, that of William Cecil at Stamford, Lincolnshire and William Sheldon at Barcheston.36 Nevertheless, the industry remains émigré-linked and with a stronger existence on paper than in products definitely attributed. Moreover, it was always in competition with imports.


Though found in an English household, an English origin for the Filoli carpet is unlikely. It seems doubtful that it could possibly be an example of a Dutch carpet and more probable that it was woven in a workshop in north Germany. More conventionally, but without any great claim to plausibility, its genesis might be sought further south, perhaps in Antwerp. An essential element in decorative schemes from the fourteenth century onwards, tapestry was most often used, particularly in large residences, as wall hangings designed to impress. Nevertheless it was a luxury item and was therefore always sought after also by the prospering merchant and lesser landholders. English sixteenth-century probate inventories certainly reveal the popularity of table carpets. One also illustrated their fate, noting ‘In the chamber where his [my Master’s] man used to lay, one covering of tapistrey which was a carpet for the banqueting board’.37

Though the stiff presentation and the sometimes awkward arrangement of both plants and birds in the Filoli carpet may strike the modern eye as un-natural, and are certainly unrecognized in other tapestries, the stylistic idioms expressed in the tapestry would have been familiar to the man who commissioned it. The house in which the tapestry now hangs was built between 1914 and 1917 by William Bowers Bourn II (1857-1936). Bourn’s parents had settled in California in1850; he inherited the most productive of many gold mines in the state, the Empire Mine in Grass Valley, and later purchased the Spring Valley Water Company which supplied water to the developing city of San Francisco, some thirty miles to the north.38 While the building, Stuart-Georgian in appearance, rose from the ground Bourn’s wife, Agnes, embarked on her plans for the interior decoration and also for the gardens. From the beginning the Bourns’ intention was that the gardens should be as important as the house and their layout was to complement its plain lines and set off its position. The theme of gardens was also to be carried inside.

The wide interior space of Filoli’s reception room and its walls, seventeen feet high at least on the ground floor, lent themselves to large scale decoration. The decision to use tapestries for the hall and for the stairway was influenced partly by current taste and fashion, and perhaps also by the recent displays of the Pan-Pacific Exhibition. Open to the public in San Francisco from February to December 1915 its succession of pavilions displayed a wide range of furnishings and fittings from seventeenth and eighteenth century Europe, on loan from European institutions.39 Further interest had perhaps been stimulated by the display of J. P. Morgan’s purchases in the Metropolitan Museum which Bourn might have visited on trips to New York.

Four tapestries were chosen; this flower-filled piece opposite which hung another, equally large, showing, under a wooden pergola, a formal sixteenth century garden sheltering wild animals and traversed by a hunt. Just visible from the reception area, two seventeenth century pieces, acquired later, hung above the staircase. Woven at Mortlake, London, they showed idyllic pastoral pursuits relevant to May and July, adapted from designs by Jerome de Potter. They are now in the San Francisco Museum of Fine Arts.40

The tapestry’s unknown first owner may have wished simply to beautify his hall or gallery or even his dining chamber, though in that case the tapestry would almost certainly have been protected during a meal by a linen cloth over its surface. He may also have been amongst those with an interest not only in plants but also in gardening.41 Just as he bought a piece of Nature and had tried to bring the outside world into his house, so too did Mr and Mrs William Bowers Bourn, bridging the centuries.42 The beauty of the tapestry is complemented now by the planting of the present gardens at Filoli. Those who see both surely share the sentiments expressed in 1618 by the Protestant clergyman-gardener William Lawson about the pleasure the arrangement of a garden could give.43 Such sentiments are true also of the skill of the tapestry’s designer:

All these by the skill of our gardener so comely and orderly placed in your borders and squares and so intermingled that none looking thereon cannot but wonder to see what nature corrected by art can do’.


My thanks go first to the conservator, Stan Derelian, who contacted the Victoria and Albert Museum; to Miss Wendy Hefford there who suggested I investigate further; to Assistant Professor Mary Barkworth and Dr Michael Piep of the Intermountain Herbarium, Utah State University, for their identification of the flowers, and to the Trustees and the Curator, Tom Rogers, at Filoli for support and interest during the time it has taken to decipher the puzzles.


1. Briefly published in S. de Ricci, Catalogue of twenty Renaissance tapestries from the J. Pierpont Morgan Collection, (Paris: privately printed, 1913), p. 28, no. XI. A survey of Knole’s tapestries is in preparation by H. Wyld which will include all the tapestries once in the house. Filoli has been  a property of the American National Trust for Historic Preservation since 1974.

