Some remarks on Islamic scientific manuscripts and instruments and past, present, and future research

David A. King

Introductory remarks

There are an estimated 10,000 scientific manuscripts in Arabic, Persian, and Turkish which, together with about 1,000 astronomical instruments, constitute the major sources for our knowledge of the exact sciences, astronomy and mathematics, in Islamic civilization. Most of these manuscripts and instruments date from after the most creative period of Islamic science, which spans the eighth to the fifteenth century. However, some late manuscripts also preserve for us earlier works which would otherwise be lost, and some late instruments bear features known to us only from early texts.

The ideal situation for documenting the history of Islamic science is the following. The manuscripts and instruments are first catalogued. When this has been done, the literary heritage and the instruments can be reviewed in a bio-bibliographical survey. Then detailed studies can be conducted of specific works or individual themes of singular importance. Thus our knowledge progresses from a reasonably sound foundation, or as sound as it Can be, given the fact that many of the major are lost without trace. Even given this optimum situation, what we could hope to recover of the Islamic scientific tradition is still very much a matter of chance.

The actual situation is, needless to say, quite different. First, the state of Cataloguing of these Islamic scientific manuscripts leaves much to be desired.  While some of the major Western collections have been catalogued with thoroughness, these catalogues were often prepared by Islamicists with little understanding of Islamic science. For the largest and richest collection in the Islamic world, the Süleymaniye Library in Istanbul, there is as yet no catalogue at all (though see below on the handlist of Max Krause). Neither is there a catalogue for the largest and richest collection of Islamic scientific instruments, namely the Museum of the History of Science in Oxford. Most of the major texts are inadequately published, some are not published at all — so our current activity is a rather haphazard operation. Also the most important individual Islamic instruments are still unpublished.

In the following brief survey of the history of publication of texts relating to Islamic science (mainly astronomy and mathematics) I shall mention only a few names, thereby inevitably overlooking many scholars deserving inclusion. However, even in the briefest survey certain names to whom we owe the greatest debt should not go unnoticed: Jacob Golius for gathering manuscripts, Wilhelm Ahlwardt and MacG. de Slane for cataloguing them Franz Woepcke and Heinrich Suter for analysing them, and Lewis Evans for starting the world’s richest collection of Islamic (and Other) scientific instruments at Oxford. In this survey I shall not deal systematically with Greek works in Arabic translation, nor Arabic works available only in Hebrew, Latin or Greek translations. Also I shall not deal systematically with astrology or geography. Colleagues will find that I have overlooked even various areas of astronomy and mathematics, let alone important medieval works on these subjects. In brief, the following remarks may tend to reflect my own interests and prejudices.

Folk science and mathematical science

We should begin by pointing out that Islamic science operated on two distinct levels. Seldom did these two traditions interact. The one was cultivated by the scholars of the sacred law (the fuqahā), who took as their authority the Qurʾān and the Prophetic sayings and devised simple procedures for regulating the calendar, organizing the times of prayer, and finding the direction of the Kaʾbah, the physical focus of Islamic religious ritual. The other was cultivated by the Muslim scientists. They favoured mathematical procedures, often extremely complicated, for the same ends. Inevitably the results obtained were different. The present writer has published several articles on the two traditions, and on the implications of their different approaches for popular timekeeping and the orientation of Islamic religious architecture and city-planning. In the same way the religious scholars developed a sacred cosmology, recently studied by Anton Heinen, which was quite different from cosmology promoted by the scholars influenced by the Hellenistic tradition. Islamic scientific cosmology has been studied mainly by S. H. Nasr, and the tradition also been surveyed by E. Jachimowicz.

Bio-bibliographical surveys

At the end of the last century Heinrich Suter made a translation of the part of the Fihrist of Ibn al-Nadīm dealing with scientists and their works. He also translated the sections of the then new catalogue of the Khedival Library in Cairo concerned with scientific manuscripts. This inspired him to compile a bio-bibliographical survey of 500 Muslim scientists and their works, which is still a major source for modern studies. Max Krause listed the early Islamic works on astronomy and mathematics in the libraries of Istanbul, providing a most useful supplement to Suter. H. P. J. Renaud did the same for the libraries of Rabat. Carl Brockelmann included the sciences in his monumental survey of Arabic literature but had only Krauseʾs incomplete overview at his disposal for the rich collections in Istanbul. Charles A. Storey did likewise for Persian literature. Fuat Sezgin has ʿredone Brockelmannʾ for the period up to c. 430 AH and devoted three spectacular volumes to mathematics, astronomy, and folk astronomy, together with astrology and meteorology. For much of the new material Sezgin relied heavily on the collections in Turkey and Iran, which he was able to visit in person. This is the only way to collect reliable information, and Sezgin’s volumes reflect this personal input; Suter and Brockelmann had compiled their works in their own libraries. Sezgin’s volumes are now the starting- point for any serious study of Islamic scientific writings. The more recent survey of Galina Matvievskaya and Boris A. Rosenfeld, covering the period 700-1700, is based on all of these and contributes new information; over 1,000 Muslim astronomers and mathematicians are featured in their work. Especially for the later period of Islamic astronomy (the thirteenth century and thereafter) my catalogue of the 2,500 scientific manuscripts in the Egyptian National Library and my bio-bibliographical survey based on it provide much new information. Also I have documented over 100 Yemeni manuscripts dealing with astronomy and mathematics.

