- 120,000 GBP





- 120,000 GBP


brass, with frame and scale, the legs and rod later replacements

diameter of globe: 23.3cm.

height of frame: 22.9cm.

total height: 33.5cm. approx.

In reasonable overall condition, some staining to globe and some engraved circular markings faded due to wear consistent with age, rod and legs later replacements, as viewed.

"In response to your inquiry, we are pleased to provide you with a general report of the condition of the property described above. Since we are not professional conservators or restorers, we urge you to consult with a restorer or conservator of your choice who will be better able to provide a detailed, professional report. Prospective buyers should inspect each lot to satisfy themselves as to condition and must understand that any statement made by Sotheby's is merely a subjective, qualified opinion. Prospective buyers should also refer to any Important Notices regarding this sale, which are printed in the Sale Catalogue.


This elegant and well-constructed celestial globe bears the hallmarks of a Safavid Iranian provenance, which, in the context of instrumentation, means late seventeenth-century Isfahan. Although unsigned and undated it looks as if it may be the handiwork of one of the leading instrument-makers of Isfahan, which, along with Lahore, was a leading place for instrument-construction in the entire Muslim world at the time (Instrument-makers in the Maghrib were equally active, particularly in Fez and Marrakesh, but their productions lacked the variety and technical sophistication of the Isfahan and Lahore instruments). Not only were the instruments of the Isfahan school made with extreme accuracy but the inscriptions were always elegantly engraved with a distinctive style of Arabic or Perso-Arabic.


No encompassing study of the scientific works of art produced in Safavid Isfahan has been conducted. What is necessary and eminently feasible would be a catalogue of Safavid instruments, such as the larger survey and catalogue and survey which Prof. S. R. Sarma has prepared for the instruments of the ‘rival’ Lahore school (See Prof. S. R. Sarma’s online catalogue of Indian astrolabes: The dozens of available Safavid instruments have not fared well in the documentation of the history of Islamic astronomy. R. T. Gunther’s monumental Astrolabes of the World (1932) featured only one Safavid astrolabe already described by W.H. Morley in 1856, which happened to be one of the most remarkable (see Morley’s description of a planispheric astrolabe constructed for Shah Sultan Husayn Safavi in the book’s introduction). L.A. Mayer’s Islamic Astrolabists and their Works (1956), alas organised alphabetically by makers’ names, mentioned a few Safavid makers and perhaps a dozen instruments (L.A. Mayer, Islamic Astrolabists and their Works). The long-awaited Répertoire of Islamic instrument-makers (by Alain Brieux and Francis Maddison, awaiting publication), will surely document numerous Safavid instrument-makers and their works. The surviving Safavid instruments have been virtually ignored in the history of Islamic art and what is available in published form are a series of technical descriptions of individual instruments in certain museum or exhibition catalogues. Most of the Safavid instruments are astrolabes, but there are also sundials and ­three brass Mecca-centred world-maps preserving direction and distance to the centre (rediscovered in the 1980s and ‘90s) of a kind previously unknown to the history of cartography. The analysis of these instruments resulted in the first listing of close to twenty Safavid makers and their instruments (See D. King, World-Maps for finding the direction and distance to Mecca, 1999, pp.263-9 (includes considerable information on seventeenth-century Iranian astrolabes, especially on the geographical gazetteers with qibla values on their maters)).

The instrument-makers who come into consideration as potential makers of this piece from the point of view of the engraving are Muḥammad Muqim Yazdi Muḥammad Mahdi al-Khadim al-Yazdi (L.A. Mayer, Islamic astrolabists, pp.74-75, and 70-71, see also lot 128). On the other hand, one would not expect either of these first-class craftsman to produce a celestial globe of the basic variety, with neither constellations nor stars. Their astrolabes have unexpected features such as plates bearing exquisite representations of the constellation figures and principal stars, mathematically-sophisticated graphs of the solar altitude in the direction of Mecca for different localities throughout the year, and historically-important information about the geographical coordinates and qiblas of hundreds of cities.

It is the distinctive engraving which enables us to ascertain the provenance of the previously-unpublished globe presented here. We may compare certain letters on this sphere to the corresponding ones on a Safavid astrolabe such as the one in the British Museum and a Safavid globe from Mashhad in the Nasser D. Khalili Collection, London, or various Safavid astrolabes now in the Greenwich National Maritime Museum. On each of these we find a bold, elegant naskh script for the names of the zodiacal signs, sometimes pointed (with vowels or other critical markings) and a similar smaller script for the other inscriptions, as well as peculiarities such as an almost circular separate nun (for fifty).

