Dyeing with Dried Japanese Indigo Leaves

The easiest way to save Japanese indigo is to dry the leaves. This is also the only option, really, when you grow a small amount of plants.


In traditional Japanese dyeing with Japanese indigo, the harvested leaves were composted (fermented) in a very specific way, sprinkling the leaf mass with water and turning it over. The timing had to be just right, and Jenny Balfour-Paul writes in “Indigo, Egyptian Mummies to Blue Jeans” that the indigo farmers referred to the packing of the leaves as “putting the baby to bed”. Every time the leaf mass was turned over, sacrifices of rice wine were made to Aizen Shin, the god of indigo.

Composting Japanese indigo was serious business – a difficult and big undertaking. The composting process can only get going if the leaf mass is sufficiently large, on the order of 100’s of kilos (or several hundred pounds). The end result were composted leaves that contained a higher percentage of indigo than the fresh ones. This mass is known as sukumo.

People who only grow a few plants (like I do) have to find a different method. Having read about it on Deb McClintock’s page, I decided to dry my Japanese indigo leaves last year. And I did manage to do so after some trial and error.

The dry leaves look like this:

Dry leaves of Japanese indigo, 2016 harvest.

Some of the leaves look a bit blue, and that does make you think there’s indio. I’ve been wondering why drying the leaves would work (the must have been good reasons for the traditional sukumo method) and I’ve come up with the following story:

In living leaves there’s no indigo, only a precursor called indican. Here, the meaning of the word precursor is a molecule that can undergo some reaction(s) that produce indigo.

Indican production is thought to be a defense mechanism for the plant. In living leaves, the indican is primarily found in a compartment within the cell called the vacuole (shown by a Japanese team of researchers in this paper).

The plant cell also contains enzymes that are able to break down indican, producing indoxyl and sugar, but these enzymes are found in other compartments of the cell.

When you pick leaves and dry them, cell membranes will break because of the loss of water. So at some point, indican and enzymes from other parts of the cell will mix, and indoxyl is formed. When two molecules of indoxyl combine, blue indigo is formed.

I used Deb McClintock’s version of John Marshall’s method but I fiddled about quite a bit, finding my way to do it. The main change is that I didn’t discard the yellow dye, so I get a green-teal instead of blue.

Green-teal with dried leaves of Japanese indigo. From left to right, the skeins are 1st, 2nd and 3rd dip in a vat made from 50 g of dried leaves (3rd skein was naturally grey). The skein on the left was dipped 3 times in a vat made from 25 g of dried leaves. I’m knitting from the first skein already, the striped boy’s jacket in the background.

For my first attempt, I used 50 grams of dried leaves to dye 3 100-gram skeins of wool. The vat stopped working early on, so I added a bit of this, a bit of that. That lead to no recipe, but the result was completely fine.

My next attempt was made during an indigo workshop I taught a while ago, and I know it was hugely optimistic to bring such a difficult project. That vat only gave a slight hint of mint green, but at least we got a lot of brilliant blues from the ordinary indigo vats.

Afterwards, I started thinking that the vat may have gone wrong because the temperature was too low. This also makes sense when thinking about this failed experiment where I kept leaves lukewarm for a longish time.

High temperature during part of the vat preparation seems to be important, and that is a part of the method I ended up with for my third attempt:

First, I simmered 25 grams of dried leaves in water (enough to cover them) for 20-30 minutes. It wasn’t a rolling boil, but some bubbling going on.

To dye blue, the first water should be discarded and new water poured on the leaves. I did not do that, so I kept the yellows from the leaves.

I added 5 grams of sodium dithionite and about 1 tablespoon sodium carbonate. Check that pH is 9, and add more sodium carbonate if it isn’t.

Then, I simmered the vat for 15-20 minutes. It seems wrong to boil a vat after adding reducing agent and base, but in my attempt where I didn’t boil it at this step, it didn’t work.

I took the pot off the heat and added another 5 grams of sodium dithionite. I let it sit until the temperature was 40-50 C, then strained the leaves out. For my first attempt, I left the leaves in to get as much out of them as possible, but that is not a good idea. At this point, they are quite slimy and stick to the yarn.

