At a glance
- Region
- al-Andalus & Europe

Quick facts
- Domain
- Mechanics, automata and clocks
- Kind
- Device
- First attested
- 1137 CE · 531 AH† (derived)
- Era
- Middle period
- Where
- Toledo
- Attribution
- Contested
Who made it
- Al-Zarqali Improver
The claim and its basis
The basins are attributed to Abu al-Qasim Abd al-Rahman known as al-Zarqali in the earliest detailed account; but the whole tale is a transmitted wonder-report with no described mechanism and no surviving remains, and the destruction accounts multiply, so the record documents the transmitted attribution, not a verified invention.
The Granadan geographer al-Zuhri (d. 1137) describes the pair as Toledo's marvels and al-Maqqari preserves his account adding the Rabat manuscript note identifying Abd al-Rahman as al-Zarqali and the Indian talisman story; Norman Roth analyses the transmission and notes the Jewish astrologer's name differs between al-Zuhri and al-Maqqari plus a fabricated attributed poem; the 2023 Kān study calls the pair an engineering puzzle later technology failed to explain.
Sources: Nafh al-Tib min Ghusn al-Andalus al-Ratib — al-Maqqari · 'Water clocks in Islamic Toledo' — Norman Roth, MEAHHEBREO · 'The Andalusian lunar water clock' — CIHAR repost of the ABC report · 'Baylatan of Toledo: the marvel of Andalusian water-fountain engineering' — Douria Kan al-Tarikhiyya · Toledo and the Tagus river photograph — Wikimedia Commons
Before it
- The water talisman of Arin in the land of India reported by al-Mas'udi
- The Andalusian fountain and water-clock tradition before Toledo
How it works
Summary
The "Baylatan of Toledo" is a pair of large water basins whose making is attributed to Abu al-Qasim Abd al-Rahman known as al-Zarqali, set in a hollow chamber on the bank of the Tagus at the Tanners' Gate in Toledo. They are reported to have filled and ebbed with the waxing and waning moon until they were full on the night of the fifteenth and empty on the night of the twenty-ninth, swallowing added water and restoring what was removed by themselves 1. The earliest detailed testimony that reached us is that of the Granadan geographer al-Zuhri, who died in 1137, preserved by al-Maqqari in the Nafh al-Tib together with the Rabat manuscript note identifying Abd al-Rahman as al-Zarqali 2 1. No mechanism is described today and no remains survive; the accounts of what ruined them multiply, so the museum records the transmitted tale and its attributed maker, not a verified invention 4 2.
What it is
Al-billitan — with the vowels the text's gloss gives — is in usage the name of a fountain basin, sometimes paralleled by the word khassa, as al-Maqqari explains in his gloss 1. The subject is therefore not a machine more complex than two basins, but a pair of large basins placed in a hollow chamber inside the Great River — the Tagus — at the place known as the Tanners' Gate outside Toledo, filling and ebbing on a system that followed the lunar month 1 3. No read source describes an internal mechanism, channels, dimensions or the material they were made of; what we hold is a behavioural description of an effect, not of a construction 4. The museum therefore records a device of transmitted description: the name, the site and the reported behaviour, not proven engineering.
The world before it
Al-Zuhri, who describes the basins, was a Granadan geographer who died in 1137 and lived in the al-Andalus environment where people knew water clocks and fountains; the modern Kān study notes that fountain-making in Islamic al-Andalus witnessed notable progress attested by the engineering works the sources preserve 2 4. A reposted Spanish report describes Toledo's standing archaeological "challenge" as proving the existence and location of the water clock al-Zarqali built on the Tagus banks, and notes that medieval scholars called it marvellous, surprising and without parallel in the world, yet "its fame oscillated between reality and legend" for want of reliable remains 5. Centuries before Toledo the Islamic world knew water-driven timekeepers, but no read source names a specific earlier device that served as the basins' model; the possible link to the Andalusian water-clock tradition is a general context, not a proven lineage 4. Inside the tale itself there is a precursor of another kind: the talisman of the city of Arin in the land of India that "turned with its finger from dawn to sunset," which al-Maqqari says al-Mas'udi mentioned — the transmitted inspiration the reported maker is said to have heard of 1. That talisman is a narrative precursor, not a documented device: we have not read al-Mas'udi's own text and know neither its form nor its maker, so it is not entered as a proven precursor.