Emilie Savage-Smith estimates that some two hundred and eighty-five Islamic celestial globes survive (E. Savage-Smith, article ‘Globes’ in Enc. Islam, III), although we can add that the number of globes surviving from the formative period of Islamic science up till c.1500 does not exceed ten. Furthermore, numerous globes, even some in major museums, were not made for any serious scientific purpose.

The number of surviving Safavid globes is remarkably small, especially when compared with the plethora of such instruments from seventeenth-century Lahore (Sarma, Catalogue of Indian instruments). Of globes from Safavid Isfahan we can point only to the one here featured. The only other individual Safavid globes are not Isfahan but from Mashhad, such as the one signed by the astronomer Muḥammad Zaman, who was productive from 1640-80, which is now in the Nasser D. Khalili Collection in London (Savage-Smith, Islamic Celestial Globes and Related Instruments, pp.248-9, no.143). His instruments were different in style and engraving from those of his Isfahan contemporaries. As with his surviving astrolabes, this globe by Muḥammad Zaman is elegantly and accurately executed; it also features some forty stars.

Emilie Savage-Smith has identified four types of Islamic globes. The first type displays the outlines of the forty-eight constellations and approximately a thousand and twenty-two stars. These are the largest and the most elaborate artefacts. Those in the second category have no constellation outlines and only a selection of the most prominent stars, usually between twenty and sixty.

The third type of globe, as witnessed here, has neither constellation outlines nor stars, but only the great and lesser circles – the ecliptic, equator, tropic, and polar circles, each of which is labelled. The latter design is not mentioned in any written source, although the engraving and construction of the globes suggest that the design may have originated in Iran in the late seventeenth or early eighteenth century.

(Savage-Smith points to a fourth form of celestial globe which appears in nineteenth-century India. This served astrologers as an aide-memoire for their art; only a few constellations were depicted, most of the surface area being covered with astrological information inscribed in Arabic, Persian, and Urdu.)

The globe and its markings

The inscriptions are engraved in a clear, bold and elegant naskh. The engraved text has been made clearly legible by insertion of black ink, which only in one inscription has loosened.

The globe bears three main sets of markings. In this description we use the standard geometrical notation for circles on a sphere: those circles centred on the centre of the globe are called ‘great-circles’; those parallel to great-circles are called ‘small-circles’. Essentially, the markings on the sphere concern the celestial equator and the ecliptic, two great-circles, the latter inclined to the former at an angle known as the obliquity of the ecliptic. This angle in ancient and medieval times decreased from about 24° to 23°30’ (Paul Kunitzsch, article ‘Mayl [= obliquity of the ecliptic]’ in Enc. Islam, 2nd. Edn.). The standard value among Safavid astronomers would have been 23°35’, but instrument-makers were more conservative and it was in any case easier to use 24°. This is the value of the obliquity used here on the Safavid globe for convenience.

The numerals on the scales are in the Arabic alpha-numerical notation called abjad (a = 1, b = 2, etc.), which was in standard use in Islamic astronomy and instrumentation.

The primary markings are the great-circles for the celestial equato da’irat muʿaddil al-nahar, literally, ‘the circle where day equals night’, and the celestial poles qutub shamali and qutub janubi, ‘north and south (celestial) poles’, as well as quṭb muʿaddil al-nahar, ‘the pole of the celestial equator’ at the northern pole. The scale of the equator runs from 0° to 360°, with labels for each 5° subdivided in single degrees. The scale is defective between 0° and 30°, for reasons that are not clear.

Circular holes of about 1.2cm diameter have been excised at each pole, with evidence that they have widened over the years due to the consistent friction of the sphere spinning. No other holes for plugs are visible on the surface of the sphere.

A pronounced pair of perpendicular great circles pass through the poles (or rather, the centres of the circles representing them). One of these is unmarked. The other bears a 360°-scale divided into labelled 5°-intervals subdivided into single degrees. On this scale are the poles of the ecliptic (see below), at 24° from the poles of the celestial equator.