When the temperature was 40-50 C, I put the pot on gentle heat to stay at that temperature. At this point, the vat is ready for use. I dipped a 100-gram skein of wool 3 times, and it turned a nice teal.

I’m impressed by the dye content of the leaves. 50 grams of dried leaves gave nice color to 300 grams of yarn, and 25 g gave a brighter color to 100 grams of yarn. My last vat was not exhausted, it had turned dark the next day because the indigo had been oxidized. I didn’t have more yarn on hand, but the vat could have given light shades on another skein.

Wool dyed teal with Japanese indigo, accompanied by fresh and partially dry leaves.

But I’ve saved the best for last: light fastness. I tested light fastness of the first skein from the 50-gram vat from July 1st to September 1st. The left side was covered and not exposed, right side was exposed to the light. I can’t really see any difference between them, and that means the light fastness rivals that of indigo blue. And that is quite impressive for a green-teal color!

Light test of Japanese indigo teal. Two months of sunlight did not affect the color.

PS: I’m growing Japanese indigo again this year. I harvested the first leaves on September 17. this year, and they are drying. They look even bluer than the ones from last year…

Læs dette indlæg på dansk

Seasonal Color Variation

An experiment with yellow to green tones of birch leaves over the summer. I didn’t see any difference, but most experiments do have different outcomes than expected.


A fresh new year calls for a new, big series of dye experiments, but I’m going to begin with an old one that was going on for so long that I never wrote about it.

“Yellow can be many things, so for each plant, I will  specify the particular yellow it gives. There will always be differences, the tone being more green earlier in the year.”

This is what Ester Nielsen writes in her introduction to natural sources of yellow in her book “Farvning med planter” (Dyeing with plants). This Danish book, published 1972, is full of useful information, but such a claim as this is just begging to be tested. I decided to use birch leaves for the test.

Birch leaves. It’s impossible to tell from the outside that they contain a good, warm yellow color.

To test the claim that colors are greener early in the growing season, tending towards yellow later, I made two 10 g test skeins of supersoft wool. I dyed each of them with 40 g of fresh birch leaves, since Ester Nielsen recommends 4 times the weight of fiber in fresh plants (twice the weight of fiber if using dried plants). I picked the first portion of leaves on May 11th, the second on July 4th.

The picture below shows the result. The skein in front is dyed with the leaves from May, the back one with thw leaves from July. They’re almost the same color, so my little experiment didn’t back up Nielsen’s claim…

Wool dyed with fresh birch leaves. The front skein is dyed with leaves picked in May, the back one with leaves picked in July.

To check for other possible differences, I tested the light-fastness of the two skeins. But again, no difference. The only thing worth noting in the light test shown below is a really good light-fastness of both yarns. The test took place over more than a month of summer.

Light test of wool dyed with birch leaves picked in May and July.

My conclusion: the time of harvest does not affect the color achieved with birch leaves. But that may only apply to birch leaves. It is possible that other plants to have a variation from yellow-green to yellow as the summer passes.

Late Summer Greens

This summer, I’ve dyed a nice pile of green wool using reed flowers and velvet pax – two dyestuffs that are a highlight of the dyer’s year. Reed flowers because they give such an electric green. You have to admit it’s a bit strange that these red flowers dye wool a wild green, but only if you get them into the dye pot absolutely fresh. If the flowers have opened or are not freshly picked, they will only give yellow. Velvet pax because its dusty greens are so lightfast. The two skeins in the back are dyed with velvet pax, the three in the front with reed flowers.

grøn green
Greens from reed flowers and velvet pax, the essence of late summer dyeing.

I’m becoming better at finding velvet pax. The first couple of years, I looked for it too late in the season. This year, I’ve found it growing several places, for example this archetypical plantation, where Dagmar is picking a big one. Just the kind of place that velvet pax likes to grow.

Dagmar picking velvet pax (with the arm that’s not broken).

Velvet pax can be found in August, and this year, everything was early, so it was there at the beginning of August. And the mushrooms were huge – I found some that were 25 cm across.

Characteristic brown tops of velvet pax, captured in a typical habitat.