The story
Al-Maqqari relates in the chapter "Wonders of al-Andalus" that the two basins were made by Abd al-Rahman — the Rabat manuscript's marginal note identifies him as "Abu al-Qasim Abd al-Rahman al-Zarqali" — when he heard of the Indian talisman; he made these two basins outside Toledo in a hollow chamber inside the Great River at the Tanners' Gate 1. He describes what were called their marvels: they filled and ebbed with the moon's waxing and waning by half a seventh each day and night; at first a little water appeared when the crescent rose, by morning a quarter-seventh was in them, by day's end half a seventh was complete, and so it continued between day and night until seven days and seven nights of the month had passed and they held half full, then the increase of half a seventh each day and night continued until they were completely full at the moon's completion on the night of the fifteenth 1. When the fifteenth night came and the moon began to wane, they decreased with its waning by half a seventh each day and night, and on the twenty-ninth of the month no water remained in them 1. Most marvellous of all, al-Maqqari reports, they "swallowed" water if it was drawn for them during the waning until nothing remained but what had been in them at that hour, and they "filled again at once" if emptied during fullness; he says they were "more amazing than India's talisman" because that one stands at the equinoctial point where night does not exceed day, whereas these two were not in the equinoctial place 1.
Then comes the ending: they remained in one chamber until the Christians took Toledo; al-Funj — Alfonso — wanted to learn their motion, ordered one of them lifted to see whence the water came and how the motion worked, and when it was lifted both stopped working, in the year 528 AH 1. Al-Maqqari then gives two further versions under "it was also said": one makes their ruin the work of the Jewish astrologer Hatim ibn Rabuwa — in the Rabat manuscript "Hatim ibn Rabuwa the Jewish astrologer" — who brought all al-Andalus's pigeons to Toledo in a single day in 527 AH and promised to make them fill by day and ebb by night, but when one was lifted he could not restore it; the other says one was lifted to steal its craft and stopped, while the other kept giving its motion, and al-Maqqari closes with "and God knows the truth of the matter" 1. Norman Roth notes that al-Zuhri's own account names the astrologer "Hamis b. Zabara," not "Hunayn"; that the pigeon story dated 527 AH tells of his predicting a Christian king's entry into Cordoba — which did occur under Alfonso VII in 1146 — and that the Ibn Zabara Jewish family is documented in twelfth-century Toledo 2. Roth adds a methodological warning: a poem supposedly by Ibn Ezra titled "Marble, the work of al-Zarqal" on these water clocks is in fact the fabrication of an unreliable nineteenth-century Jewish writer and cannot serve as evidence 2. There is also a locational difference between the accounts: al-Maqqari places the basins in "a hollow chamber inside the Great River," while al-Zuhri's version as reposted describes "two great ponds in a house on the outskirts of Toledo on the banks of the Tagus, not far from the Tanners' Gate" — a difference of detail that cannot be silently reconciled 1 3.
An internal chronological problem also deserves recording: the tale keeps the basins in "one chamber" until "the Christians took Toledo," then dates the uprooting to 528 AH — so the text itself has the device working for many years after the city's transfer to Christian rule before its lifting, an arrangement scholars read with caution that alone proves no construction or sabotage date 1 2.
The maker(s)
The attributed maker in the account is "Abd al-Rahman," whom the Rabat manuscript's marginal note matches to "Abu al-Qasim Abd al-Rahman al-Zarqali" — the same Ibrahim ibn Yahya al-Zarqali also known in Latin as Arzachel 1 2. Al-Zuhri's account names him "Abu al-Qasim ibn Abd al-Rahman known as al-Zarqil" in the description reposted from a Spanish press report 3. This man was of Toledo itself: he grew up there as an instrument-maker and engraver — the laqab "al-Naqqash," "the engraver," being the origin of his Latin name Arzachel — and worked from 1048 in Sa'id al-Andalusi's team for al-Ma'mun of Toledo and then al-Mu'tamid of Seville, dying in 1100 on one account 1. The museum records al-Zarqali's role as "improver," not "inventor": the tale itself describes him as having heard of the talisman and made "its like," that is, he is said to have arranged a transmitted design rather than the accounts crediting him with creation from nothing 1. The Jewish astrologer Hamis b. Zabara — or Hunayn in al-Maqqari's version — is not entered as maker or partner; his role in the accounts is that of a supposed saboteur or restorer, he is not Muslim, and the museum's rules do not make him a maker but keep his documented role in the tale's history 2 1. Alfonso — al-Funj — appears as ordering one basin lifted to learn its motion, a dismantling ruler's role, not a maker's 1. No read source names an Andalusian patron who ordered or financed the making, so the commissioning authority and financier are unknown in the text 1.
