Sawwort (Serratula tinctoria)


Sawwort is a perennial plant with purple thistle-like flowers and serrated-edged leaves. It produces colours similar to those from weld (Reseda luteola) and, as with weld, the main dye component of sawwort is luteolin. According to Dominique Cardon (Natural Dyes 2007), sawwort has been used in Europe since the Middle Ages, especially in areas where weld was not harvested, and in Tuscany during the 14th and 15th centuries it was as highly regarded as weld.

Sawwort is a dye plant I have wanted to try for a long time but had not been able to find, although it is native to Britain and grows wild here. Recently I discovered a supplier of native wild plants (, from whom I bought several sawwort plants last Autumn. Sawwort should not be harvested for dyeing until the second year but as usual I was too impatient to wait until next year, when my plants would have been more mature.

I used about 100% dyestuff to weight of fibres and this produced a dyebath strong enough to enable me to use the same dyebath 3 times for successively paler yellows. The results were very pleasing and I shall probably order more sawwort plants this Autumn, especially as sawwort is a perennial plant and should prove more reliable as a garden plant than weld, which is sometimes not easy to grow. (This year I had no self-seeded weld plants and the seeds I sowed in the Spring germinated but the seedlings failed to develop because of the wet weather and are still no bigger than they were in April.)

The photo below shows, from left to right: alum mordant x 3 samples, alum mordant & iron modifier x 2 samples, alum mordant & copper modifier x 2 samples. The first 3 samples show a range of yellows from (a) the original dyebath, (b) the first exhaust dyebath & (c) the second exhaust dyebath

Dyes of the Celts

The word “Celt” is apparently derived from the Greek word “keltoi”, meaning “barbarian”, and is used to describe tribal societies in Iron-Age and Roman-era Europe, who spoke Celtic languages and were loosely tied by similar language, religion and cultural expression.

In preparation for a workshop on Celtic Dyes that I led recently at Fishbourne Roman Palace near Chichester, I have been doing some research into the dyes that may have been used by the Celts in Iron -Age Britain (c.600BC – 50AD).

The proto-Celtic culture in Europe was the Hallstatt culture (c.800BC – 450BC), named after the site of rich grave finds at Hallstatt in Austria. This culture then spread over much of Europe, into Britain, France, central Europe, the Iberian Peninsular and Northern Italy. The conditions in the Hallstatt salt mines, where the graves were discovered, meant that the textiles found there were relatively well preserved and analysis carried out on some of them gives an indication of the dyes and techniques used by the Hallstatt Celts.

Woad and weld were identified on textile fragments and also tannin, although it is not possible to identify the precise source of this tannin. Other unidentified yellow dyes were also found (perhaps sawwort or chamomile) and there is a possibility that lichen purple may also have been used. The red dyes analysed are interesting – there was no trace of any of the madder-type sources of red but some indication of unidentified insect dyes, possibly Polish or Armenian cochineal, and also of kermes. Both white and naturally pigmented wool was dyed and there is evidence of over-dyeing to create further shades. The issue of mordants is problematic; iron and copper were identified on several fragments and aluminium was identified on one or two fragments. However, it is possible that these minerals were present because of contamination within the salt mine, rather than because they were intentionally used in dyeing.

For my experiments I used white and naturally-coloured brown and grey wool. As I think it is unlikely that Iron-Age dyers in Britain would have had easy access to mineral alum. I decided not to use an alum mordant. Some samples were pre-treated in tannin from oak bark and others were unmordanted. The dyes I used were: weld (Reseda luteola), hedge bedstraw (Galium mollugo), lichen purple (Ochrolechia tartarea), oak bark (Quercus spp.) and woad (Isatis tinctoria). Although there was no trace of Galium spp. dyes on the Hallstatt textiles, I felt it would be historically correct to use native bedstraws in my experiments, as there is evidence from elsewhere that these plants were used for dyeing in the Iron-Age. Some samples were then over-dyed in woad and some samples were also treated in an iron-water solution after dyeing.


Some of the results of Celtic dye experiments.

The top row shows the colours from lichen purple with vinegar added to the dye-bath (except for a couple of woad samples on the extreme right). The second row shows lichen purple colours, without vinegar added to the dye-bath. The third row shows the colours from hedge bedstraw (Galium mollugo). The fourth row shows the colours from weld (Reseda luteola) and the bottom row shows the colours from oak bark. On each row some of the samples have been over-dyed in woad.

