← OrreryAll Orbits storiesAn Orrery publication

Points of Interest

Geber and His Tribe

You will have heard that an eighth-century alchemist named Jabir ibn Hayyan, the father of chemistry, discovered the strong acids and left his name in the word gibberish. The recipe is real, but it turns up in Latin some five hundred years after he is supposed to have lived. The Arabic books behind his fame look like a century's work by a secretive circle, and the man himself may never have existed. Who, then, did the chemistry?

The myth

Now, you will have met this gentleman in a chemistry lesson, if your teacher was a good one. Jabir ibn Hayyan, born in 721, is the father of chemistry. An information sheet for teachers from Queen Mary University of London says he invented or discovered distillation, and it is quite plain about the rest: "Jabir Ibn Hayyan discovered both these acids using distillation", meaning hydrochloric and nitric. He then combined them, it says, into aqua regia, the liquid that dissolves gold. A 2024 report from Saudi Arabia, on his being honoured at the King Salman Science Oasis, adds sulfuric acid and silver nitrate. In medieval Europe he was Geber. And if you enjoy words, you may have been told that his jargon was so thick that it gave us "gibberish".

It is a generous story, told with real pride, and it does a useful job. It reminds people that chemistry did not begin in a European cellar. So let us treat it kindly and treat it as a case. We have a famous man, some very strong acids and a great many books. Our question is a simple one. Who was in the room?

A small cast, one of whom may not be there

Jabir ibn Hayyan of Kufa, said to be a pupil of the sixth Shiite imam, Ja'far al-Sadiq. Here is the telling detail: no historical source that we know of mentions him before about 900.

Abu Sulayman al-Sijistani, head of a scholarly circle in Baghdad, who died around 981. He disliked alchemy, so keep that in mind when he speaks.

Ibn al-Nadim, the Baghdad bibliographer, who listed Jabir's books in 987 and stood up for him.

Richard Russell, who signed himself "a Lover of Chymistry" and put Geber into English in 1678, from his house "at the Star in New-market in Wapping, near the Dock". His title page promises "The Most Famous Arabian Prince and Philosopher".

Marcellin Berthelot, the French chemist who went through the old manuscripts in the 1890s.

E. J. Holmyard, an author of school chemistry books and Jabir's warmest English friend.

And Paul Kraus, a young Prague-born scholar with a long list of languages, whose story does not end well.

The recipe with the wrong passport

Let us begin with the acids, because that part can be checked.

Open Russell's Geber at the book called The Invention of Verity, chapter 23, "Of Solutive Waters and Incerative Oyls". There it is. Take vitriol of Cyprus, saltpetre and alum, and distil off a water, "for it is very Solutive". That is a recipe for what we now call nitric acid. Then comes the lovely part. Dissolve a fourth part of sal ammoniac in it, and the water grows sharper, "because that dissolves Gold, Sulphur, and Silver". That is aqua regia, and historians cite this Latin text for the earliest known recipe for nitric acid and the first documented aqua regia.

So the recipe exists, and it has Geber's name on it. Case closed? Well, look at the date on its passport.

Holmyard, who was inclined to think Geber genuine, set out the difficulties himself in his Makers of Chemistry in 1931. No Arabic originals of these Latin books have been found. Their contents differ from the Arabic Jabir. Their style is that of the medieval schoolmen. He answered each point as well as he could, but he made one remarkable admission: "the present writer knows no Arabic book of Jabir's so systematically arranged as The Sum of Perfection."

There is more. Modern scholars put the main Latin book, the Summa perfectionis, slightly before 1310, and neither Albertus Magnus nor Roger Bacon, who would both have been interested, mentions it. Hermann Kopp had already gathered evidence that the Latin and Arabic authors were different men. Berthelot then compared them directly. His verdict on the Summa (my translation of his French) was that it "was certainly not written by Geber" and looked like a purely Latin work by a thirteenth-century impostor. M. Plessner, in the Dictionary of Scientific Biography, calls the clincher Berthelot's recognition that one Latin Geber, the Book of Seventy, really is a translation of a book listed by Ibn al-Nadim. The five famous Latin titles are not on that list.

Who wrote them, then? In the 1980s the historian William Newman proposed Paul of Taranto, a thirteenth-century Franciscan from southern Italy who wrote on alchemy under his own name. It is a tentative identification, though many accept it as likely.

Not everyone does, and you should hear the other side. Ahmad al-Hassan, a historian of Arabic science, called Newman's case "a maze of bewildering assumptions" and argued that "the Summa is based in its entirety on Arabic alchemy". He also said that "several recipes for nitric acid are given in Jābir's Arabic works". Holmyard had pointed the same way, naming a book called The Chest of Wisdom. Al-Hassan's is the minority view, and I have not been able to read those Arabic recipes. So the first nitric acid may be Arabic after all. Hold that thought, because we are about to ask whose Arabic.