Big, fat spiders are another joy of late summer. This one, which is possibly the fattest spider I’ve ever seen, lives outside our house. When I was sticking my camera right in its face, the neighbor’s big dogs started barking. Immediately, the spider lifted its front legs as if to attack. I chose to run away, so I only got a good shot from underneath the spider, where its pattern looks a bit like eyes. I think it’s a very light colored cross spider, since its body is pointy at the back. After reading that they can bite if provoked, I think my decision to flee was not a bad one.

My pet spider.

Summer is also the time of year to test light-fastness. I tested a handful of colors on the windowsill from early July to mid August, and their light-fastness was quite different.

  1. Old polypores, the two top ones warm baths and the lower one a cold bath that brewed outside for some weeks. None of these yellow browns are very light-fast.
  2. Velvet pax, the color didn’t change. I’ve seen this light-fastness in previous test, so it really is that good!
  3. Orange Cortinarius mushrooms, I don’t know which species. Not that light-fast
  4. A matrix of madder and indigo, showing that saturated colors are much more light-fast than pastels
  5. Sorrel root, not very light-fast
  6. Birch leaves. Surprisingly light-fast
  7. Weld. Surprised by the fact that it’s less light-fast than number 6…
  8. Henna on alpaca. I’d say this is a medium light-fastness
  9. Calendula flowers. Surprisingly light-fast
Light testing summer 2016.

I’ve also dyed with tansy, which doesn’t give green, but “just” yellow on alum mordanted wool (no pictures of that). But when I admired the flowers, I suddenly wanted to check if they really do stick to Fibonacci numbers.

The Fibonacci series begins with two ones, and then the next numbers are found by adding the two previous ones:

1, 1, 2, 3, 5, 8, 13, 21, 34, 55, etc.

The last time I thought about Fibonacci numbers were for calculating the numbers of my Vindauga blanket where rectangles obey the golden ratio, approximated by the ratio between neighboring numbers in the Fibonacci series, eg. 55/34 = 1.61.

Below is a close-up of a tansy flower. And as promised, the numbers of rows of tiny buds are Fibonacci numbers – 13 clockwise rows and 21 counter-clockwise.

Tansy flower obeying Fibonacci’s sequence.

Reed Flowers


Reed flowers are in season! I took the photo above on a beautiful August day at the lake. The sound of the wind through the reeds is positively mind-cleansing, which apparently I’m not the only one to think, judging by the number of YouTube videos of just that phenomenon. So now you too can enjoy it, even if you are in a skyscraper somewhere:

But I can promise you there was nothing mind-cleansing about my search for this dyestuff. They often grow a little bit out into the water, in places that are a bit hard to reach. I found a good patch of them growing in shallow water, I just had to cross a small forest. It looked fine, but it was actually a bottomless swamp, which I sank into to mid-thigh. Afterwards, I was a bit shocked, but otherwise fine! And I picked about 100 g of the silky soft, discreetly burgundy colored flowers: reedflowers

I rushed them to the dye pot – you have to use them fresh – and this is what I got on 20 g of supersoft in the first bath, 10 g in the second: first a lovely green, then a cold yellow:



Mordant 10% alun

Water Tap

Yarn Supersoft 575 m/100 g

Yarn:Dyestuff ratio 5:1 in first bath, 10:1 in second

Conclusion I love this green color, it’s beautiful and has a good light-fastness

Possible improvements The only one I can really think of is, that reeds should spontaneously grow in places that are easier to reach for the natural dyer

Last summer, I collected reed flowers while on holiday, and brought them home with me. I didn’t know that you have to put them fresh into the dye pot. The results were pleasing enough, although more towards yellow than green.

I have knit with the reed flower dyed wool from last year, and used the remaining scraps for light testing. These were on the windowsill for about a month from mid July to mid August. With this daylight calculator and a simple spreadsheet, I find that the samples for the 1st, 2nd, and 3rd bath of 2014 reed flowers got 572, 497, and 481 hours of daylight, respectively (since they spent the intervals 9/7-15/8, 10/7-11/8, and 11/7-11/8 on the windowsill). This is what they look like afterwards: the left side of each card has been covered, and the right side exposed:


A very good result indeed. You can tell that the paper has yellowed slightly from the exposure, but the yarn has only changed very slightly. The change is towards the yellow, so the green component seems to fade, but the color intensity is quite unchanged.