The flowers in William Morris's Holland Park carpet seem determined to fill every inch of the floor. Large blossoms open against deep blue. Pale stems loop between them. Around the edge, a red-and-cream border moves to a broader, heavier rhythm. There is plenty to look at, but the parts do not all ask for attention at once.
That relationship between abundance and order is a good way into Morris's carpets. They belong to the British Arts and Crafts story, but they also carry the results of his close study of Persian and other historical textiles. Their interest lies in what he did with that knowledge: flowers became a way of organising a surface, not simply decorating one.
Cover: Holland Park carpet, designed by William Morris, late 19th century. The Metropolitan Museum of Art, Bequest of Frank A. Munsey, 1927, 27.139.3. Public Domain.
What Morris learned from Persian carpets
Morris's admiration was practical as well as aesthetic. He examined historical carpets, collected them and advised the South Kensington Museum, now the V&A, on acquisitions. In 1893 he supported its purchase of the Ardabil Carpet. These were encounters with actual woven objects, not just a taste for an imagined Eastern style.
Art historian Caroline Arscott, in Morris Carpets, explains that Morris wanted to learn from the organisation and vitality of older carpets without simply reproducing their designs. His own patterns were to suggest growing plants and a connected world. Arscott's reading is about relationships within ornament, not a dictionary in which every flower has one fixed meaning.

The Holland Park carpet makes that distinction visible. Its centre, paired large forms and enclosing border provide a strong structure. Within it, the flowers turn in different directions and smaller branches keep the blue ground active. Readers familiar with Persian rug medallions will recognise some of this compositional language, but the result is a British design, not a Persian carpet under another name.
A drawing that already thinks in pattern
In this drawing, a flower can contain another plant. A broad leaf becomes a route between blossoms. The brown background is not an empty remainder: its shapes help separate the pale blue curves and yellow petals. The pattern works through these interlocking areas as much as through the outline of any single flower.
Morris's plants are recognisable without behaving like botanical illustrations. They bend, flatten and change scale to suit the design. The V&A describes a similar balance of observation and stylisation in his wallpapers. On a carpet, this approach lets a leaf remain readable as a leaf while also doing a larger job in the composition.
Hammersmith carpets were a workshop achievement
The name Hammersmith refers to the hand-knotted carpets associated with Morris & Co. Production began at Morris's London home in 1879 and moved to Merton Abbey in 1881. The name remained, so it should not be read as a guarantee that every example was woven at the earlier address.
Designing and making were closely related, but they were not the same credit. Arscott describes women at Merton Abbey tying the wool tufts into the foundation, row after row. Calling an object a William Morris carpet identifies its design authorship; it does not mean that he personally knotted the whole piece.
The Met records its Holland Park example as wool with symmetrical, or Ghiordes, knots, at 25 knots per square inch. That relatively broad construction suits substantial shapes. It also reminds us why fine drawing cannot be judged by knot count alone. Here the pleasure is in the relationship between generous flowers, curling stems and the firm edge of the carpet.
Not every Morris carpet was hand-knotted
Morris's reputation for handcraft can obscure another part of the business. His firm also sold machine-made carpets, including flatwoven Kidderminster and pile carpets made at Wilton. These were distinct products, not different names for the hand-knotted Hammersmith range.
The Tulip and Lily pattern belongs to the Kidderminster story described by the V&A. In the Met's example above, the flowers form a close, continuing repeat. There is no large central medallion to command the rectangle. The scrolling lines make the individual units feel connected, while the restrained green and yellow palette keeps them close together visually.
Compare that surface with Holland Park. One works through a field of similarly sized units; the other gives you conspicuously larger flowers and a strongly contrasting frame. The difference is not simply how much ornament each contains. It is how that ornament is distributed.
Look at the spaces between the flowers
The smaller drawing is easy to miss in the whole-carpet view. Come closer and the blue field is threaded with pale curls, leaves and tiny flowers. Step back and those marks recede, leaving the larger blossoms and red border to organise the view. The carpet offers different things at different distances.
This is the useful design lesson to take from Morris, without turning a present-day room into a Victorian reconstruction. A floral rug can have both detail and a clear overall shape. If you are comparing floral rugs, notice which parts remain distinct from across the room and which emerge only nearby. The museum works shown here are historical examples, not items in our current collection.
For the separate question of what to place beside a patterned rug, our guide to mixing rugs, curtains and upholstery looks at scale and contrast in a furnished room. Morris's carpets offer something to linger over first: not just the flowers, but the lines and spaces that make them belong together.
Sources and image records
- Caroline Arscott, “Morris Carpets,” RIHA Journal 0089 (2014), especially paragraphs 3-4, 7, 12-14 and 26-29.
- Victoria and Albert Museum: William Morris textiles; William Morris and the V&A; William Morris and wallpaper design.
- The Metropolitan Museum of Art: Holland Park carpet, 27.139.3; Tulip and Lily, 1998.276. Public-domain photographs.
- The Little Flower reproduction and its public-domain declaration are linked in its caption. Its identification follows that image record; no examination of the original drawing is claimed.
Laissez un commentaire