For astronomical instruments we have Robert T. Guntherʾs monumental Early Science in Oxford (volume II) and Astrolabes of the World, both outdated and alas neither corrected nor updated for six decades. Gunther was no Arabist and was poorly advised by the Arabists at Oxford. On a Sounder foundation is the bio-bibliographical survey of Muslim instrument- makers by L. A. Mayer, now updated and expanded in the Répertoire of Alain Brieux and Francis Maddison, long awaited and now shortly to appear. The Répertoire and my forthcoming catalogue of all historically important Islamic instruments promise to arouse some interest in a much- neglected field. A catalogue of about 130 Islamic globes was published in 1985 by Emilie Savage-Smith. Each instrument has something to contribute to the overall picture, and most previous studies of Islamic instruments have been hit-or-miss ventures. Authors have invariably felt obliged to explain what an astrolabe is before embarking on the description of a single instrument, inevitably overlooking some of what we now consider to be the basic features: dimensions, inscriptions, construction marks, accuracy of markings, star-names, and the like. There is no substitute for a catalogue in which all of these are included.

Critical editions of scientific texts

The optimum procedure for dealing with the texts is a critical edition using all available manuscripts or a facsimile in the case of a unique manuscript, a translation into a European language, and a commentary based on the current understanding of the subject. The names of scholars who have contributed to our subject in this way is small indeed, but the sum total of their achievements is impressive. The serious study of Islamic astronomy and mathematics by scholars who were capable of reading the original texts began appropriately enough with Jacob Golius in Holland in the seventeenth century. Goliusʾs edition of the Arabic text of the epitome of Ptolemaic astronomy by al-Farghānī (Baghdad, c. 850), together with a Latin translation and partial commentary, was published in Amsterdam in 1669. Earlier, in fact already in 1594, an edition of the Arabic text of the recension of Euclid’s Elements attributed to Naṣīr al-Dīn al-Ṭūsī (Maraghah, c. 1260) had been printed in Rome.

In Paris in the nineteenth century a series of first-rate Orientalists had at their disposal the rich collection of Arabic manuscripts in the Bibliothèque Royale. Caussin de Perceval published three chapters from the introduction of the astronomical handbook (al-Zīj al-Ḥākimī) of Ibn Yūnus (Cairo, c. 990) dealing with his own observation records and those of his predecessors. The Sédillots, father and son, chose to study the astronomical handbook of Ibn Yūnus, the treatise on astronomical instruments by al. Marrākushī (Cairo, c. 1280), and the introduction to the astronomical handbook (Zīj-i Sulṭāni) of Ulugh Beg (Samarkand, c. 1425). The father made an analysis of the entire introduction to the tables of Ibn Yūnus, which was never published and is not known to exist, but fortunately was extracted by J. B. Delambre in his Histoire de l’astronomie médiévale (1819). The father also made a translation of the first half of the treatise of al- Marrākushī, namely, the part dealing with spherical astronomy and sundials, while the son later summarized the second part dealing with other

instruments. The son published in 1847—53 the Persian text and a translation of the introduction to the Zīj of Ulugh Beg. However, none of these scholars published any of the many tables in the two zījes.

In 1831 F. Rosen published the text and English translation of the Algebra of al-Khwārazmī (Baghdad, c. 825). Twenty years later Franz Woepcke published the Algebra of ʿUmar al-Khayyām (N. Iran, c. 1080) with French translation, and in 1874 two treatises on the geometrical instrument called the ‘perfect compass’.

A French translation of the treatise on constellation figures by ʿAbd al- Raḥmān al-Ṣūfi (Shiraz, c. 965) was published by F. C. H. C. Schjellerup in St Petersburg in 1874. The Arabic text was published in Hyderabad 80 years later (see below).

E. Sachau published the text and translation of two monumental works by al-Bīrūnī (Ghaznah, c. 1025): the work on chronology entitled Kitāb al- Āthār al-bāqiyah (1878, 1879), and the encyclopaedia on India known as Kitāb fi tahqīq mā lil-Hind (1888, 1910).

The French Arabist R. Dozy published in 1873 the Arabic text of the Calendar of Cordova of ʿArīb b. Saʿd (Cordova, 961), reprinted in 1961 with a French translation by Charles Pellat. H. P. J. Renaud published the text of an almanac attributed to Ibn al-Bannāʾ (Marrakesh, c. 1300). Pellat has recently published the text of several medieval Egyptian almanacs, and Daniel M. Varisco made an edition and an annotated translation of the important agricultural almanac of the Rasulid Sultan al-Ashraf (in press).

In 1891 A. Carathéodory published in Constantinople the text of Naṣīr al-Dīn al-Ṭūsī’s treatise on the spherical quadrilateral. This was, I think, the first scholarly edition of an Arabic scientific text to be printed in the Islamic world.

In the early years of this century the main contributions to our subject were made by scholars writing in German, with the notable exception of the Italian Nallino. Heinrich Suter published in 1914 the Latin text of the astronomical handbook, introduction, and tables, of al-Khwārazmī (Baghdad, c. 825), lost in the original. Many decades later Otto Neugebauer produced an English translation with a masterful commentary on the text and tables. Suter and, after him, Carl Schoy brought out German translations of numerous Arabic treatises. Carlo Nallino published the Arabic text and a Latin translation of the astronomical handbook of al-Battānī (Raqqah, c. 910). Max Krause prepared the edition of al-Qānūn al-Masʿūdī, the astronomical handbook of al-Bīrūnī, later published in Hyderabad. He himself published the text of the Arabic version of Menelaosʾs Spherics.

