Al-Jazari's Elephant Clock, 1206
The Islamic Golden AgeTwenty-five years at an Artuqid court
Around 1180 a craftsman named Badi' al-Zaman ibn al-Razzaz al-Jazari entered the service of the Artuqid rulers of Diyar Bakr, whose seats were Amid - today's Diyarbakir - and Hisn Kayfa, today's Hasankeyf. He stayed roughly twenty-five years, working for Nur al-Din Muhammad, then his son Qutb al-Din Sukman, then Nasir al-Din Mahmud, who ordered him to write down everything he had made. The result was a manual of fifty machines sorted into six categories, in al-Jazari's own count. Even its date is unsettled: surviving copies disagree about when his service began, one giving 575 of the Hijra and others 577, which places the finished book in 1204 or 1206. Most editions print 1206, and the argument is unresolved.
What happens every half hour
The clock is a wooden elephant with a water tank in its belly. A bowl pierced by a small hole floats in that tank and takes roughly half an hour to fill and sink, and as it descends it drags cords that walk a scribe's pen around a dial to mark the minutes. When the bowl bottoms out, the cords trip a release in the castle on the elephant's back. A falcon drops a ball from its beak into the mouth of a dragon; the ball's weight swings the dragon down on its axle, and that swing hauls the drowned bowl back to the surface; the dragon tips back and discharges the ball into a vase; the mahout brings his mallet down on a cymbal while a bird on the cupola whistles. Then the bowl sinks again. The rate is tunable by hole and bowl, as it had to be: an hour then meant a twelfth of daylight and stretched in summer.
Almost nothing in it is new
Almost every part of that has a genealogy. The perforated sinking bowl is an Indian timekeeper, the ghati, attested in Sanskrit sources centuries earlier. Float regulation of flowing water belongs to a Greek and Hellenistic tradition that Baghdad had been translating and extending since the ninth century. Al-Jazari names his own debts: the Banu Musa brothers of Baghdad, whose book described roughly a hundred devices; a monumental water clock he credits to Archimedes, from a text scholars now judge pseudonymous; al-Saghani on candle clocks; Hibatallah ibn al-Husayn, who died in 1139, on musical automata. He is careful about what he improved and what he simply took. This era received as hard as it transmitted, and said so in its own footnotes.
The reading that is not his
You will often read a symbolic decoding of the machine - the elephant for India, the dragons for China, the phoenix for Persia in most versions and for Egypt in others, the water-work for Greece, the turban for Islam. The decoding is fair as a description of where the parts came from, but it is a modern reading and it is not in al-Jazari's treatise. His prose is a foreman's: hole sizes, cord lengths, how to solder a tank so it does not weep. He is careful to say what he improved and what he simply took, but he offers no interpretation of the figures at all. The borrowing is genuine; the sermon about it is ours. That is the better fact, because it means the mixing was ordinary enough to need no comment.
Copies, translators and rebuilt clocks
No original machine survives, and we cannot be certain this one was ever built. Copies of the book are what survive: the earliest, dated 1206, in the Topkapi library in Istanbul, with others in Oxford, Leiden, Paris and Dublin, plus loose illustrated leaves in museums. The Metropolitan Museum's elephant-clock leaf was copied in 1315 by the scribe Farrukh ibn Abd al-Latif, probably in Syria or Iraq, nearly sixty years after the Mongols took Baghdad; the copying did not stop. Donald Hill, an engineer and a historian, translated the book in 1974 and had to redraw much of it, since the diagrams are schematic and inconsistent about sizes and placement. The rebuilt clocks in Dubai, Le Locle and at King Abdullah University in Saudi Arabia run, and a touring replica drew crowds at the Science Museum in London in 2010 - each one embedding a modern reading of an ambiguous drawing.
Key Dates to Remember
- c. 1136 — Al-Jazari is born in the Jazira, upper Mesopotamia. The town is not recorded and the birth year is conventional rather than firmly documented.
- c. 1180 — He enters Artuqid service; his own book dates this to 575 or 577 of the Hijra depending on which copy you read.
- 1206 — The Book of Knowledge of Ingenious Mechanical Devices is completed for Nasir al-Din Mahmud; the oldest surviving manuscript carries the same year.
- 1315 — The scribe Farrukh ibn Abd al-Latif finishes an illustrated copy, probably made in Syria or Iraq, the source of the elephant-clock folio now in the Metropolitan Museum.
- 1974 — Donald R. Hill publishes the standard English translation, working mainly from a Bodleian manuscript and redrawing the ambiguous diagrams.