2. ‘Of Gardens’, in The collected works of Francis Bacon, ed. J. Spedding, R. L. Ellis and D. D. Heath (London 1876, reprinted London: Routledge/Thoemmas Press, 1986), vol VI, pt ii, pp. 485-492.

3. M. Mauquoy-Hendrickx, Les Estampes de Wierix, (Brussels: Bibliothèque Royale Albert Ier, 1978-84), four vols, I, plates 43, 65, 94.

4. E. Hartkamp-Jonxis and H. Smit, eds., European tapestries in the Rijksmuseum, Amsterdam, (Amsterdam: Waanders Publishers: Rijksmuseum, 2004), pp. 273-281.

5. R. Sackville-West, Knole, Kent (London: The National Trust, 2003); R. Sackville-West, Inheritance, (Bloomsbury 2010).

6. S. M. Alsop, Lady Sackville: a biography, (London: Weidenfeld and Nicolson, 1978), pp. 184-86.

7. de Ricci, Twenty Renaissance Tapestries, p.1.

8. Morgan Library New York, Archives, J. P.Morgan papers Call no. ARC 1196, List of tapestries belonging to J Pierpont Morgan, no 27 Box 13, file 14. The archives of the dealers French and Co are held by the Getty Institute.

9. C. J. Phillips, History of the Sackville family, 2 vols, (London: Cassell, 1929); Sackville-West, Knole. Pictures of the property are at

10. The Knole archives are deposited in Kent Archives, Maidstone; reference for 1690 inventory, U 269/E 3, f.25; for 1694, U 269/069/1, a bundle of 27 sheets, the useful ones are nos. 9, 20, 21; for 1702 and 1705/06, U 269/79/2.

11. Kent Archives, U 269/ E 4, Inventory 1765, U 269/E 5, Inventory 1799.

12. These conclusions are based on information from the Conservator. Consultation with the College of Arms, London, failed to produce an identification because the description (de Ricci, Twenty Renaissance Tapestries, p. 28) is incomplete; 1 & 4 blazoned argent three billets sinople, 2 & 3 three billets sinople between a chevron azur – ie lacking the tincture of the field.  Green (sinople) is now visible only from the beck.  However, the identification suggested to de Ricci, that the family was Delves, is incorrect.

13. A. G. Bennett, Five centuries of tapestry, (San Francisco: The Fine Arts Museum of San Francisco and Chronicle Books, 1992), no. 19; W. Rytz, “Der Tausenblumenteppich mit dem Wappen Philipps des Guten in Bern. Seine Bedeautung fur die Geschicte der Pflanzenbildung und deren Auswertung”, Jahrbuch Bern Historical Museum, Bern 1961.

14. The tapestry can be viewed at; size: 7ft 1 ¾ inches x 22 ft 01/4 inches (2.18 x 6.72 m.); G. Wingfield-Digby, The Victoria and Albert Museum, Catalogue of Tapestries Medieval and Renaissance, (London: HMSO, 1980), pp. 47-8, plates 58a,b. Another example is pictured in G. Delmarcel and E. Duverger, Bruges et la Tapisserie, (Bruges, 1987), Cat 6; an earlier example, with two roundels bearing arms flanked by palm trees, ibid., Cat.4.

15. Hartkamp-Jonxis and Smit, European tapestries in the Rijksmuseum, figs.104, 107 and Cat.Nos.66-69.

16. On display in the Hamburg, Museum fur Kunst und Gewerbe, Inv. 1900-520, Brussels, c.1550; the ground there, however, is dark blue; S. Franses, Giant Leaf tapestries of the Renaissance 1500-1600, inaugural exhibition 21st October-25th November 2005 (London: S. Franses, 2005), fig.19.

17. F. De Nave, and D. Imhof, eds., Botany in the Low Countries, end of the 15th century-ca.1650, (Antwerp: The Plantin-Moretus Museum and the Stedelijk Prentenkabinet, 1993), explores the complicated inter-relationship of illustrations in more detail, pp. 11-18. A. Arber, Herbals, their origin and evolution, 3rd edition revised W. Stearn, (Cambridge: Cambridge University Press, 1986), is an indispensable introduction.

18. P. Bosqued Lacambra, “Aproximacion a la flora y vegetacion en los tapices de Bruselas del siglo XVI del Patrimonio Nacional de Espana”, in Felipe II el rey intimo. Jardin y naturaleza en el siglo XVI, Aranjuez 1998, (Madrid: Sociedad para la Conmemoracion de los Centenarios d Felipe II y Carlos V, 1998), pp. 77- 102. My thanks to Professor Guy Delmarcel for this reference.