In 1934 R. Ramsay Wright published a facsimile edition of the Arabic text of the handbook on astronomy and astrology by al-Bīrūnī entitled al-Tafhīm fi ṣināʾat al-tanjīm, but the translation that he published on the facing pages was originally made from a manuscript of the Persian version. (This is clearly indicated in the introduction (pp. XII—XIII) but is not generally known)

ln 1936 Karl Garbers published the Arabic text of the treatise on sundial theory by Thābit b. Qurrah (Baghdad, c. 900). A German translation of this work, preserved in the oldest surviving Arabic scientific manuscript (dated 370/ 980 or 81), was made in 1937—38 by Paul Luckey. Alas the edition and translation of the treatise on sundial theory by Ibrāhīm b. Sinān (Baghdad, c.925), prepared by the same scholar in 1944 as a doctoral dissertation, is still unpublished.

In the 1940s and 1950s numerous Arabic texts were published by the Osmania Oriental Publications Bureau in Hyderabad, Deccan. While some of these are ‘uncritical editions’ simply copied from the manuscripts, others like the text of al-Bīrūnī’s zīj called al-Qānūn al-Masʿūdī are more reliable. The texts span the entire spectrum of Islamic science, from the Kitāb al- Anwāʾ on folk astronomy by Ibn Qutaybah (Dinawar, c. 860) to various works of al-Bīrūnī and Ibn al-Haytham. Whatever faults these editions may have, most of the texts represented would still be unpublished had it not been for the labours of the Hyderabad press. In Europe the 1940s and early 50s were bad years for our field. But new scholars emerged to continue in the same spirit as before World War II, and the scene of activity moved away from Germany to other parts of the world, notably Spain (Millás Vallicrosa and his school), Lebanon (Kennedy and his school), and the Soviet Union (a group of scholars inspired by A. P. Youschkevitch).

To J. Millás Vallicrosa we owe a good survey of all the available material on the astronomer al-Zarqāllī (Toledo, c. 1075), and an edition of the treatise on the astrolabe by Ibn al-Ṣaffār (Cordova, c. 1000) as well as editions of various works in Hebrew. His successor Juan Vernet published the introduction to the astronomical handbook of Ibn al-Bannāʾ, without the tables, Vernetʾs younger colleague Julio Samsó has contributed translations of a commentary on various treatises by Abū Naṣr b. ʿIrāq published in Hyderabad.

The Russian Arabist P. Bulgakov edited in 1962 the newly-discovered treatise Taḥdīd nihāyāt al-amākin on mathematical geography by al-Bīrūnī. E. S. Kennedy published translations of, and commentaries on, several works by al-Bīrūnī, including the Taḥdīd and the Ẓilāl on timekeeping in general and shadows in particular. He also produced a facsimile of al-Kāshīʾs treatise on the equatorium with annotated translation and has prepared an edition of the astronomical tables al-Zīj al-Khāqānī of al-Kāshī. Together with David pingree and other scholars he had brought out the astrological history of Māshāʾallāh (Baghdad, c. 800) and a book on theories underlying zījes by al-Hāshimī (Baghdad, c. 900?). David Pingree has provided us with the text of a Byzantine Greck version of the lost astronomical handbook of al-Fahhād (Shirvan, c. 1175), a Byzantine version of the Toledan Tables, and the Arabic text of the corpus of pseudo-scientific material known as the Picatrix. The same scholar has published all of the remaining fragments of the astronomical writings of some of the earliest Muslim astronomers, notably al-Fazārī, Yaʿqūb b. Ṭāriq and Abū Maʿshar. Marie-Thérèse Debarnot has made available text, translation, and commentary for al- Bīrūnī’s Maqālīd ʿilm al-hay’ah, a very important work on spherical trigonometry and spherical astronomy.

Gerald R. Tibbetts of London has produced a translation of the Kitāb al- Fawā’id by the fifteenth-century Indian Ocean navigator Ibn Mājid, the most substantial Arabic treatise on navigation. Earlier a French translation of this and other works by the same author had been published by G. Ferrand. More recently Ibrahim Khoury of Damascus has provided us with the Arabic texts of all surviving treatises by Ibn Mājid and the sixteenth-century navigator Sulaymān al-Mahrī.

E. Millás Vendrell has published the Latin version of a commentary of Ibn al-Muthanna (Andalusia, probably tenth century) on the astronomical tables of al-Khwārazmī. More recently B. R. Goldstein produced the text of two Hebrew translations thereof, again with translation and commentary. The same treatment he afforded to the treatise on planetary theory by al- Bitrūjī (Seville, c. 1190). His investigations of the interesting celestial spheres attributed to Ptolemy but not contained in his known works led him to discover that the relevant passage in Ptolemy’s Planetary Hypotheses had been omitted from the published text, and so he published a facsimile of the unique London manuscript, with translation of, and commentary on, that passage. Régis Morelon has recently made available the astronomical works of Thābit b. Qurrah (Baghdad, c. 900); again the treatment is optimal text, translation and commentary. Khalil Jaouiche has also published in the same way the treatise on the balance (qarasṭūn) by Thābit.

In 1975 Paul Kunitzsch of Munich published the treatise of Ibn al-ṣalāḥ (Baghdad, c. 1150) criticizing the stellar coordinates found in available versions of the Almagest. (For the same scholar’s other works on the Almagest see below.)

Anton Heinen undertook as a doctoral dissertation at Harvard University the difficult task of editing an important text extant in numerous copies, namely, the treatise on sacred cosmology Kitāb al-hayʾah al-saniyyah fi al- hayʾah al-sunniyyah by al-Suyūṭī (Cairo, c. 1500), which he provided with translation and commentary. One of his most important discoveries was that there was a whole tradition of such treatises before and after al-Suyūtī which await detailed study.