The Romans commented on the brightness of the clothing of the native Britons and the results of my experiments indicate that, even without alum as a mordant, it was certainly possible to produce bright colours in the Iron Age. The Celts also produced fabrics patterned with checks, stripes and plaids and this must have added to the impression of brightness.

Memories of Summer

As the colder weather sets in and Christmas approaches, I find my thoughts returning to the warm, sunny days of Summer. One of my most enjoyable activities this Summer was introducing my 3-year-old granddaughter, Milly, to the delights of using plants for colour and this proved very popular. As I wanted to avoid any confusion between “dyeing” and “dying”, I decided to call our experiments “making colours for white wool”.

We collected onion skins, dyer’s broom tops, dyer’s chamomile flowers, dahlia flowers and marigolds (Calendula) and we also used some of the frozen viola petals (as used for the ice-flower experiments described in an earlier post). We put these into glass jars, together with a skein of alum-mordanted wool and we added an iron spike to the jar with the marigold flowers. These jars were then left in the sun and Milly regularly checked the progress of the colour on the skeins each time she visited us. When she was satisfied with the colour on each skein, she pronounced the experiment finished and we then rinsed, washed and dried the skeins. She now has a small sample of each colour in her own personal scrapbook.

I am looking forward to introducing Milly to more dyeing experiments next Summer, when the weather is warm enough for working outside.

This photo shows some of the experiments in progress. From left to right: frozen viola petals, dahlia flowers and onion skins

This photo shows the results of our experiments. From left to right: dyer’s chamomile, frozen viola petals, dyer’s broom, marigolds (plus iron spike), dahlia flowers and onion skins

Anglo-Saxon dyes – woad

Woad (Isatis tinctoria) was the Anglo-Saxon source of indigo blue. Although woad is not a true native plant (i.e. it was not present here before the formation of the English Channel), it is thought that it was introduced in the neolithic age when farming began. Some of the earliest textile fragments show evidence of having been dyed with woad and it was probably one of the first dyes to be used. As extracting blue from the indigo-bearing plants is somewhat more complicated than the method of extracting colour from most other plants, it may seem strange that blue was among the first dye colours. However, the indigo-bearing plants, including woad, were generally considered to have healing properties and it may be that their use as dyes developed from their use medicinally. For example, if woad leaves were applied to damaged skin as a poultice, perhaps together with urine, which was regarded as an antiseptic, the conditions necessary for extracting blue from the leaves might have developed. These conditions would be heat (from the skin) an alkaline medium (from the urine as it became stale) and bacteria from the urine. So one can imagine that, if the poultice was removed to reveal blue skin beneath it, people would have been able to work out how to use the leaves to dye textile fibres. Another possible scenario might occur if woollen fleece was being cleaned in a tub of urine and someone dropped woad leaves into the tub by mistake. The woad leaves might remain in the tub long enough for the urine to act on them and could, in effect, create a woad vat in the tub. When removed, the fleece would become blue on contact with the air. Once people noticed the presence of the woad leaves in the tub, they would probably have been able to work out why the fleece had become blue. All this is purely speculation, of course. Woad vats would have been organic in the Anglo-Saxon period and might have been made using stale urine, which provides both the source of alkali and the bacteria needed to make the vat active.  Woad leaves may have been harvested and used fresh or they may have been allowed to ferment and processed into woad balls and stored for later use. Another method of dyeing with woad may have been the fermentation vat, made using wood-ash water as the source of alkali and madder and bran to induce fermentation and remove the oxygen from the vat. The recipes for these vats can be found in "Colours from Nature". The first photo below shows pale blue shades from woad. In the second photo the first three skeins show a range of shades from a woad fermentation vat. (The other skeins show lichen purple and black achieved by dyeing.


Anglo-Saxon colours from oak leaves and acorns

Although walnut hulls are often the dye of choice for browns, I decided to use oak leaves and acorns in my tests, because the walnut tree is not native to Britain and walnuts may not have been widely available during the early Anglo-Saxon period. I harvested the acorns and oak leaves in early Autumn and dried the leaves before use. As oak leaves and acorns are rich in tannin, no mordant is needed.

The first photo below shows colours from oak leaves. An alum or clubmoss mordant produced slightly more yellowish colours and a tannin mordant made the colours deeper. An alkaline modifier increased the depth of these colours. Mid-grey was achieved on unmordanted fibres modified in iron and a very dark grey was achieved on tannin-mordanted fibres modified in iron.

