The Niyamināmā of Mūsā Mālimī

Synopsis of batches 1 and 2 — manuscript pages 2 to 33  ·  352 transcribed lines

The two navigators

Ahmad ibn Mājid was born about 1432 at Julfar, on the coast of what is now Ras al-Khaimah, into a family of pilots — his father and grandfather had held the title before him. He wrote some forty works on navigation, in verse and in prose, of which the greatest is the Kitāb al-Fawāʾid fī uṣūl ʿilm al-baḥr waʾl-qawāʿid, the Book of Profitable Things concerning the First Principles and Rules of Navigation, finished about 1490. It is the fullest account surviving of how the Indian Ocean was sailed before the Portuguese came: the lunar mansions and the thirty-two rhumbs, star altitudes measured in fingers, the monsoon seasons, and sailing directions from East Africa to the China Sea. A later tradition made him the pilot who brought Vasco da Gama from Malindi to Calicut in 1498, but the story is late and doubtful, and Tibbetts — whose Arab Navigation in the Indian Ocean before the Coming of the Portuguese, 1971, remains the standard study and is used throughout here — set it aside.

Mūsā Mālimī is known only from this book. The title gives his house and his island — Arāiragē in Hulhudhoo, Addu Atoll, the southernmost inhabited atoll of the Maldives, a degree south of the equator — and his calling. Mālimī is the Dhivehi form of the Arabic muʿallim, the same word Ibn Mājid uses for the master navigator: the man who knew the stars, the coasts and the seasons, as distinct from the rubbān who commanded the ship. Nothing else of him survives. The front cover carries three dates — 1908, 1919 and 1920, against 1339 of the Hijra — at least one of them in a later hand, and which marks the compilation is not yet settled.

The two books are five centuries and four thousand kilometres apart, and they are not copies of one another. They share a star vocabulary, a compass of thirty-two rhumbs and a stretch of the Arabian coast where their figures agree place for place. But the Niyamināmā keeps a Maldivian tradition underneath the Arabic — Dhivehi star names written in Arabic letters, the nakaiy calendar doing the work Ibn Mājid gives to the Arabic seasons, and a survey of the archipelago and the Indian coasts that has no counterpart in the Fawāʾid at all. Reading them against each other is what this synopsis is for.

Names are given as the manuscript spells them, romanised; identifications and the arithmetic behind them are editorial.

The shape of the book

Five sections, each opening with a red section mark. The book moves from how to sail, to where places are, to how to compute, and ends in a coordinate survey.

One · The sailing directions

Two independent variables govern the choice of route.

The current is primary.

The book opens on one, not on a bearing: in the Adha season, if the vavoi runs strong, take this route; if the vavoi is slack, take that one. Twenty-one lines turn on a current.

Five states, not two: running hard, strong, slack, none at all, and if it is running. Routes come in pairs so the pilot chooses from what the water is doing rather than from the calendar.

The season is secondary, and expressed in Nakaiy.

Season and current are not locked together — the book’s first line places a westward-setting current inside a southwest-monsoon season, which is why every route carries both variants. A season can also be named by a single Nakaiy (the Adha season, Nakaiy 6) or by the state of its current.

Two · Fixing position

Twenty pairs of islands are declared level north and south — the same meridian, therefore the same longitude — with six more given as simply level. One entry attaches a figure: level north and south, five minutes.

Where two islands are not on a meridian, position comes from a star.

Since every star gives both a rising and a setting, the two forms are halves of one system — the same 32-point sidereal rose used for steering, turned sideways for cross-bearings. The pole star is the exception that proves it.

Three · The computations

Pages 20 and 21 are the first calculation in the book rather than observation.

The 178 is the figure that shows measurement rather than assumption. The true autumn-to-spring interval is 178.8 days, because the earth runs faster in January. The 184 for the other half should be about 186 — one half is right and the other two days short.

The equinox is the day the sun struck the middle; the solstice is the mīl striking its corner, north or south. The majrā answer is an altitude, not a bearing — the same currency as everything else, checkable against the sky on arrival.