Arithmetic was well served by the late Ahmed Saidan of Amman. He published the arithmetic of al-Uqlīdisī (Damascus, c. 950); the arithmetic of Abū al-Wafā (Baghdad, c. 975); the mathematical works of Ibrāhīm b. Sinān; and several other texts. Sami Chalhoub of Aleppo in 1986 brought out the al-Kāfī fi al-ḥisāb of al-Karajī (Baghdad, c.1000).

Islamic treatises on algebra have fared rather well. We have already mentioned Rosenʾs edition of the Algebra of al-Khwārazmī. Aydın Sayılı published in 1962 the Algebra of a contemporary of al-Khwārazmī named Ibn Turk, with English translation. Roshdi Rashed has edited the Algebra of Sharaf al-Dīn al-Ṭūsī (Damascus, c. 1165) and various treatises by Ibn al- Haytham, and, together with Ahmed Djebbar, the Algebra of ʿUmar al- Khayyām, and, together with S. Ahmed, the Algebra of Samawʾal (Baghdad, c. 1180). Adel Anbouba in 1964 brought out the text of al-Badīʿ fi al-jabr by al-Karajī. ʿAbdelhamid Sabra and Barbara Harper provided critical editions of the Ḥall shukūk Kitāb Uqlīdis of Ibn al-Haytham and the Sharḥ Muṣadarāt Uqlīdis by the same author, respectively. J. Len Berggren has recently published the correspondence of the two late-tenth-century mathematicians al-Kūhī and al-Ṣābiʾ, and Yvonne Dold-Samplonius has translated a Kitāb al-Mafrūḍāt by an unknown author.

Gerald Toomer produced an edition of the unique Mashhad manuscript of the treatise on burning mirrors of Diocles, lost in the original Greek and now available with English translation and commentary. To Toomer we also owe a splendid English translation of Ptolemy’s Almagest, an edition of Books V-VII of the Conics of Apollonius (available only in Arabic), and a survey of the various components in the Latin Toledan Tables, based mainly on Arabic sources. Jan Hogendijk has published edition and ū.anslation of the Tamām Kitāb al-Makhrūṭāṭ, the ‘completion’ of the incomplete Conics of Apollonius by Ibn al-Haytham (Cairo, c. 1025). Tzvi Langermann of Jerusalem has recently brought out a treatise on planetary astronomy by Ibn al-Haytham.

To four members of a group of graduates from the Millás-Vernet-Samsó School in Barcelona, Roser Puig, M. V. Villuendas, Mercè Comes, and Emilia Calvo, we owe edited texts and commented translations of the treatise on trigonometry by Ibn Muʿādh (Jaén, c. 1020) ; two treatises on the universal

Plate by al-Zarquāllu (Toledo, c. 1070); a treatise on a universal plate of his own invention by Bāṣo (Cordova, c. 1295); and a study of various Andalusian treatises on the equatorium. From Joan Carandell we have an edition of the treatise on sundial construction by Ibn al-Raqqām (Tunis, c.1325).

Two new books reached me a mounth before this conference: from Maria Angeles Navarro of Granada the edition of an anonymous Andalusian almanac, and from George Saliba of Princeton an edition of the non- Ptolemaic astronomy of al-ʿUrḍī (Maraghah, c. 1260).

The following works of the same genre are not yet in print. Jamil Rageb has prepared a detailed study of the Tadhkirah, a work on theoretical astronomy by Naṣīr al-Dīn al-Ṭūsī (Maraghah, c. 1260). George Saliba has prepared an edition, translation, and commentary for the treatise on plaretary theory by Ibn al-Shāṭir (Damascus, c. 1350), which will surely by on the same scholarly level as his new volume on al- ʿUrḍī. Forthcoming are also editions of the Arabic and Latin texts of the Spherics of Theodosius by Richard Lorch, incorporated in a study which will contribute substantially to our understanding of transmission from Greek to Arabic and to Hebrew and Latin. S. M. R. Ansari has promised an edition of the Zīj of Jai Singh (Jaipur, c. 1725), the publication of which will an important step for our knowledge of late Indian astronomy. My own analysis of the spherical astronomy in the Zīj of Ibn Yūnus, completed many years ago, and my more recent analysis of the corpus of tables for timekeeping that was used in Damascus from the fourteenth to the nineteenth century, are still unpublished. Both would be more useful if the relevant manuscripts were available in facsimile editions (see below).

It would be amiss not to mention the monumental edition of the Optics of Ibn al-Haytham currently published by A. I. Sabra. (This tribute to the scholar of Basra was being printed in Kuwait before the outrages of 1990.) Likewise the texts and translations of various works on mechanical engineering and clocks by Sevim Tekeli, Donald Hill, and Ahmed Y. al- Hasan have advanced our knowledge of this important aspect of Islamic science.

That the above list is incomplete will be obvious to anyone who looks at the entries for al-Bīrūnī and Ibn al-Haytham in Sezgin’s bio-bibliogaphic surveys, let alone leafs through the entire volumes. My purpose is simply to convey an idea of the kind of international effort that has gone into building a firm foundation for our subject. Although so many works are already available in one form or another no one should think that our task is almost complete. On the contrary, the above-mentioned publications represent just a modest fraction of the prodigious output of the scientists of medieval Islam.