The second photo below shows colours from acorns. The comments on mordants and modifiers, made above for oak leaves, also apply to acorns. The  dark grey was achieved on tannin-mordanted fibres modified in iron.


Anglo-Saxon dyes – lichen purple

Lichens of the species Ochrolechia and Umbilicaria give beautiful brilliant purple and red shades when treated in stale urine or a solution of water and ammonia and no mordant is required for these dyes. This purple colour seems to have been used relatively rarely by the Anglo-Saxons, probably because the lichens needed to produce it only occur in restricted regions of Britain, mainly in the North and West in hilly areas or rocky, coastal districts

Analysis of the dyes used in textiles from the early Anglo-Saxon period shows that purples from lichens were used in embroidery, narrow woven bands and accessories, such as bags and headdresses, rather than to dye larger fabrics. Bearing in mind the scarcity of purple-producing lichens in southern and eastern England, this is perhaps not surprising.



This photo gives some indication of the beautiful purples available from lichens but does not do justice to the brilliance of these colours, which sadly are not very lightfast.




I do not recommend using lichens for dyeing, except in very small test dyebaths, as lichens grow very slowly and may take a long time to regenerate. Lichens should never be harvested indiscriminately and some may be protected species and should never be gathered. It is very important to be sure you have correctly identified each lichen before even considering collecting any. However, even a small piece of lichen the size of a large coin can yield enough purple dye for most test purposes.

Purple-producing lichens are prepared by soaking them in stale urine or in a solution of 2 parts water to 1 part ammonia. Use a strong glass jar with a  well-fitting lid and shake or stir the solution every day. It can take several weeks for the purple colour to develop. When the solution is a rich purple in colour, strain off the liquid into a dye pot and add the fibres to be dyed, plus more water if necessary. If you have used ammonia, make sure not to inhale any of the rather unpleasant fumes. (Stale urine can be equally unpleasant, of course!) Then heat the solution gently to simmering point and simmer gently for about 30 minutes. Take great care when heating the solution, as ammonia can catch fire very easily. Then allow the fibres to steep in the solution overnight. Then remove the fibres, squeeze the excess liquid back into the pot and re-use the solution until it is exhausted.

If the dyed fibres are steeped in an acidic modifier (for example in a solution of clear vinegar and water) they will become redder in tone. Using an alkaline modifier, such as wood-ash-water, will make the fibres more purple in tone.

The 4th skein from the left in the photo below shows some of the variations from acid and alkaline modifiers.




Anglo-Saxon dyes – weld and dyer’s broom

Weld (Reseda luteola) and dyer's broom (Genista tinctoria) were used by Anglo-Saxon dyers for yellows and remain in use today because of their reliability. Yellow tends to be the natural-dye colour that fades most rapidly but yellows from weld have better fastness properties than most other yellow dyes. Dyer's broom has similar dye pigments to weld and also has good fastness.

The first photo below shows colours from dyer's broom and the second photo shows colours from weld. (Note: the difference in depth of colour between these two dyes is because I used a higher percentage of dyer's broom than of weld. If the same percentage of each dye is used, I would usually expect very similar depths of colour to result.)

Using an alum mordant gives the brightest yellows and clubmoss mordant produces very similar shades. If these dyes are used without a mordant the colour is considerably less bright. However, if an alkaline modifier is applied to these unmordanted dyed fibres, by soaking them in wood-ash-water, the colour becomes almost identical to the yellow achieved on an alum mordant. So, even if early dyers did not have access to mineral alum, they would still have been able to achieve bright, clear yellows from weld and dyer's broom.

The use of tannin as a mordant makes the colours deeper than they would have been without a mordant but this yellow lacks brilliance.

The use of an iron modifier after dyeing produces moss green shades on all fibres, although the green achieved on unmordanted fibres and those mordanted with tannin is duller.



Anglo-Saxon dyes – Madder

Madder (Rubia tinctorum) is one of the most ancient dyes and a particularly useful and reliable source of red. Other plants in the madder family (Rubiaceae) include the native plants lady’s bedstraw (Galium verum), wild madder (Rubia peregrina), hedge bedstraw (Galium mollugo), dyer’s woodruff (Asperula tinctoria) and woodruff (Galium odoratum).

