Six days ago, the “TürkiyeToday” website published a story revealing that lost works by the ancient mathematician, Apollonius of Perge, had been found in an Arabic manuscript in Leiden by Prof. Jan Pieter Hogendijk. The story has since reached other outlets, including the Jerusalem Post.
This story would be very exciting, if it was true. But after some investigation, I find that it’s complete nonsense. The “lost works” have been known for centuries.
Let’s start with the story as it appears online now:
Lost works of ancient mathematician Apollonius of Anatolia found in rare Arabic manuscript
By Koray Erdogan Feb 5, 2025
Scientists have uncovered two lost books by Apollonius, the renowned mathematician from Anatolia, in an Arabic manuscript housed at the Leiden University Libraries in the Netherlands. This extraordinary find sheds new light on the preservation and transmission of ancient Anatolian knowledge during the Islamic Golden Age.
Apollonius (262 B.C.–190 B.C.) (Apollonius of Perge) is celebrated for his groundbreaking work in geometry, particularly in his book The Conics of Apollonius. This work, which introduces the concepts of hyperbolas, ellipses, and parabolas, was one of the most influential in ancient Anatolian mathematics. However, only four of the original eight books of “The Conics” were available to European scholars during the Renaissance.
The missing books—five and seven—are now found preserved in an 11th-century Arabic manuscript, a translation that had been lost to history.
The manuscripts, which were acquired by Dutch orientalist and mathematician Jacob Golius during his travels to the Middle East in the 17th century, form part of a vast collection of nearly 200 manuscripts that he brought back to Leiden University.
Golius’s acquisition of these texts not only enriched Western scientific scholarship but also played a crucial role in the rediscovery of lost ancient works.
The newly revealed Arabic translation of Apollonius’s lost books is accompanied by detailed illustrations and exquisite Arabic calligraphy. The Dutch mathematician and historian of science, Jan Pieter Hogendijk, emphasized the importance of these manuscripts as symbols of the intellectual achievements of Islamic scientists, noting their precision and artistic quality. “These manuscripts are a testament to the mental discipline and focus of their creators,” Hogendijk stated.
The story the rest of the story is padding, although it refers to a recent volume published in Leiden, “Prophets, Poets and Scholars” as if it was the source.
In fact the news story does not make sense. Apollonius of Perge did indeed write a textbook on geometry called the “Conics”, in 8 books. Only the first four books are preserved in Greek. Books 5 to 7 are indeed preserved only in an Arabic translation.
But this Arabic translation has been known for centuries. None other than Edmund Halley, of Halley’s Comet, published an edition of the Arabic text of these books with Latin translation. A modern edition and English translation appeared in 1990 by Gerald Toomer:
Apollonius of Perga; Toomer, Gerald J. (1990). Conics, books V to VII: the Arabic translation of the lost Greek original in the version of the Banū Mūsā. Series: Sources in the History of Mathematics and Physical Sciences. Vol. 9. New York: Springer (1990). DOI:10.1007/978-1-4613-8985-9. ISBN 978-1-4613-8987-3.
In the preface to the Toomer volume, indeed, on p.viii, we read:
I wish to acknowledge a particular debt of gratitude to my former colleague, J.P. Hogendijk. He read the whole book in draft form, corrected a number of errors, made several suggestions for improvement, and offered some original contributions, which I have gladly incorporated. I need hardly say that all remaining imperfections are solely my responsibility.
Today I have communicated with Prof. Hogendijk. I learn that he was unaware of the news reports, despite some of them claiming to have an email interview with him. The Prophets, Poets and Scholars publication from Leiden does indeed exist, and he has a paper in it. This he lists on his home page here as item 69. Here’s the bibliographical detail. But he kindly sent me a copy of the paper, and this makes no such claims.
Jan P. Hogendijk, Jacobus Golius and his Arabic manuscripts on the exact sciences, pp. 114-123 in Arnoud Vrolijk, Kasper van Ommen, Karin Scheper and Tijmen Baarda, eds., Prophets, Poets and Scholars: The Collections of the Middle Eastern Library of Leiden University , Leiden: Leiden University Press, 2024. ISBN 9789087284077 (print); 9789400604520 (ePDF), https://doi.org//10.24415/9789087284077
So… the news story is pure rubbish. One feels for Prof. Hogendijk, to whom all this nonsense is being attributed.
* * * *
Ancient Greek mathematical texts are today a field of limited interest, perhaps. In 17th century Europe, on the other hand, they were of real interest to contemporary mathematicians, trying to solve the same problems.
But reading the preface of the 1990 Toomer book reveals some quite fascinating details of the process whereby the text actually came to light. Here are some abbreviated extracts from p. xxi onwards.