Four · The landfall table

Three altitudes at once: the pole star falling from 10 to 4 as Huvan nakaiy and Sal rise from 3 to 9. Two counter-running scales and one fixed one make a three-star fix.

The figures agree with Tibbetts’s pole-star column place for place, so the Addu pilot and Ibn Mājid work one coastal ladder in one unit. Every identified entry sits within half an isbaʿ of its true latitude, and the run ends at Ras al-Failak beside Guardafui — exactly where Tibbetts’s ladder ends. There is no Red Sea in it.

The section opens with an alignment rather than an altitude: when the rising star of Vega and the two rising stars of Arcturus stand level, the ship is at Socotra.

The stars used throughout:

Five · The coordinates

From page 28 the book becomes a survey, and this is the most remarkable thing in the two batches. Faresmaathodaa is taken as the line — it lies at 0.147° N, nine minutes off the equator — and every island is given its distance north in degrees and minutes.

Mean error 3.2 minutes of arc — about six kilometres — with three islands inside a fifth of a minute.

The list then crosses to the Laccadives — Kavaratti, Agatti, Androth, Amini, all within a few minutes — and to Ceylon, where longitude appears alongside latitude.

A constant offset of 22.4° places the prime meridian in the Atlantic off the African bulge — the Fortunate Isles origin of the Ptolemaic tradition. Sangamakanda Point, identified independently from Ibn Mājid, lands within eleven minutes of its true longitude on that offset. Devinuwara — Dondra Head — comes out at twelve seconds of arc.

What the two batches establish

This is not Ibn Mājid in Dhivehi. The book shares his star vocabulary, his rhumb system and his Arabian landfalls, but keeps a Maldivian tradition underneath — Dhivehi star names written in Arabic letters, the nakaiy calendar doing the work of the Arabic seasons, and a survey of the archipelago with no counterpart in the Fawāʾid. Devinuwara appears under its Sinhala name where Ibn Mājid uses the Arabic form, which suggests the Ceylon coordinates are independent rather than copied.

And the pilot could fix a position absolutely, not only relatively. Latitude from a named zero on the equator, longitude from a prime meridian, a rule for the course between any two points, and a declination method to correct the sun — with the answer always returned as an altitude a man on a deck could verify.

What the later pages add

The coordinate survey does not stop at Ceylon. Later pages carry it up the Coromandel to Pulicat, round the head of the Bay of Bengal to Satgaon, along the Arakan shore to the Andamans, and on through the Straits of Malacca to the China coast — the furthest entry standing at 35° 23′ north, which is Nanjing’s latitude.

The northern Coromandel and the west coast of Malaya are the best-surveyed stretches in the book. Kunimedu, Chennaipattinam, Mylapore and Pulicat run in coastal order within five to eleven minutes of latitude; Kedah, Dinding, Selangor, Malacca and Johor within one to eighteen, three of them inside three minutes. Madras appears twice, as the fishing town and the temple town, ten minutes apart in the manuscript and five in reality.

The longitude offset does not hold that far east. The 22.43° derived from Ceylon fails progressively: the five Malay ports all read about three degrees low, and would need about 25.4°. Either the scale compresses with distance from the origin, or a second origin is in use beyond the Bay of Bengal.

Two of the eastern pages carry a group of longitudes in the 102–105 range among entries whose neighbours lie far to the east — five on the Bengal page, three on the Arakan page. Those figures fall in exactly the range the Ceylon and Coromandel pages use, which would be explained if a leaf of the earlier list had been copied in among the eastern entries.

The eastern material also dates the book. Harishpur and Balasore held East India Company factories from the 1630s, Satgaon was Bengal’s principal port before the river silted, and one entry reads Hollandia, the Dutch. That places the source material in the seventeenth century or later, whatever the date of the copy.

Open questions

The 184 in the declination rule, against the near-exact 178.

The unexplained 11 in the majrā rule — at these latitudes no cosine correction is needed, so it must be a unit conversion.

Three landfalls on the Arabian table remain unidentified: Haseeraa on the Omani coast, Kariāree at pole star 7, and Wazyāḥ at the Horn.

Pages 34 and 35 missing from the photographs.