Although madder (Rubia tinctorum) was available during the Roman period, it seems to have been replaced during the early Anglo-Saxon period by the native Rubiaceae (for example:lady’s bedstraw and wild madder). This suggests that the Romans imported madder as a dried dyestuff, rather than growing it in Britain, and that it disappeared with the departure of the Romans. There is evidence that madder began to be imported from France in the 7th century and by the later Anglo-Saxon period it had become a very common dye.

The analysis of dyes in textiles of the early Anglo-Saxon period seems to indicate that reds, like purples, were mainly used for narrow woven bands, headdresses, embroideries and accessories, such as bags, rather than for larger fabrics. Where dyes were used in larger fabrics (and dyes were detected in only one-third of the larger fabrics analysed) these were mainly dyes that give shades of yellow, blue and brown, plus green from blue and yellow dyes used in combination. It is also possible that, at least in the early Anglo-Saxon period, reds and purples were colours reserved for people of high status.

An alum mordant is necessary for true reds from madder (Rubia tinctorum) and reds achieved on a clubmoss mordant are very similar to those achieved on fibres mordanted with mineral alum. Without a mordant, madder gives colours in the orange to coral range and using a tannin mordant gives similar but slightly deeper colours. An alkaline modifier, such as wood-ash-water, makes the colours pinker in tone and an iron modifier makes the colours browner. The addition of chopped crab apples to the dyebath makes the colours brighter.

The first photo below shows a range of shades from madder (Rubia tinctorum)

In my experience it is not easy to obtain true rich reds from lady’s bedstraw, wild madder, dyer’s woodruff and woodruff. I grew all these for several years in my old garden but never managed to get the sort of red obtained from madder, also grown in my garden. The use of an alkaline (wood-ash-water) modifier moves these colours further towards red and an iron modifier makes the colours browner. I intend to continue experimenting with these dyes in the madder family – if, that is, I can grow enough roots to make the tests worth the effort. Lady’s bedstraw grows wild at the sides of many country roads in this area but it is against the law to uproot wild plants, so I shall have to rely on my home-grown plants and it will be a while before they are mature enough to harvest.

The second photo below shows colours from wild madder roots (upper left), wild madder dried tops (4 samples at lower left) and lady’s bedstraw (right)

The third photo shows colours from woodruff roots (Galium odoratum)



Anglo-Saxon mordants

This is the next post about my Anglo-Saxon-style experiments.

Today’s natural dyers tend to use mainly aluminium mordants and sometimes iron and copper, which may also be used after dyeing as colour modifiers. Chrome, which was not introduced as a mordant until the 19th century, has been popular among some dyers but is avoided by many nowadays because of its toxicity. Similarly tin, which was first used as a mordant in the 17th century (mainly with cochineal to produce bright reds) is now less frequently used, partly because of environmental considerations and partly because it can make fibres brittle. Copper was used by early dyers in the Mediterranean world and also in India in the classical period AD300 – 700 but, of the metallic mordants mentioned above, only alum and iron seem to have been used by the Anglo-Saxons. It is also possible that pots made of metals, such as iron or bronze, may have been used as dyepots and this may have had an effect on the colours. However, the only evidence of dyestuffs staining pots in the early Anglo-Saxon period occurred in pots made of clay and dyeing may have frequently been carried out in clay pots, which would probably not have had a significant effect on the colours produced.

Many of the dyes used by Anglo-Saxon dyers will fix adequately without a mordant and I think it is likely that many fibres would have been dyed in this way. Some textile fragments from the period show evidence of an alum mordant and mineral alum from the Mediterranean was probably used during the later Anglo-Saxon period. However, there is some doubt as to whether this mineral alum would have been available earlier in the period, so alternative mordants may have been used, such as aluminium extracted from clubmosses. Plants rich in tannins, such as oak galls and blackberry leaves and shoots, may also have been used as mordants.

In my experiments I tested most dyes in the following ways:

  • Without a mordant
  • With a 10% mineral alum mordant
  • With alum extracted from clubmoss
  • With a tannin mordant (from oak galls or bramble/blackberry leaves and twigs)
  • With an iron mordant / use of an iron pot

As clubmosses are rare in Britain, I would not advocate their use, except in very small quantities for experimental purposes. For my tests I used the following recipe for clubmoss as a mordant:

Use 200% clubmoss (Lycopodium spp), chop it up and put it into a pot filled with water, then heat to 40C. Hold at this temperature for 3 days. On the fourth day boil it up briefly and then strain off the liquid. Add the fibres to be mordanted, heat slowly to 40C and then allow to cool  Repeat this process daily for 3 days. Then remove the fibres and squeeze the excess liquid back into the solution, which can be re-used once. In order to be sure that aluminium had been extracted and then absorbed by the fibres, I used fibres mordanted with 10% mineral alum and unmordanted fibres as “controls” in the dyebath. The results were particularly clear with madder dye – the unmordanted fibres dyed to a coral shade but both the alum-mordanted and the clubmoss mordanted fibres dyed to an almost identical red shade.

