Potassium Carbonate, K2CO3.
A comment from Randy Asplund on Facebook 8/2015 I found very helpful:
"Potash vs. Potash. A while back someone asked me a question about pigment making and the answer involved potash. I don't remember who or exactly what, but I was just thinking that I ought to be more clear and differentiate between the two kinds of potash and the difference between that and "lye" ("ley"). The medieval potash, aka "lixivium" is Potassium CARBONATE (K2CO3) and THAT is the best thing to use in recipes for making things like iris green. That's what they used in the middle ages. You make it. Making it from hardwoods and vines is a diluted product.
The other potash is "caustic potash" which is KOH (potassium HYDROXIDE). You can buy both in crystal form. Be careful, because the crystals are very potent, and the hydroxide version is likely to kill your color.
Avoid using the kitchen lye because it is really caustic. It is NOT what you want."
Materials and Techniques of Medieval Painting
from woad pg 136-7
Alchemy of Paint
From Geffrei Maudeleyne the way to make it is found here:
From another source Prof. D. V. Thompson translated, De Arte Illuminandi pp. 36-7, footnote 47 "Lye made from the wood of the vine or of oaks is specified in Section 12. p. 8 above. It is of course, to be understood that the lye is made from the ashes of these woods. This is the product generally indicated by the Latin lixivium; its alkalinity depends largely upon the potassium carbonate, 'potash,'
K2CO3, extracted from the ashes. Oak wood, vine branches, beech wood, and particularly varieties of Artemisia, yield ashes rich in this alkaline carbonate. A word of warning must be given to experimenters not to use caustic potash, potassium hydroxide, KOH. Both KOH and NaOH2, sodium hydroxide are often called lye nowadays, but their use would often lead to failure in carrying out rules in which the milder carbonate solution, the common medieval lixivium, is called for. Solutions of commercial K2CO3 may be used safely, to avoid the inconvenience of lixiviating ashes for the purpose. If the lixivium is to be boiled with quicklime, as occasionally directed (e.g. Sloane 416, folio 137r. 'A fare fina lacha'), the potassium carbonate lye should be replaced by the hydroxide; but in the absence of special indications, the carbonate lye should be understood. See note 83, p. 45 below"