19. E.A.B.Barnard and A.J.B.Wace, “The Sheldon tapestry weavers and their work”, Archaeologia,78 (1928) : pp. 255-314; purchase of the Sudeley hanging, Catalogue of household furniture, Hill Court, 22 April 1845, lot 59, lot 60, copy at Sudeley Castle; H. L.Turner, “Finding the Weavers; Richard Hyckes and the Sheldon Tapestry works”, Textile History 33, no. 2 (2002) : pp. 137-161; idem, “Tapestries once at Chastleton House and their influence on the image of the tapestries called Sheldon: a re-assessment”, Antiquaries Journal 88 (2008) : pp. 313-46, doi

For a more relaxed account of the story see, now archived at search term tapestries.

20. R. Strong, 600 Years of British Painting, (Denver: Denver Museum, 1999), p. 56, illustrated. The tapestry is now in a private collection. It was sold twice in the 1920s; sale catalogue, Christie’s London, (London) 10 May 1923, (sale 63), Lot 93; Christie, Manson, Woods, (London) 17 July 1930, Lot 155 and more recently Christie’s New York, 10 January 1995, lot 145.

21. H. Goebel, Wandteppiche, III, Die Germanischen und Slavischen Landen, (Leipzig: Klinkhardt & Biermann, 1933), vol. III, ii, p. 132, plate 108.

22. V. Woldbye and C. A. Burgers, Gewen Boeket, Exhibition Catalogue, 20 Nov 1971-30 Jan 1972, Rijksmuseum, Amsterdam, (Amsterdam: Rijksmuseum, 1972), p. 81, no. 53, size: 6 ft 11/2 inches x 9 ft (1.87 x 3.02m). See also V. Woldbye, “Flemish Tapestry weavers in the service of the Nordic kings, in G. Delmarcel, ed., Flemish Tapestry Weavers Abroad, Emigration and the Founding of Manufactories in Europe, (Leuven University Press, 2002), pp. 91-112.

23. E. A. Standen, “The Carpet of Arms”, Bulletin of the Metropolitan Museum of Art, 20 (1962) : pp. 221-31; Size : 7ft 6 inches x 16ft 4 inches, (2.28 m x 4.97m)

W. Wells, “The Luttrell Table Carpet”, Scottish Art Review, xi no. 3 (1968) : pp. 14-18; Size : 6ft 4 inches x 18 ft 7 inches (1.93 m x 6.66 m.).

24. Sale Catalogue, Sotheby’s Hannover, (Schloss Marienburg) 5-15 October 2005, lot 1410; Goebel, Wandteppiche,, volume III in two parts, plates, ii, 39 a,b,c; 51; 53, 54; 75.

25. E. A.Standen, European post-Medieval Tapestries and Related Hangings in the Metropolitan Museum of Art, (New York: The Metropolitan Museum of Art, 1985), pp. 199-202, no.30.

26. Wingfield-Digby, The Victoria and Albert Museum, Catalogue of Tapestries, Cat. 63, 64-65, plates 91-92, the story of Jacob, T.180-185-1925.   A second set, similar but not identical, was sold Sotheby’s London 20-22 November 1985, lots 460-462; these, like the arms of Sacheverell, were once called Sheldon, cf ibid, Cat.71c,79, plate 102c and Size: 19 x 18½ inches (0.46m x 0.48m); more probably dated 1578-1580, see and associated pdf  For Filoli's border ribbon see Guy Delmarcel, Flemish Tapestries, (London,1999), 185

27. C. Schellenberg, ‘Zwei Wirkerein des 17 Jahrhunderts’, Jahrbuch der Hamburger Kunstsammlungen 4 (1959) : pp. 939-50, fig.3. See also R. Bauer, ‘Flämisch Weber im deutschsprachingem Raum’ in Delmarcel, ed., Flemish Tapestry Weavers Abroad, pp. 63-89.

28. K. Brosens, ed.,European Tapestries in the Art Institute of Chicago, (New Haven and London: Yale University Press, 2009), pp. 363-367, Cat 61; G. Delmarcel, ed. Flemish Tapestry Weavers Abroad, Emigration and the Founding of Manufactories in Europe, Proceedings of the International Congress, Mechelen, October 2-5, 2000, (Leuven: Leuven University Press, 2002).

29. Hartkamp-Jonxis and Smit, European tapestries in the Rijksmuseum, pp. 273-281

30. As quoted in Brosens, Chicago, p. 364 from G. van Ysselsteyn, Geschiedenis der tapijtweverijn in de noordeljke nederlanden: Bijdrage tot de geschiedenis der kunstnijverheid, 2 vols, (A.W.Sijjhoff’s Uitgeversmaatschappij, 1936,) p.549; E. Duverger, Antwerpse kunst inventarissen uit de zeventiende eeuw, 13 vols, Fontes historiae artis Neederlandicae 1. (Koninklijke Academie voor Wetensschappen, Letteren, en Schone Kunsten van België, 1984-2002).