One last clue from the Latin. Near the end of the Summa the author says, in Russell's English, "We have not treated of our Science with a continued series of Discourse, but have dispersed it in divers Chapters", so that the unworthy should not get hold of it. The reader must go around collecting the scattered pieces. Put that in your pocket.

The witness who knew another man

Now to Baghdad, and to the awkward thing about the Arabic books. Doubts about them are nearly as old as the books themselves.

Plessner reports that the earliest historical mention of Jabir is a hostile one. Al-Sijistani, in his Notes, denied that Jabir wrote the corpus at all, and said the true author was one al-Hasan ibn al-Nakad of Mosul, a man he knew personally. Remember his dislike of alchemy. He may have been scoring a point, and I could find nothing else about this al-Hasan. Still, it is quite something for the first witness in the case to name a different suspect.

Ibn al-Nadim, a few years later, tells us there were people who said Jabir had never existed, and others who would allow him only one book, the Book of Mercy. He insisted that Jabir had been real. But notice that the argument was already going on in the tenth century.

And how many books were there? Modern counts run to about 600 titles known by name, some 215 of which survive. Kraus reckoned the total, counting treatises and chapters, at about 3,000, a method the historian Nomanul Haq criticises as inflated. Holmyard, in 1931, drew a charming conclusion from the sheer bulk. It proved the old man must have lived a long life, since "no man who died young would have had time to write them."

There is a warm biography in that same book of Holmyard's: the druggist father executed for his politics, the orphan sent to Arabia, the friendship with the imam, the great ministers at Harun al-Rashid's court. Holmyard admits that "much is conjectural". The court stories come partly from the Jabirian books themselves; Holmyard marks one of them with his own note, "says Jabir". The witness, in other words, is also the accused. Later scholars have been unkind to it. Kraus rejected the court stories in 1931, and Thijs Delva argued in 2017 that the druggist was very unlikely to be the father.

Mr Kraus counts the books

Enter Paul Kraus, born in Prague in 1904. He trained in Berlin, left for Paris when the Nazis came to power, and moved to Cairo in the late 1930s. He did something rather simple and enormous. He gathered the manuscripts and read them.

Plessner sets out what he found. The corpus was the work of a school, not one man. Its learning and vocabulary presuppose the Greek translations of Hunayn ibn Ishaq's ninth-century circle, so a pupil of an imam in the eighth century could hardly have used them. It refers to theological arguments that point to the same later period. Nobody cites these books until the first half of the tenth century. And Kraus heard in them the veiled propaganda of an Ismaili sect. He dated the whole thing to about 850–950, and that date has been widely accepted.

The case is not entirely shut. Fuat Sezgin argued that there really was an eighth-century Jabir who wrote many of these works; Pierre Lory, in the revised Dictionary of Scientific Biography, calls that "mere suppositions". Haq also defended an early date. Lory allows that there may be an older kernel inside the corpus. He rejects Kraus's Ismaili label, and Leonardo Capezzone in 2020 places the group among extremist Shiites instead. As for the man, Lory quotes Manfred Ullmann: "The very existence of Jabir ... still seems very doubtful." A person called Jabir may have written something once. The trail goes cold there.

Capezzone adds the detail I promised you. The group kept to a rule he calls the "voluntary dispersion of knowledge". Roughly speaking, no one book holds the whole art, and the reader has to chase the rest. Does that sound familiar? You will have it in your pocket. It may mean that the Latin author knew his Arabic models well, or it may be a point for al-Hassan. Secrecy was also a common habit among alchemists, so I would not convict anybody on it.

As for Kraus, his great study came out in Cairo in 1942–43. He had married Bettina Strauss, the sister of the philosopher Leo Strauss, in 1941, and she died giving birth to their daughter in 1942. He was accused of misusing library funds. In October 1944 he was found dead in the house of Albert Hourani, where he had rooms. The Egyptian police called it suicide, though his family did not accept that. Leo Strauss adopted the child. The third volume on Jabir was never written.

Why the truth is the better story

So here is the reveal. As far as the evidence runs, there is no single father of chemistry in this case. There is a family of people using a borrowed name.

First come the Arabic writers, whom Capezzone calls a community "gathered under the pen name" of Jabir, working over about a century. In their world, he notes, putting a master's name on your book was a "guarantee of authoritativeness and success". What did they do behind it? They left the oldest known instructions for getting an inorganic salt, sal ammoniac, out of plants, blood and hair, and the oldest known systematic classification of chemical substances. And they had a stranger ambition. With their imam hidden from the world, they wrote about making human beings by art. Capezzone reads the project as one "to generate an artificial imam in a laboratory". You will not find that on the teachers' information sheet.

Then, centuries on and across a sea, comes a Latin writer, perhaps a friar in southern Italy, who takes the same famous name, presumably because it will get him read. Under it he writes what became one of the most widely read alchemy books in the late medieval West. Then comes Mr Russell in Wapping, whose title page has by now made the man an Arabian prince.