How it works
No read source describes the mechanism; what we know is the claimed behaviour, not the construction 4. In al-Maqqari's account — which in substance is al-Zuhri's — water began to flow into the basins "through invisible channels" as soon as the new crescent appeared, reaching half a seventh of the water by dusk the next day, increasing by night and day at a rate of half a seventh every twenty-four hours; the basins were half full at the week's end and completely full to overflowing the following week 3 1. From the night of the fifteenth, when the moon begins to wane, the water decreased at half a seventh per day, until on the twenty-ninth nothing remained 3 1. Most remarkable in the description is that the basins "swallowed" excess water or returned what was removed by themselves to the correct measure — a self-correcting property no text that reached us explains 1. The modern Kān study tries to interpret them as an engineering puzzle, saying they relied on mechanical filling and emptying tied to the days of the lunar month, but its proposed explanation remains a hypothesis, not a confirmed analysis 4. A reposted Spanish press report offers a wholly different archaeological hypothesis: a strange hydrological-cycle spring in the Las Tinerías area near the Arab baths of Yuso may have supplied the basins with water through the groundwater, yet excavations have not proven the link and the scientists themselves stated that reliable remains are absent 5. The excavations recount that on certain days water rose to flood the excavated area, gradually submerging the tanning ponds until they filled, and on another day began lowering its level, leaving the ponds empty — a strange rising-and-falling flow linked to al-Zuhri's description of the basins, though the link is not scientifically proven 5. A history of Toledo's streets notes that in 1973, at the foot of San Sebastián hill near the river, ruins of an old building were found, perhaps part of the Adabaquín gate or remains of al-Zarqali's famous clock, and that researcher Clara Delgado in her 1987 book "Islamic Toledo" set out location criteria with two possible installation points 5. Precise reconstruction is impossible: no plan, dimensions, material or known mechanism exists 4.
The spread
No read source names another city that made the like of the basins, no workshop that transmitted them, and no traveller who carried them 1. What we observe is the spread of the account itself: from al-Zuhri's twelfth-century description to al-Maqqari's seventeenth-century version, then to modern studies treating it as an open engineering puzzle 2 4. Its classifier in the "Wonders of al-Andalus" is itself a literary category gathering the marvels of cities and monuments, not a register of documented inventions; and the Kān study places the episode in a wider frame: that the political and military upheavals accompanying the conquest of Islamic cities by the Spaniards affected civil-scientific projects, some known and many unrecorded 4. The Baylatan thus spreads as a text, not as a craft; the museum's current photograph of Toledo and the Tagus is setting context for the reported site, not an image of the device 6.
What it was built on
At the level of the tale, the reported design rested on news of the Indian talisman of the city of Arin that "turned with its finger from dawn to sunset," which al-Maqqari says al-Mas'udi mentioned 1. That talisman is a transmitted narrative precursor inside the same text, not a device we read or verified; al-Mas'udi's own text was not read independently, so we claim no proven technical precursor 1. At the civilisational level, the Kān study places the basins within a broader Andalusian tradition of fountain- and water-clock-making, without naming any specific earlier device as their model 4.
What was built on it
No read source names a later device built on the basins or imitating their system; their fate in the accounts is an end — one was lifted and both stopped working 1. Later water clocks in history are an independent technology with no documented lineage to them; the modern archaeological hypothesis concerns a possible water source, not a working replica 4 5.
Impact on daily life
Inside the tale itself, the reported effect is that the basins enabled Toledo's people "to know the time and day from night" and "to reckon the lunar day," as al-Zuhri's version reposted in the Spanish report puts it 3. This is narrative effect inside a historical text: there is no evidence of a user count, documented daily use, or that the device ever worked 4. The value here is historical and epistemic: an account documenting an Andalusian engineering ambition to bind water to the months, not proof of living use 4.
How it changed history
The basins changed no documented history, but the tale changed what we know of al-Andalus's historical image: it became a stock example of the "wonders of al-Andalus" in the books of geography and history, and the subject of a modern study of fountain engineering and of an ongoing archaeological debate about its site and water source 1 4 5. Roth's analysis shows the tale itself carries marks of textual transmission — the astrologer's name differing between al-Zuhri's and al-Maqqari's accounts, the pigeon story dated 527 AH, and a fabricated poem Roth exposes — a lesson in how wonder reports take shape 2.
Who invented it?
No one holds a certain answer. The transmitted account attributes the craft to Abd al-Rahman known as al-Zarqali, matched by the Rabat manuscript's note, and described by al-Zuhri in the earliest detailed testimony 1 3. But there is no described mechanism, no surviving remains and no independent confirmation, and the sabotage accounts multiply (al-Funj's order, the astrologer's plot, theft), so the attribution remains transmitted and contested, not proven fact 4 2. The museum therefore records an attributed "improver," not a certified "inventor": even inside the tale the device was made on a transmitted model, not on documented creation 1.