31. D. Starkey, ed., The 1547 Inventories of Henry VIII, ( London: Harvey Millar, 1998), nos. 12713 and cf 12888; 12719, 13027, 12147, 13026, 9059 and cf 12146. A yard or yerde is 3 feet or 0.915 m.

32. J. G. Nichols, ed., “Inventory of the Goodes of Dame Alice Hungerford, 1523”, Archaeologia 38 (1860) : pp. 353-372, esp. p. 364.

33. Canon Scott Robertson, transcribed, “Leeds Castle, Kent, Inventory, AD 1532”, Archaeologia Cantiana, xv (1883) : pp. 382-85, esp. p. 382.

34. TNA (The National Archives, London), LR 2/115, f.73v.

35. National Art Library, London, Special Collections, KRP.D30, f.104.

36. H. L. Turner, “Tapestries once at Chastleton House”, pp. 329-31.

37. A. Heal, ‘A Great Country House in 1623’, The Burlington Magazine 82 (May 1943) : pp. 108-116, esp. p. 113b.

38. F. Egan Last Bonanza Kings, the Bourns of San Francisco, (Reno and Las Vegas: University of Nevada Press, 1998), pp. 196-219; more information and pictures of the house at

39. F. M. Todd, The story of the Exposition, (New York: G.B.Putnams’ Sons, 1921), five vols.

40. A. G. Bennett, Five centuries of tapestry, pictured in second edition only, Appendix C; additional information, with grateful thanks, from Wendy Hefford.

41. T. Scheliqa, “A Renaissance Garden in Wolfenbuttel, North Germany”, Garden History 25 no 1 (Summer 1997) : p. 1-27,; J. Roberts, “Gardens of the Gentry in late Tudor Period”, Garden History 27 no. 1 (Summer 1999) : pp. 89-108,; Paula Henderson, The Tudor House and Garden: architecture and landscape in the sixteenth and early seventeenth centuries, (New Haven and London: Paul Mellon Centre for Studies in British Art, Yale University Press, 2005).

42. The Bourns may have had another reason to purchase the tapestry, for it is endowed with a curious personal link. Their only child, a daughter Maud, married Arthur Rose Vincent in 1910. It was for them that Bourn bought the Irish estate of Muckross, Killarney, Co Kerry. Long before Bourn’s interest, however, the estate had been the property of Henry Arthur Herbert who, in 1781, married the Honourable Elizabeth Sackville of Knole (b.1762), names likely to have been on the property deeds,Phillips, Sackville Family, ii, p. 179. Pictures at

43. W. Lawson, A New Orchard and Garden, (London: 1618) facsimile of the 1656 edition with introduction by M. Thirsk, (Totnes: Prospect, 2003), Chapter 17. Also available on EEBO.

Appendix I - List of Plants


Solanum dulcamera


Malus sp.


Vaccinium myrtillus

Bitter Orange

Citrus aurantium

Bleeding hearts

Dicentra spectabilis


Centaurea cyanus


Borago officinalis

Canterbury Bells

Campanula medium


Silene sp.


Prunus avium


Clematis sp.


Aquilegia vulgaris


Primula veris

English Daisy

Bellis perennis



Ficus caricara


Digitalis purpurea

Garden pea

Pisum sp.


Vitus sp.


Crataegus laevigata


Viola tricolor


Iris germanica

Lily of the valley

Convollarius sp.


Lilium sp.


Pulmonaria sp.


Mandragora officinarum


Mespilus germanica


Melilotus officinalis


Vinca minor

Pheasant’s eye

Adonis annua


Cucumis melo


Narcissus sp.

Oxeye daisy

Leucanthemum vulgare


Paeonia mascula


Dianthus sp.


Prunus domestica


Punica granatum

Pot Marigold

Calendula officinalis


Primula vulgaris

Poppy (field)

Papaver rhoeas


Cydonia oblonga


Rosa sp.


Ruta graveolens

Scarlet pimpernel

. Anagalis arvensis


Prunus spinosa


Galanthus sp


Sorbus domestica


Malcolmia maritima


Fragaria verna

Summer snowflake

Leucoium vernum


Tulipa sp.


Viola odorata

Wood Anemone

Anemone nemorosa


Lonicera periclymenum


Pyrola rotundifolia

Text:  © 2011 H. L. Turner MA DPhil
Layout and Design:  © 2011 Kent Archaeological Society