In these tests I used 100% bramble/blackberry leaves and twigs as a tannin mordant. I simmered them for about one hour to extract the tannin, then strained off the liquid, added the fibres and simmered them for about 45 minutes, then left them in the liquid to cool overnight. (25% oak galls can be used instead of bramble leaves and twigs.)

I also experimented with the use of iron and wood-ash-water, which is alkaline, as colour modifiers after dyeing. It is also possible that stale urine may have been added to dyebaths and this would have increased alkalinity.

To make iron water, put some pieces of scrap iron or rusty nails in a large container with a well-fitting lid and fill it up with a solution of two parts water to one part clear vinegar. Leave the iron to steep in this solution for a week or two, until the solution is orange in colour. When you use the iron water, strain it through a fine-meshed sieve or a piece of muslin.

To make wood-ash water, put the ashes from a wood fire into a large glass or plastic container with a well-fitting lid. Fill up with water and leave the ashes to steep for a week or two, until the liquid is yellow in colour and feels slick or slimy to the touch. When you use the solution, pour or siphon it off without disturbing the ash sediment. Wood-ash water can be used as an alkaline modifier and as the source of alkali in woad vats. It can also be added to dyer’s broom and weld dyebaths to increase the depth of colour.

My experiments are intended to demonstrate some of the possible methods used by Anglo-Saxon dyers and to give an idea of the colour range they might have achieved. Although we don’t know exactly how early dyers worked, there is enough evidence to indicate that they would have been able to achieve a wide range of bright, rich colours from the relatively small number of dyes they used.

I will write about individual dyes in later posts.

Natural Dyes for Basketry Materials

Although I have occasionally dyed cane, willow and raffia experimentally in madder and indigo dyebaths, I have not done extensive tests with a range of dyes on basketry materials. Recently, Sussex basket-maker, Jackie Sweet, spent a day here with me and we experimented with some natural dyes on willow, cane and Phormium tenax (New Zealand flax).

Jackie Sweet is a talented and accomplished basket-maker, with a wealth of knowledge and experience, and her baskets range from more robust traditional English willow baskets to delicate woven containers, which are works of art in themselves. I have always loved baskets and I have some from many parts of the world in my collection, so the possibility of assisting Jackie in using natural dyes as part of the creative basketry process was something I was looking forward to.

Of course, the natural brown, beige or tan colours of the basketry materials would mean that the dye colours achieved would not have the brightness and clarity of colours achieved on white or cream-coloured fibres. Jackie’s main aim was to achieve deep colours that would provide contrasts with the natural colours of the basketry materials and enable her to incorporate more elaborate pattern designs in her work. She had treated some of the willow and cane in alum beforehand, by soaking the materials in a solution of 10% alum. This meant we were able to compare the results on both alum-mordanted and unmordanted materials.  I decided to experiment first with indigo, madder and onion skins, as I thought strong dyes would be needed, especially if the undyed materials were brown in colour. We also used iron and washing soda as colour modifiers.

We simmered the fibres for about 30 minutes in the prepared dyebaths and the results were quite good, especially on the alum-mordanted materials. The iron and washing soda modifiers intensified the colours but the green I had anticipated from onion skins plus iron did not appear. Overall, Jackie was pleased with the results and decided to experiment further, possibly with some natural dye extracts.

I suggested it might be worth experimenting with cool dyeing, using large plastic containers, as this would enable Jackie to dye larger quantities in a single dyebath. However, while cool dyeing should work well with indigo, I suspect strong dye solutions and considerable patience might be necessary for this method to work with other dyes. I also recommended extracting the dye colour by simmering first, so the dyebath would be hot when the fibres were added.

All in all, I spent a thoroughly enjoyable day with Jackie and I look forward to hearing about her progress and to working with her again.

The first photo below shows the undyed materials and the second photo shows the dyed materials on the left with some undyed materials on the right for comparison. The third photo shows a woven design worked by Jackie using naturally-coloured and madder-dyed Phormium tenax.