Today
Today neither basin nor a copy exists; what remains are texts: al-Zuhri's account as it reached us through al-Maqqari, Roth's analysis of the transmission chain, the Kān study treating it as an engineering puzzle, and an archaeological hypothesis reposted from Toledo's water conference proposing that the Las Tinerías spring may have supplied the basins through the groundwater 4 5. The museum's current photograph of Toledo and the Tagus gives only the geographical setting of the reported site, not an image of the device 6. The illustration on this entry's page is a conceptual teaching chain: from the transmitted talisman report to the attributed maker to the reported fill-and-ebb mechanism — not construction plans or a reconstruction 1.
Surviving examples
No material witnesses survive: no basin, channel, inscription or plan; their reported fate is an end — "it was lifted and both stopped working" in one version, and a story remained 1. What survives is texts: al-Zuhri's twelfth-century description as preserved by al-Maqqari in the seventeenth century, Roth's modern analysis exposing transmission variants, the Kān study carrying it into an open engineering debate, and the archaeological hypothesis excavations have not yet proven 2 4 5.
In culture
The tale belongs to the "wonders of al-Andalus" in the books of Andalusian geography and history, and the Kān study names it "the marvel of Andalusian water-fountain engineering" 1 4. New cultural material has appeared around it in the twenty-first century: a Spanish press report on "al-Zarqali's water clock resurfacing from oblivion," and Toledo's water conference where archaeologist Julián García Sánchez de Pedro presented his Las Tinerías spring hypothesis 5. Roth warns that a poem supposedly by Ibn Ezra about the water clocks is in fact the fabrication of a nineteenth-century Jewish writer unworthy of trust, so it is no genuine cultural evidence 2.
For young readers
Imagine two big basins in a house by a river, one filling while the other empties as the moon changes — that is how Andalusian books described the "Baylatan of Toledo" attributed to al-Zarqali 1. Nobody today knows how they really worked, and they may never have existed; the story itself is part of Toledo's history 4.
References
- 1 — al-Maqqari, Nafh al-Tib min Ghusn al-Andalus al-Ratib, chapter on the wonders of al-Andalus: the Baylatan of Toledo (local scanned edition, scan pages 242–243).
- 2 — Norman Roth, "Comunicación breve: relojes de agua en la Toledo islámica," MEAHHEBREO 41 (1992), pp. 151–152.
- 3 — CIHAR repost of the ABC report "The Andalusian lunar water clock" (14 August 2023), carrying al-Zuhri's text on the basins and the Las Tinerías spring hypothesis.
- 4 — Abdullah Ali Nuh and Karima Ahmad Awad, "Baylatan of Toledo: the marvel of Andalusian water-fountain engineering," Douria Kan al-Tarikhiyya 16(61), September 2023, pp. 58–73 — abstract and bibliographic record only; the article body was not read in full.
- 6 — Wikimedia Commons photograph of Toledo and the Tagus (Gaguilarq, CC BY-SA 3.0), used as setting context only.
Connections
- Improved by
- Al-Zarqali
- Exemplified by
- Baylatan of Toledo
People and things also explored
The Toledo baylatan: twin basins credited to al-Zarqali working on the moon's cycle· Al-Zarqali's independent equatorium· The Toledan Tables· Al-Zarqali's universal astrolabe plate· Al-Zarqali's moving solar-apogee model· Cordoba· Toledo· Baylatan of Toledo
Sources
- Nafh al-Tib min Ghusn al-Andalus al-Ratib — al-Maqqari , pp. 242–243
- 'Water clocks in Islamic Toledo' — Norman Roth, MEAHHEBREO , abstract and pp. 151–152
- 'The Andalusian lunar water clock' — CIHAR repost of the ABC report , al-Zuhri quotation
- 'Baylatan of Toledo: the marvel of Andalusian water-fountain engineering' — Douria Kan al-Tarikhiyya , abstract
- 'The Andalusian lunar water clock' — CIHAR repost of the ABC report , archaeology hypothesis
- Toledo and the Tagus river photograph — Wikimedia Commons , file description
Cite this page
islam.museum contributors. “Baylatan of Toledo — twin water basins that followed the lunar month”. islam.museum, https://islam.museum/en/invention/baylatan-of-tulaytula/. CC BY-SA 4.0. Last updated 2026-11-15.
Improve this page on GitHub · Report an error · Help complete it · JSON