Surveys of individual topics

Critical surveys of various categories of Islamic scientific literature — dealing with, for example, theoretical astronomy, astronomical instruments, arithmetic, mensuration, theoretical geometry, and algebra — would be useful. Examples of publications of this kind are: a collection of the fragments of the works of the earliest Muslim astronomers, al-Fazārī and Yaʿqūb b. Ṭariq, by David Pingree; and an overview of early Islamic algebra by Adel Anbouba; a paper by Galina Matvievskaya on the theory of quadratic irrationals; the numerous works of Paul Kunitzsch on Arabic star- names in the Islamic world and in Europe; the paper of Jan Hogendijk on treatises dealing with the construction of the heptagon; the writings of Wilbur Knorr on the construction of two mean proportional and angle trisection; the various papers by J. Len Berggren on the construction of azimuth curves on astrolabe plates; the study by Richard Lorch of a text on isoperimetry; the contributions of Jacques Sesiano on magic squares; the writings of Sonja Brentjes on the Nichomachean tradition in Islamic mathematics; the various studies by George Saliba on non-Ptolemaic planetary models; and my own studies of different categories of procedures — from folk astronomy, as advocated by the scholars of the religious law, and from mathematical astronomy, as proposed by Muslim scientists — for regulating the lunar calendar, timekeeping and determining the times of prayer, and finding the qiblah for any locality. Numerous volumes of studies have been published in Barcelona by the Millás-Vernet-Samso school, and the texts they have edited will be mentioned below.

There were regional schools of astronomy in the late medieval period which are so distinctive in their interests and tastes that they merit separate studies. Julio Samsó has prepared a splendid survey of Andalusian astronomy which will be published in 1992 on the occasion of an exhibition in Madrid on Andalusian science. Aspects of Toledan astronomy in the eleventh century are already treated in an elegant article by Lutz Richter- Bernburg. E. S. Kennedy has published general articles on astronomy in Seljuq and Timurid Iran and Central Asia. The present writer has surveyed the history of astronomy in Egypt and Syria, the Yemen, and the Maghrib, and Ahmed Djebbar has written several studies on the history of mathematics in the Maghrib. For Ottoman Turkey and Muslim India Ekmeleddin İhsanoğlu and S. M. R. Ansari are respectively in control of the sources and have promised general surveys. For the influence of Ottoman astronomy in the Balkans we have the useful studies of Jasminko Mulaomerovic of Sarajevo. The influence of the Islamic tradition on the Hebrew tradition and vice versa is dealt with in a series of masterful studies by Bernard R. Goldstein. This having been said, it must be pointed out that there are no reliable surveys of early Islamic astronomy, of the reception of Greek words in Arabic, of the transmission of Arabic works to Europe, or indeed of Islamic astronomy in general.

Articles on different aspects of Islamic astronomy and mathematics are to be found in the Festschrift volumes of A. Aaboe, W.Hartner, and especially that of E. S. Kennedy, to which 35 of the world’s leading scholars in the history of science contributed. Other such articles can be found in the major journals of the history of science, and in three journals specializing in the history of Islamic science, namely, the Journal of the History of Arabic Science (Aleppo), Zeitschrift für Geschichte der arabisch-islamischen Wissenschaften (Frankfurt am Main), and Arabic Science and Philosophy (Cambridge, UK). Also important are the proceedings of various conferences, notably those held in Aleppo in 1976, in Istanbul in 1977 and 1986, and in Tunis in 1989.

Astronomical handbooks with tables

A large component of astronomical manuscripts is tables for one purpose or another — the standard tools of the medieval astronomers. E. S. Kennedy has documented the astronomical handbooks with tables and explanatory text. His survey of Islamic zījes published in 1956 listed 125 of such works, and we now know of close to 200. A zīj deals with all the standard material of medieval astronomy and contains page after page of tables for calendrical conversion, computation of solar, lunar, and planetary positions, prediction of eclipses and visibility of the moon and planets, stellar coordinates and lists of localities with longitudes and latitudes. Families of zījes were popular in different regions, and often the parameters underlying the tables will have been updated. The parameters underlying a table and the computational errors it may contain facilitate relating one zīj to another, especially now that we have the electronic computer at our disposal and can generate in minutes tables containing thousands of entries that would have taken a medieval astronomer weeks to compute. The importance of this vast corpus of material will be obvious to any historian. I have a survey of Islamic tables for timekeeping and regulating the times of prayer ready for publication; these tables were not generally included in zījes, and they are as important for the study of Islamic religious practice as for the study of Islamic science.

Articles on Muslim scientists

In addition to the various bio-bibliographical works cited above, we should mention the articles on individual Muslim scientists that have appeared in the Encyclopaedia of Islam and the Dictionary of Scientific Biography.  These are usually written by the person most qualified to write on the scholar in question and contain useful bibliographies. The articles on Abū Maʿshar, al- Bīrūnī, Ibn al-Haytham, al-Karajī, al-Uqlīdisī, Sharaf al-Dīn al-Ṭūsī and al-Khayyām(ī) in the DSB by D. Pingree, E. S. Kennedy, A. I. Sabra, R. Rashed, A. Saidan, A. Anbouba, A. P. Youschkevitch, and

B. A. Rosenfeld (scholars from the United States, Egypt. Jordan, Lebanon, and the former Soviet Union) serve as outstanding examples.

Republishing medieval biographical and bibliographical works on Muslim scientists

A reprint of Gustav Flügelʾs 1871—2 edition of the Fihrist of Ibn al-Nadīm (Baghdad, c. 975) would be most useful. Unfortunately no reliable critical edition of theṬabaqāt al-umam of Ṣāʿid al-Andalusī (Toledo, c. 1065) is available yet (the Beiruṭ 1985 edition leaves much to be desired). It will be a long time before any scholar produces a new critical edition of the important biographical works by (Ibn) al-Qifṭī (Aleppo, c. 1230) and Ibn Abī Uṣaybiʿah (Damascus, etc., c. 1250); the Old Cairo editions are well worth reprinting (the former was reprinted by al-Muthannā in Baghdad, undated). Fonunately we now have a reliable edition of the Irshād al-qāṣid by Ibn al- Akfānī (Cairo, c. 1330), published by Jan Just Witkam in 1989.