The second principal markings are for the ecliptic, falak al-buruj, the apparent path of the sun against the background of fixed stars. The ecliptic here is inclined at an angle of 24° (mayl kulli) to the celestial equator, da’irat muʿaddil al-nahar. The poles of the ecliptic lie on the circle perpendicular to the celestial equator, which circle is appropriately labelled da’ira marra bi-aqṭab arbaʿa, ‘circle passing through the four poles’.

The points where the ecliptic intersects the celestial equator are unlabelled for the vernal equinox, nuqṭat iʿtidal kharifi, ‘autumnal equinox’. The points where the ecliptic intersects the celestial colure are labelled nuqṭat inqilab ṣayfi, and nuqṭat inqilab shitawi, ‘summer and winter solstices’. The distances of each solstice from the equator, 24°, are marked on the colure scale, mayl kulli, ‘maximum (solar) declination’, or simply ‘obliquity of the ecliptic’. A great-circle labelled da’irat al-mayl, ‘circle of declination’, between the north celestial pole and the celestial equator (if not beyond), seems to simply indicate that distances perpendicular to the equator are called declination.

The ecliptic itself is divided in 12 equal sections for the zodiacal signs, each 30° being labelled with the signs of the ecliptic

al-ḥamal – al-thawr al-jawza’ al-saraṭan ­– al-asad – al-sunbula

Aries ­– Taurus – Gemini – Cancer – Leo – Virgo

north of the celestial equator and

al-mizan – al-aqrab – al-qaws – al-jady – al-dalw – al-ḥut

Libra – Scorpio – Sagittarius – Capricorn – Aquarius – Pisces

to the south of it. The spaces for the signs are divided into labelled 5°-sections subdivided into degrees. The signs represent the artificial and to some extent arbitrary divisions of the ecliptic conceived in Antiquity; in the present context they have nothing to do with astrology (Paul Kunitzsch, article ‘Mintakat al-burudj’ in Enc. Islam, 2nd edn.).

Perpendicular to the ecliptic at the end of each of the first and second signs in each quadrant are lightly-engraved great-circles which meet at the poles of the ecliptic. The first of the eleven great-circles through the vernal equinox is labelled khaṭṭ niṣf al-nahar, ‘line of midday’ or ‘meridian’.

The third set of markings are lightly engraved small circles which relate to one or other or both of the celestial equator and the ecliptic. One has to imagine the ensemble rotating about the celestial axis, so that certain points execute what are called madars in Arabic, usually rendered ‘day-circles’ but more precisely indicating ‘circles of rotation’.

Two small-circles parallel to the celestial equator passing through the solstices are labelled madar ra’s al-saraṭan, and madar ra’s al-jady, ‘the day-circles of the first points of Cancer and Capricorn’. There are also two sets of unlabelled small-circles at distances (declinations) 10° and 20° north and south of the equator, inside the solstitial circles for obliquity 24°. (The circle for declination 10° north features slight but noticeable errors).

Five small-circles around the north celestial pole have radius 5° unlabelled; 10°, unlabelled; 21°, unlabelled; 24°, labelled madar quṭb al-buruj, ‘the circle of rotation of the pole of the ecliptic’; 26°, unlabelled. Around the south ecliptic pole is a small-circle of radius 30° labelled da’irat ʿarḍ, ‘the circle of rotation of the latitude’, whose significance (if any), is unclear. Four small-circles around the south celestial pole are at 7°, unlabelled; 16.5°, labelled madar yawmi, ‘daily circle of rotation’; 23°, madar quṭb al-buruj, ‘the circle of rotation of the pole of the ecliptic’; and 28°, unlabelled. Around the south ecliptic pole is another small-circle of radius 30° labelled da’irat ʿarḍ, ‘the circle of rotation of the latitude’, as above.

The brass rod of 25.7cm has flattened ends which enable the globe to be placed in any orientation on the circular scale on the flour-legged frame. This brass frame appears to be a later replacement, although the brass annulus of width 2cm on top of it seems to be original. This is engraved with a scale of 360° divided into labelled 5°-intervals subdivided into single degrees. The forms of the numerals are mainly those used by Safavid instrument-makers but the maker betrayed himself in certain places, as with the fa’ for ‘80’ at 280°, which has the dot in the middle of the bowl of the letter. The frame bears no trace of ever having supported a vertical semi-circular scale which would have enabled the ensemble to be used to display the rotation of the globe with respect to the horizon for a specific latitude.