Previously we looked at the claim that Galen knows of soap. In the same article we find the claim:
Zosimos the alchemist  (c. A.D. 250) mentions both soap (σαπώνιον) and soap-making (σαπωναρικὲ τέκνη).
148. Berthelot, Collection des anciens Alchemistes Grecs, 1888, vol. ii, 142.3, 143.7. (This contains Zosimos; French translation here).
It is always good practice to check these claims. Fortunately we can. And if we look at the online copy of the Greek, we will see both these phrases for soap (sapo) where they are supposed to be. But what is the context?
As might be supposed, this is an alchemical recipe, in a book full of these. On p.143 of the French translation we find the text:
III. 8. On the same divine water
1. Taking some eggs, whatever quantity you like, boil them, and after breaking them, remove all the white, but don’t use the shell. Taking a male-and-female glass, which is called an alembic, put into it the yellows of the eggs, using the following amounts: 1 ounce of yellow, calcined egg-shell, two carats, neither more nor less, but exactly as written. Then stir; then, taking some more eggs, break them and throw them into the alembic with the stirred yellows, so that the whole eggs are covered by the yellow.
A process of distillation then follows, and the “water” is what comes out of this.
So what of our soap-making? A little further on it says:
2. … Then mix with the ashes other egg-yellows, as in the art of soap-making; stir together the wet and dry materials, and put them all into an alembic. Do the same operation as before, but changing the recipient of the water, that is the rogion.
The notes make plain that in this case “the ashes” is a sulphur residue; but the point is that Zosimos knows that soap is made by mixing ashes – from wood, in the case of soap-making – with a fatty substance.
(At this point I must confess that I knew little about how soap is made, but a google search tells me that wood ash, or “potash”, contains Potassium Hydroxide (KOH), which is the alkali in soft soaps, which combines with the fat – usually oil – to make soap. Hard soaps, in bar form, require Sodium Hydroxide.)
But there are two references to sapo in this text. The other is close to the end:
Now after 41 days, remove the alembic from the hot place and let it cool completely for 5 days. Once the 5 days have elapsed, place the alembic on the sawdust ashes and extract the divine water from it; not into your hand, but into a glass vessel. Then, taking this water, put it into an alembic, as before, and heat it for 2-3 days. After removing it, stir and expose to the sun on a shell. When the product becomes as compact as soap, warm an ounce of silver and cover it with this solidified water, i.e. 2 carats of dry powder, and you will have some for.
The total number of days of the operation is 110 days, according to what Zosimos the Christian and Stephanus say. As for me, having foraged from everywhere like a bee and plaited a crown with many flowers, I do homage to you, my master. Next I will explain what are the devices. Take care in Jesus Christ, our God, now and forever and in all the ages of ages. Amen.
The compact soap, from which moisture has evapourated, is all very well. But the last paragraph betrays that we are not dealing with a text written directly by Zosimos, whatever his date; but with a compilation of fragments, assembled from Zosimos, and from Stephanus – perhaps a commentator – in a form that looks suspiciously like a pupil’s notes.
In the circumstances, can we be sure that these references to soap making are original to Zosimos? Rather than a later addition?
It is slightly depressing to find the evidence for the early use of soap so fragile. The author of these statements did indeed know how to make soap, that much is clear. But was he Zosimos, or some much later writer? Without knowing more about the transmission of these works – and technical works, by their very nature, tend to be revised, amended, “corrected”, and so on, when copied – it might be rash to be certain.
All the same, I suspect that it is simpler and involves fewer hypotheses if we assume that the statements are indeed original, than to posit an interpolator; in which case, we may reasonably assume that Galen, in the late 2nd century, is probably using the word “sapo” in the same way as Zosimos, and therefore the use of soap does indeed start to appear in the times of Marcus Aurelius or thereabouts.