Facsimile editions of important manuscripts

Whilst there is no substitute for critical editions of texts there are so many important primary sources and so few workers that is very useful for our subject to have faithful facsimile editions. Facsimile editions have been published recently in Frankfurt of the following works:

1. the unique Escorial manuscript of a recension of the Mumtaḥan Zīj of the astronomers of al-Maʾmūn (Baghdad, c. 830);
2. the unique Munich and Damascus manuscripts of the two halves of the treatise on astrology by Ibn Hibintā (Baghdad, c. 850);
3. good manuscripts of the treatise on the use of the astolabe and the book on the constellations by ʿAbd al-Raḥmān al- ṣūfi (Shiraz, c. 950);
4. a manuscript of the Algebra of Abū Kāmil (Egypt, c. 875);
5. the Istanbul and Leiden manuscripts of the Kitāb fī Ḥall shukūk kitāb Uqlīdis of Ibn al-Haytham (Cairo, c. 1025);
6. good manuscripts of the Jāmiʿ al-mabādiʾ wa-al-ghāyāt of al-Marrākushī (see above);
7. the unique Istanbul manuscript of the Kitāb al-Dalāʾil, a work on folk astronomy by al- Ḥasan b. Bahlūl (Syria, c. 990).

However, these facsimiles are not always faithful to the manuscripts. It is to be hoped that the criticism of reviewers of this series will be taken into consideration and that the deplorable practice of removing original folio numbers, ‘fixing up’ text missing because of worm-holes, touching up  illustrations to make them prettier, and merging fragments of a text from different manuscripts without indication will be abandoned henceforth.

It would be of great use to the field to have proper facsimile editions of the following works of major importance (by ‘best manuscript’ in the following list I mean the copy which is closest to the original):

1. the Istanbul and Berlin manuscripts of the astronomical handbook of Ḥabash al-Ḥāsib (M. Tt). Debarnot has published a detailed survey of the former, which is closer to the original);
2. the Escorial manuscript of the astronomical handbook of al-Battānī (for comparison with the published edition);
3. one of the Berlin manuscripts of the treatise on the construction of the astrolabe by al-Farghānī, which manuscript also contains the only known copy of two on the construction and use of the astrolabe by al-Khwārazmī;
4. the Leiden and Oxford and Paris fragments of the astronomical handbook of Ibn Yūnus (see above);
5. the best manuscript (to be determined) of the astronomical handbook of Kūshyār b. Labbān (Iran, c. 1000);
6. the Leiden manuscript (which is probably the best available) of al- Bīrūnīʾs treatise on the construction of the astrolabe;
7. the best manuscript (to be determined) of al-Bīrūnīʾs astronomical handbook (for comparison with the published edition);
8. the unique Istanbul manuscript of al-Bīrūnīʾs Taḥdīd (for comparison with the published version);
9. the best (Cairo?) manuscript of the important treatise on astrology entitled Rawḍat al-munajjimīn by Shāhmardān Rāzī (Rayy? c. 1070);
10. all available manuscripts of the treatise on mathematics by Yūsuf al- Muʾtaman b. Hūd, King of Saragossa (c. 1060) (to Jan Hogendijk and Ahmed Djebbar goes the credit for bringing this important work to the attention of colleagues);
11. the unique Paris manuscript of the astronomical handbook known as Dastūr al-munajjimīn, probably compiled in Alamut in the early twelfth century (a preliminary study of this important document has been conducted by F. Zimmermann);
12.  the Vatican or the London manuscript of the Zīj by al-Khāzinī (Merv, c. 1120), a work of major importance which has yet to be properly investigated;
13. the unique Oxford manuscript of the introduction to a zīj by Ibn al- Hāʾim (Andalusia, c. 1205), a key to the understanding of the zīj tradition of Andalusia;
14. the unique Hyderabad manuscript of the astronomical handbook of Ibn Ishāq (Tunis, c. 1195), the most important single astronomical work compiled in the Maghrib, discovered by chance during a visit to the Andra Pradesh State Library in 1978 and currently under investigation in Barcelona;
15. the best available manuscript (to be determined ----- possibly Cairo) of the Zīj-i Īlkhānī of Nasir al-Dīn al-Ṭūṣī (Maraghah, c. 1260);
16. The Topkapı manuscript of the ‘middle works’, minor Greek works which were translated and edited by Muslim scholars;
17. the Cambridge manuscript of al-Zīj al-Muzaffarī by Muḥammad b. Abī Bakr al-Fārisī (Aden, c. 1260), a work which is based on the lost Arabic works of al-Fahhād, now available in the Greek version published by David Pingree (see above);
18. the Cairo manuscripts of two Rasulid annual almanacs (Taiz 726/1326 and 808/1405), discovered during the cataloguing project in the 1970s (no other complete ephemerides that are this early have survived);
19. the Cairo manuscript of the treatise by the Rasulid Sulṭān al-Ashraf (Taʿizz, c. 1295) on the construction of the astrolabe and sundial and on the magnetic compass;
20. the Leiden manuscript of a Mamluk work on timekeeping called Kanz al-yawāqīt, currently being studied by Margarita Castells of Barcelona;
21. the best manuscript (to be determined) of the astronomical handbook of Ibn al-Bannāʾ, of which only the introduction has been published (see above);
22. the best manuscript (to be determined) of the Shāmil Zīj, the main astronomical handbook that was used for several centuries in what is now Iraq and the border regions of Southern Turkey;
23. the fine Dublin manuscript of the corpus of tables for timekeeping and regulating the times of prayer that was used in medieval Cairo; the Paris manuscripts of two versions of al-Zīj al-Muṣṭalaḥ, the most popular astronomical handbook in medieval Egypt;
24. the unique Dublin manuscript of the treatise on astronomical instuments by Ibn al-Sarrāj (Aleppo, c. 1325);
25. the Oxford manuscript of the astronomical handbook of Ibn al-Shāṭir (Damascus, c. 1365);
26. the Paris manuscript of the corpus of tables for timekeeping and regulating the times of prayer by Shams al-Dīn al-Khalīlī (Damascus, c. 1365), used in Damascus until the nineteenth century.