Since it was known from Apollonius’ introduction to Book I of the Conics that they were originally in eight books, there was considerable interest in 17th-century Europe in recovering the missing Books V-VIII. As we have seen (p. xviii) Books V-VII were extant in Arabic, both in the original translation made by Thabit for the Banu Musa, and in various reworkings. The study of Arabic was beginning to expand in Europe in the later 16th century, and underwent a real flowering in the 17th. Furthermore a manuscript of one of the Arabic versions of the Conics had reached Italy in 1578, and others came to northern Europe from 1629 on. Yet nothing of significance for the later books3 was published until 1661, and no translation of the version closest to Apollonius’ original, that of the Banu Musa, appeared until 1710…..
Among the oriental manuscripts which were given in 1578 by Ignatius Ni’matallah (“Neama”, “Nehama”), Patriarch of Antioch, to Cardinal Ferdinando dei Medici, later Grand-Duke of Tuscany, and founder of the Medicean printing-press in Rome, the first Arabic press, was a compendium of the Conics by Abu’l-Fath Mahmud al-Isfahani. Its importance was recognized by the man in charge of the press, Giambattista Raimondi, who intended to publish an edition, but had still not done so when he died in 1614. At that date there is some talk in the correspondence of Galileo about publishing the work, but then interest lapsed until 1645…
In the meantime two other Arabic manuscripts of the Conics had been brought to Europe. In 1629 Jacobus Golius returned to Leiden from a prolonged visit to the east, bringing back a large number of manuscripts in Arabic, Persian and other languages. Among them was the splendid codex of Apollonius’ Conics, in the version commissioned by the Banu Musa, which is now in the Bodleian Library ..[ms. “O”] .. This was given to him by his countryman David Leleu de Wilhem “for the public good.” Knowledge of the existence of this manuscript (or rather of the apograph which had been made from it of Books V-VII) was rapidly disseminated by the catalogue of the manuscripts deposited by Golius in the Leiden library which the enterprising Gassendi had printed at Paris in 1630. Golius promised to publish the recovered books, and seemed ideally qualified to do so, since he was both an accomplished Arabist and a competent mathematician (shortly after his return from the east he was appointed to the Professorships of Arabic and Mathematics at Leiden).
However, not only did he fail to publish the lost Apollonius himself, but he manreuvred adrOitly and successfully to prevent anyone else doing it for the rest of his life lhe died in 1667), jealously guarding access to his manuscript and the partial copies of it which existed at Leiden, and actively deterring others who were inclined to try their hand, for instance by claiming that he alone could read the difficult script of the manuscript (which is in fact extremely legible). In the years 1644-1646 he was apprehensive that the English mathematician John Pell would anticipate him by publishing Books V-VII from the manuscript of Ravius (on which see below), but seems to have dissuaded him by pretending that he was about to produce his own version. Golius might be excused for not publishing the Apollonius himself on the grounds that he was not idle during these years, being engaged on, amongst other things, his great Lexicon Arabico-Latinum (Leiden, 1653). But his dog-in-the-manger attitude towards the efforts of others to get access to the work, or to publish alternative versions, caused increasing frustration and disgust throughout Europe and among his own countrymen, which can be traced through the decades of the 1630’s and 1640’s in the Mersenne correspondence. …
When Golius died in 1667 his valuable collection of manuscripts, including the Apollonius (see p. xxii) went not to the University of Leiden, but to his personal heirs. Negotiations between them and Cambridge University for the sale of the manuscripts as a lot fell through, and they remained, inaccessible, in the hands of the heirs for nearly 30 years. However, Golius was no longer in a position to restrict access to the partial copies of the Conics which he had handed over to the Leiden library, …
In 1696 Golius’ heirs finally decided to auction his manuscripts. Bernard persuaded Narcissus Marsh (then Archbishop of Dublin), a great patron of scholarship and interested in orientalia, to give him carte blanche to buy at the auction. Although ilt he traveled to Leiden for the sale in October 1696, and bought for Marsh a large number of interesting mss., including the famous Apollonius. But he never recovered from the effects of the winter voyage, dying at Oxford shortly after his return, on Jan. 12, 1697, aged 58, with most of his ambitious plans, including his edition of the Conics, uncompleted. However, the manuscript was now where it would be made available to the man who was to do the most for Apollonius in modern times, Edmond Halley.
Halley’s great edition of Apollonius’ Conics was published at Oxford in 1710. It was the editio princeps of the Greek text, and is until the present publication the only translation of Books V-VII based on the original Arabic version (that supervised by the Banu Musa): all later translations of these books derive from Halley’s Latin version. …
The manuscript of Golius is now in the Bodleian, where it is Marsh 667 (although apparently now labelled Arch. o. c. 3). It was written in 1070AD in Azerbaijan, but was in Aleppo by 1627, where it was acquired by Leleu de Wilhelm, for Golius, as indicated by a note in the volume.
While the news story is nonsense, let us hope that it draws attention to the possibilities of real discoveries, waiting to be found in Arabic translation. Maybe someone will be inspired to work more on these collections! That would be a nice outcome.