Al-Biruni's proposed simplified calendar gearing

A small wheel can make a pointer turn slowly, but fewer wheels do not by themselves make a calendar correct. Al-Biruni described an older device called the Moon Box, then proposed simplifying its solar gearing. This entry concerns that particular proposal: a four-tooth gear driving a fifty-two-tooth gear, not invention of the whole older device ¹.

Donald Hill derives a364day year from the ratios, whereas the translated passage mentions360days. The difference remains visible. This is not an announcement of a self-running clock or surviving exact calendar, but a technical text combining inheritance, modification, transmission errors and limits of reconstruction ².

The Moon Box mechanically represents mean motions: pointers and a phase disc moved through calculated tooth ratios. A user advances an input alidade by hand according to daily divisions; a complete input revolution takes a week. That is not a self-powered motor, nor does the pointer directly observe a body's actual position in the sky ³.

The bounded contribution changes the solar reduction path. The four-tooth gear shares the forty-tooth lunar wheel's arbor, so they turn together; the four then meshes with the fifty-two. A shared arbor transfers the same rotation, whereas a mesh changes speed according to tooth counts. Confusing these arrangements changes the whole calculation ⁴.

Hill associates the older machine with Hellenistic and Byzantine settings and says its original inventor is unknown. That is the scholar's historical inference, not proof that one complete machine survived continuously between two cities. The accessible Arabic account's antiquity does not make al-Biruni the first inventor of gearing or calendars ⁵.

The text also describes other craftsmen's arrangements before the proposal attributed to its author. An instrument's presence in a scholar's book does not make every described device his invention. Three layers remain separate: inherited device, craftsmen's variants and al-Biruni's particular concluding simplification ⁶.

The source actually read here is Hill's1985study, presenting an edition, translation and technical discussion from manuscript witnesses. We read the English introduction, translation and commentary, not the complete Arabic apparatus or independent originals. The full discussion of astrolabe constructions contains this material; it is not the unrelated biographical Istiab by another author ⁷.

This reading supplies neither an exact composition day nor a verified workshop for the proposal. The tenth/eleventh-century frame is a lifetime horizon, not a manuscript-derived construction date. Ghazna is biographical context, not a factory address established here. Accurate attribution matters more than filling every field with an attractive unsupported story ⁸.

Al-Biruni's individual Muslim identity is documented, not inferred from his name or state. The museum's Sunni frame is explicit, and Gunay Tumer's Sunni reading remains visible in his biography. Reported Shia and Zaydi affiliations remain labelled contrary accounts, without making identity disagreement an automatic judgement about a gear's or equation's attribution ^{9 10}.

This entry does not replace his earlier Earth-measurement, interpolation, displacement-vessel or qibla-setting work. A shared scholar does not merge different instruments: here the question is rotational ratio and a described machine. Nor do we invent names or affiliations for the earlier device's unknown maker or incompletely identified craftsmen simply to increase a count ¹¹.

The basic design lists eight wheels with7,10,19,24,40,48,59and59teeth. Counts alone cannot reconstruct a machine: one must know what meshes, what shares an arbor and what carries each pointer. Hill warns about erroneous drawings and partly inferred assembly; his suggested mountings and dimensions are not converted into certain physical specifications supplied by the author ¹².

The arithmetic begins with one input revolution every seven days. If a ten-tooth gear drives forty teeth, the latter's speed magnitude is one quarter of the former's, requiring four weeks or28days for a revolution. That is an approximate sidereal lunar cycle, not the interval between crescents or a promise of crescent visibility on a particular date ¹³.

The phase-disc path uses seven and fifty-nine teeth, producing a59day revolution. That revolution represents two phase cycles, giving a mean29.5day synodic month. Its difference from28does not contradict the preceding arithmetic: the same everyday word, month, covers different measurements, return relative to the stars and return relative to the Sun's direction ¹⁴.

In al-Biruni's simplification, the four-tooth wheel turns with the forty on one arbor. We do not divide speed by ten again just because their sizes differ: they do not mesh with each other. Four teeth then drive fifty-two, making the solar pointer's revolution thirteen times longer than the sidereal lunar revolution. Our modern period-magnitude calculation is7×(40/10)×(52/4)=364days ¹⁵.

The ratio reverses when moving from speed to period: final speed is one thirteenth of the lunar arbor's speed, but revolution time is thirteen times longer. Each external gear mesh reverses direction; this teaching chain computes period magnitudes only. It does not claim to solve every arbor's direction, tooth contact or pointer mounting in a functioning replica.

The earlier design's reconstruction gives59×(59/19)×(48/24), about366.4211days. Against Hill's365.2422year reference, the earlier output is about1.1789days too long, while the364day proposal is1.2422days short. It is therefore not more accurate than the earlier arrangement on every measure: it is an attributed simplification, not an unconditional mathematical victory ¹⁶.

If both pointers start aligned, period error does not appear all at once; the angular difference accumulates with elapsed days. After364days the proposal's pointer returns to its starting point while the reference year has not completed a revolution. This is a modern mathematical explanation, not an archaeological experiment we performed. The translation's360and Hill's reconstructed364remain a disclosed discrepancy, not a number silently corrected in the original ⁴.