The list goes on, and I have not even reached the Zīj of Ulugh Beg. Also I am well aware of the need for an accessible version of the Arabic translations of the Almagest of Ptolemy and the Elements of Euclid. George

Saliba is contemplating editing the former, but in the case of the latter there is no hope for a properly edited text, so a facsimile edition would certainly be very useful. An indication of the difficulty of preparing critical editions of such works is provided by the complexity of Paul Kunitzschʾs recent edition of the Arabic version of Ptolemy’s star catalogue.

It would give our subject an enormous boost if various manuscripts of a number of short treatises (majāmīʿ or Sammelbände) could be made available in facsimile editions.  These are:

1. MS Bankipore 2468 (surveyed by Ahmed S. Saidan and Jan Hogendijk);
2. MS Istanbul Aya Sofya 4830 and 4832 (uncatalogued, some of the contents are listed by M. Krause);
3. MS Cairo Dār al-Kutub MR 40 and 41 (catalogued by the present writer), two Ottoman collections of Arabic recensions of Greek treatises (some not available elsewhere) and early Islamic treatises on astronomy and mathematics;
4. MS Paris B.N. ar. 2457 (catalogued by MacG. de Slane), supposedly in the hand of al-Sijzī (c. 970);
5. MS Oxford Bodleian Marsh 633 (catalogued);
6. MS Damascus 4871 (surveyed by Jamil Rageb and E. S. Kennedy);
7. MS Princeton Yahuda 373 (catalogued by Rudolf Mach), an Ottoman compendium of treatises on instruments.

All but the second are catalogued or described in brief in the secondary literature. All but the last contain early works on astronomy and mathematics which have not been properly studied. Each one is a goldmine for researchers, but it is often tiresome to acquire microfilms and a great trial to use them. To have these volumes available on one’s own bookshelf would be a dream. In each case the best-qualified scholar could be selected to write an introduction.

Illustrations in Islamic scientific manuscripts

At the conference on “The Islamic Book” organized in Washington, DC, in November 1990, by Dr George Atiyeh, I addressed the topic of illustrations in Islamic scientific manuscripts. The study of these has barely begun. Suffice it to say that no more than twenty extracts from Arabic manuscripts were contained in John E. Murdoch’s Album of Science — Antiquity and the Middle Ages (1984), richly illustrated with over 500 plates. Credit must be given to S. H. Nasr for providing his book on Islamic science prepared for the Festival of Islam in London in 1976 with a superb set of illustrations from Islamic scientific manuscripts, many reproduced in colour. My own

survey of the Cairo manuscripts contains some 240 more mundane extracts, many illustrations of planetary models or instruments, and other examples of different kinds of tables, handwriting of individual astronomers, and various items of historical interest. There is an obvious need for a companion volume to the album mentioned above which would feature illustrations in Islamic manuscripts, let alone for art-historical investigations of different categories of illustration.

Some important astronomical instruments

Instruments are another important source for our knowledge of the development of Islamic science, and we are now in a much stronger position than we were recently to evaluate the astrolabes, quadrants, globes, and sundials of the various schools in the Muslim world, dating from the ninth to the nineteenth century. The majority are housed in museums, but a substantial minority are in private collections. Historical instruments already command very high prices at auctions.

I have recently published a lengthy and richly illustrated article on Islamic instrumentation; although only an Italian version is available in print, the English original will also be published soon. For the time being L. A. Mayer’s book on Islamic astrolabists and Emilie Savage-Smith's catalogue of Islamic globes serve to locate many of these items. The forthcoming Répertoire of Brieux and Maddison and my Catalogue (both mentioned above) are guaranteed to cause an upsurge of interest in Islamic instruments.

A collection of original instruments is outside the scope of most academic organizations, but for teaching and display purposes a collection of facsimiles can be worthwhile. Modern techniques enable facsimiles of historical instruments to be made at reasonable prices, and there are various commercial concerns which undertake this. The Institute of Arabic-Islamic Science in Frankfurt has had prepared some 200 facsimiles of Islamic instruments, but this private collection is not to the general public.