Which brings us to gibberish. Dr Johnson's Dictionary of 1755 says the word is "probably derived from the chymical cant, and originally implied the jargon of Geber and his tribe". The etymologist Anatoly Liberman is firm that Johnson was wrong: "Gibberish is certainly not a derivative of a proper name." But look at Johnson's last three words. He got the word wrong and, quite by accident, the people right. It was Geber and his tribe all along.

And the tribe, whoever they were, agreed on one thing, across the centuries and a change of language. The Arabic books, as Holmyard quotes them, say: "The first essential in chemistry is that thou shouldest perform practical work and conduct experiments". A Latin Geber, in Russell's English, claims to record what "we have operated, seen with our Eyes, and handled with our Hands". Neither writer, it seems, signed his real name. Both told you to go and try it. For a story about borrowed names, that is a good deal of honesty about the part that mattered.


On sources. The Latin Geber is quoted from The Works of Geber, translated by Richard Russell (London, 1678), read in the Internet Archive's Early English Books scan and machine-read text, with long s and spelling lightly regularised; the title page, Russell's signed note, Invention of Verity chapter 23, Sum of Perfection book 2 chapter 14 and the Investigation passage on what was "seen with our Eyes" were all read there. I have left out the quantities because the scan's reading of them is unclear. Reading the recipe as nitric acid and aqua regia is the modern interpretation, not Russell's. The Latin text was not checked. The priority claims, the date "slightly before 1310", the silence of Albertus and Bacon, Newman's identification and its tentative standing, and the book's wide readership are from Wikipedia's "Pseudo-Geber" and "Paul of Taranto" articles, citing Newman (1985, 1991) and Karpenko and Norris (2002); Newman's edition and articles were not reached, so I know his case only through summaries and his critic. Al-Hassan's quotations are from "The Arabic Origin of the Summa and Geber Latin Works: A Refutation of Berthelot, Ruska and Newman on the Basis of Arabic Sources", on his site history-science-technology.com; he names Sunduq al-hikma among Jabir's works, and I did not see the recipes or check how old that book is. Holmyard is quoted from Makers of Chemistry (Oxford, 1931), in the Digital Library of India scan on the Internet Archive; that he wrote school textbooks is from the Archive's listings. Berthelot is from La Chimie au moyen âge, volume 1 (1893), in a poor machine-read scan; the translation and the word "impostor" for his "apocryphe" are mine. Plessner's and Lory's statements, including the Ullmann quotation, are from the Complete Dictionary of Scientific Biography entries on Encyclopedia.com. Ibn al-Nadim I know only through Plessner, Capezzone and Wikipedia; Fück's 1951 translation is paywalled, and I did not use an unofficial upload of Dodge's 1970 translation. Al-Sijistani's Notes are known here only through Plessner, and the remark about his hostility to alchemy is from Wikipedia's Jabir article, citing Kraus and Delva. The online text of Plessner gives Hunayn's death as 847; the usual date is about 873, so I have said only "ninth-century". The sources do not quite agree on the first mention: Wikipedia says about 900, while Plessner calls al-Sijistani (d. about 981) the earliest historical mention and has the first citations in the first half of the tenth century. The first-mention date of about 900, the title counts, Kraus's and Haq's numbers, the sal ammoniac instructions, the classification, Kraus's 1931 rejection of the court stories and the Delva summary are also from that article, which cites Kraus (1942–43), Lory and Haq. Leonardo Capezzone, "The Solitude of the Orphan", BSOAS 83 (2020), read in the open-access PDF, supplies "pen name", "guarantee of authoritativeness and success", "voluntary dispersion of knowledge", artificial generation and "artificial imam"; the last is his interpretation, drawn from Jabirian treatises I have not read, and "extremist" is my gloss on his "Shiite extremist". His reading is one recent scholar's. Kraus's life is from Wikipedia's article on him and the Encyclopaedia Judaica entry on Encyclopedia.com; that article flags thin sourcing, and it gives the death as 10 or 12 October. Judaica says suicide, and the family's contrary claim rests on the same Wikipedia article. Joel Kraemer's essay "The Death of an Orientalist" (1999) was not read, because the only full copy I found was a pirate upload. Kraus's own volumes were not read directly either; his case comes through Plessner, Lory, Capezzone and Wikipedia. Johnson's sentence is as quoted by Michael Quinion at World Wide Words; the dictionary's online edition did not return the entry. Liberman is from OUPblog, 17 December 2008, and the claim that Johnson's derivation is wrong is his; other etymologists also treat the word's origin as uncertain. Etymonline and Britannica refused automated access. The Queen Mary sheet is a teaching resource, undated; the Saudi report is from ajel.sa, dated 14 September 2024. Not confirmed: whether any Jabir existed; who al-Hasan ibn al-Nakad was; who wrote the Summa; whether nitric acid appears first in Arabic or Latin; how Kraus died.

Original narrative history. Sources named in-line.  ·  Orbits is published by Orrery.