Witnesses in different manuscript collections demonstrate textual circulation, not automatically manufacture of simplified calendars wherever a copy arrived. Hill explains variant readings and drawings; this entry supplies no unsupported sales map or European workshop chain ¹⁷.

Nor does a later writer's study establish widespread everyday operation. An engineering proposal can survive in a book without a surviving manufactured example. Unverified geographic and artifact ties are therefore left empty rather than filled with museums displaying modern models whose connection to this particular proposal has not been established here.

The proposal uses gearing, tooth ratios, compound arbors and approximation of different motions through regular periods. These are inherited layers in the account, not all independent discoveries by its author. Modern explanation also depends on critical reading: is a tooth count correctly transmitted, and does it agree with the stated period? ¹⁸.

For one craftsmen's variant Hill derives about327.72days, while witnesses give399or359; another variant yields354days. These are not silent substitutes for the proposal's364day year. They cannot be grouped beneath a claim that the machine was accurate without identifying which arrangement and which reading are meant ¹⁵.

Today the material supports a joint lesson in ratio design and textual criticism. One can compare a simpler alteration with a longer train, or calculate a single tooth's effect, without claiming ancestry of every modern clock or computer. Functional resemblance cannot replace evidence of transmission ¹⁹.

The museum's teaching diagram is a generic modern chain: manual input, reduction, compound arbor, pointer and error check. It neither reproduces Hill's protected plates nor claims a complete Moon Box model. No program needs to run for a reader to understand the ratio.

Hill suggests that reversibility might help investigate dates, including debt dates. That is a possible interpretation of usefulness, not a transaction archive proving adoption by a court or market. The distinction brings the question close to daily life without inventing a particular merchant's story ²⁰.

Pointers might also help visualize month and year, but do not make an observational month28days or resolve a month's beginning by gearing alone. Mean-motion representation, assigning a calendar date and seeing a crescent are different operations. This entry offers no religious ruling; it explains an engineering function and its limits.

Its defensible historical value is a text showing alteration and criticism, not a slogan that one machine created modernity. The account separates inheritance from craftsmen's methods, and Hill specifically attributes the final proposal to al-Biruni. Comparison shows that simplifying assembly may bring a different period error ⁶.

Counting fewer wheels cannot establish a linear march from error to truth. A simpler arrangement might be easier to understand, but proving lower manufacturing cost or greater operating reliability would require experiments and evidence unavailable here. We describe the ratio change rather than invent a historical efficiency measurement.

The attributed Sunni interpretation concerns the scholar's identity as Tumer presents it, not automatic priority for every device described. Hill's academic reading treats the Moon Box as older with an unknown inventor, and credits this particular simplified gearing proposal to al-Biruni. Other Shia/Zaydi affiliations remain labelled reported accounts, not a verdict chosen from arithmetic ^{10 21}.

The bounded answer is therefore al-Biruni's proposal in this passage, not invention of all calendars or gears. The text does not establish that it was built or survives. Naming a study after him does not make every earlier wheel his personal innovation.

A reader can check the calculation with a pencil: four weekly-input revolutions give one forty-tooth revolution, and thirteen revolutions of that arbor give one fifty-two-tooth revolution. Changing units to days gives364. Scaling all tooth counts proportionally preserves the mathematical speed ratio, not dimensions or mechanical strength.

The accompanying image shows modern involute teeth, not the historical machine's triangular profile. It cannot establish that a historical four-tooth gear meshes successfully, or supply its scale. It compares terms, arcs and dimensions while explicitly separating general engineering explanation from artifact reconstruction ^{22 15}.

The evidence used is an edition, translation and commentary from manuscript witnesses, not an original box we inspected. Hill warns of assembly not fully specified in the text and errors in some figures: correct ratios do not settle how the parts fit. We did not independently inspect originals or read the complete Arabic apparatus ²³.

Hugo4's image derived from Bokai is modern, public-domain material retained unchanged. Its availability does not confer historical authenticity, and its modern tooth shape cannot witness the shape of medieval teeth ²².

Calendar wheels may invite labels such as ancient computer or first astronomical clock. Neither priority is needed to appreciate the difficulty of representing unequal cycles through whole-number tooth counts. The pleasure lies in understanding a simple ratio and also calculating its limits ²⁴.

The grey detail with coloured arcs and dimension lines is not presented as a portrait or surviving machine. By putting text, engineering reading and numeric disagreement together, the museum offers knowledge as something open to criticism, rather than a list of unqualified triumphs.

Imagine a disc taking four weeks to turn. Attach another disc to the same arbor: both turn together even if their sizes differ. Then let a little four-tooth gear drive a fifty-two-tooth wheel. The large wheel needs thirteen small-wheel turns to return to its starting position. Four weeks multiplied by thirteen makes fifty-two weeks, or364days.

But the reference year is slightly longer, so the machine's pointer gets ahead. The machine has not stopped; its ratio is approximate. That is both the attraction and difficulty: making a model to understand the sky does not make the sky follow its teeth. This arithmetic example is modern, not a historical toy or building experiment attributed to the author ¹⁶.

• Hill: introduction139–142,translation156–158,commentary158–163; these English sections read and158/163visually inspected, not complete Arabic apparatus or originals ²⁵.
• Inherited individual Muslim basis and affiliation disagreement, without a new inference ^{26 27}.
• Commons: actual current image, description and public-domain rights; modern comparison only ²⁸.