The following are some of the most important Islamic astronomical instruments:

1. the astrolabe of Nasṭūlus dated 315/927 or 928 and representative of the earliest Islamic astrolabes (Kuwait, Dār al-Āthār al-Islāmiyyah);
2. the spectacular astrolabe of the well-known astronomer al-Khujandī dated 374/984 or 985 and representing the culmination of Abbasid instrument-making (private collection);
3. the spectacular astrolabes of ʿAbd al-Karīm al-Miṣrī dated 638/1240 or 1241 ? (London, British Museum), and of Jalāl al-Kirmānī dated 830/1426 or 1427 and dedicated to Ulugh beg (Copenhagen, David Collection);
4. the globe by Muḥammad son of the astronomer Muʾayyad al-Dīn al-ʿUrḍī, made in Maraghah in the thirteenth century (Dresden, Mathematisch-Physikalischer Salon);
5. the large astronomical clock in the Mosque of the Qarawiyyīn in Fez, constructed in 717/1317 to replace an earlier one (only the clock-face remains: the driving mechanism has disappeared); 830/1426 or 1427 and dedicated to Ulugh Beg (Copenhagen, David collection);
6. the remarkable astrolabe of Ibn al-Sarrāj of Aleppo, dated 729/1328 or 1329, which can be used for any terrestrial latitude in five different ways and which is without doubt the most sophisticated astrolabe ever made (Athens, Benaki Museum);
7. the highly complex sundial of Ibn al-Shāṭir, made in 773/1371 or 1372 for the Umayyad Mosque in Damascus (Damascus, Dār al-Āthār [fragments of the original] with a faithful nineteenth-century reconstruction still in situ in the mosque);
8. the remarkable cartographic qiblah indicator made in Isfahan c. 1700, with which the qiblah can be found for any locality between Spain and China directly from the map (private collection).

These e are just a few of over one thousand Islamic instruments that survive, a small fraction of those that were actually made. Of particular historical interest are those which are mentioned in contemporary treatises. Examples of these are the universal astrolabe of Ibn al-Sarrāj, which in the fifteenth century came into the possession of the Egyptian astronomer ʿIzz al-Dīn al-Wafāʾī. The latter complained that his illustrious predecessor had not written a treatise on the use of the instrument, so this he proceeded to undertake himself. Before the discovery of al-Wafāʾīʾs treatise it was by no means obvious how some of the component parts functioned. Or the astrolabe signed by the Rasulid Sulṭān al-Ashraf in 690/1291: al-Ashraf authored a book on the construction of the astrolabe, and appended to this work are some remarks by his teachers on all of the astrolabes which he made. Amongst these we can recognize the sole surviving one, now in the Metropolitan Museum of Art in New York. But many important instruments are described only in texts. Richard Lorch is currently editing all available texts on non-standard astrolabes.  These were invented in the ninth and tenth centuries, and we have treatises on them by a series of Muslim scholars, notably al-Sijzī, al-Bīrūnī, al-Marrākushī and Ibn al-Sarrāj, but not one single instrument of this kind has survived the ravages of time. Al-Sijzī actually mentions the patrons for whom the different kinds of newly- invented astrolabes were made.

Collected works

The collected papers of several colleagues of past and present generations are now available in reprint volumes. C. A. Nallinoʾs widow brought out his unpublished papers on astronomy, astrology, and geography in 1944 (a reprint of these would be worthwhile.) The numerous papers of J. Millás Vallicrosa and Juan Vernet have also been reprinted. Georg Olms Verlag published some of the collected papers of Eilhard Wiedemann (2 vols.) and Willy Hartner (2 vols.). Thanks to the Frankfurt-based Institute of Arabic- Islamic Studies we have the collected papers of various scholars of the nineteenth and early twentieth century: Franz Woepcke (2 vols.), Carl Schoy (2 vols.), Heinrich Suter (2 vols.), and Eilhard Wiedemann (3 vols. of papers not reprinted by Olms). The works of Paul Luckey, Max Krause, and H. P. J. Renaud merit the same treatment. The same Institute has recently published six volumes of early studies on Islamic astronomical instruments; whilst many of these are quite out-of-date and some have been superseded, the volumes are of great value and will save researchers many hours of work collecting photocopies of nineteenth-century journals. The American University of Beirut paid tribute to its professor emeritus E. S. Kennedy by reprinting some 75 papers either written or inspired by him (he published mainly books) in one large volume. Thanks to the British publisher Variorum we have reprints of the more recent papers of Bernard R. Goldstein (1 vol.), D. A. King (2 vols. and a third in press), Paul Kunitzsch (1 vol.) and Franz Rosenthal (3 vols.). The works of David Pingree and Julio Samsó are planned for the same series.

There are several works on aspects of Islamic science which merit reprinting and a few which have been reprinted. One example is Aydin Sayılı’s The Observatory in Islam: written over thirty years ago this work is still unsurpassed, but it is difficult to obtain, and despite the fact that it was reprinted once in 1981 it merits a much wider audience than it has enjoyed hitherto. Also Kennedy’s 1956 zīj survey, long out of print, has recently been reprinted by the American Philosophical Society.

Concluding remarks

This is an exciting field in which to work. Not least does one soon learn a healthy respect for the scientists of medieval Islam, but there is also a succession of modem scholars of diverse backgrounds and nationalities who have laid a clear path for us to follow and set standards which we should constantly strive to uphold. Also one is not working alone but enjoys the Privilege of working with a particularly collegial group of fellow-researchers in different parts of the world.

The most difficult task facing us is the editing of the texts, and it is also the most important in the long term. Some of us shy away from it. As a graduate student I had a professor with some understanding of the difficulties of editing texts who advised me: ‘It doesn’t matter what text you work on, just make sure there’s only one manuscript!’ Fortunately most of the scholars mentioned above have been more bold.

Acknowledgements: This final version of my paper has benefited from the corrections and additions of various colleagues. Particular thanks go to Dr Jan Hogendijk (Utrecht), Prof Henry C. King (Little Chalfont, UK), Prof Paul Kunitzsch and Dr Richard Lorch (both of Munich), Prof George Saliba (New York), and Prof Julio Samsó (Barcelona), without in any way involving them in any remaining omissions or oversights.