Nihayat al-sul fi tashih al-usul

Nihayat al-sul fi tashih al-usul explains Ibn al-Shatir's astronomical models; Victor Roberts studied its solar and lunar constructions in1957. This record distinguishes the inherited book from the scholarship actually read. Reading the article does not mean the museum inspected every folio of the manuscript on which its author relied ^{1 2}.

King describes a sequence beginning with a lost Ptolemaic handbook, followed by Taliq al-arsad, which described new procedures and parameters and has no known copy in his account. Nihayat al-sul explains the models' reasoning; the surviving new handbook contains tables based on the theory and parameters. Different functions are not interchangeable titles for one book, and surviving tables do not make a lost observational work available to read ².

Dizer lists these among a wider body of instrument, quadrant and timekeeping writings. Shared authorship does not erase their different genres. Instructions for a physical quadrant are not automatically a proof of lunar theory, and the manufacture date of an astrolabe does not date this book's composition ³.

Roberts lists five witnesses known in1957: Marsh139, Marsh290, Marsh501 and Hunt547 at the Bodleian, and Arabic1116 at Leiden. He used a microfilm of the first, describing sixty-four folios and a copying date768AH/1366, within the author's lifetime. These are his dated descriptions, not a complete current inventory or evidence of an autograph ¹.

This explains the open chronological bound in the lunar-model record: a1366 copy places the text at or before that date without fixing its invention year. We have neither inspected the original nor collated witnesses. Citation anchors therefore identify printed article pages, not manuscript folios supposedly read by the museum ¹.

Roberts describes dispensing with the eccentric carrying circle and adding a second epicycle. He does not find the solar model's improvement over its Ptolemaic counterpart clear in his account, whereas reducing exaggerated lunar-distance variation is a specific benefit. The lunar result cannot certify that every result throughout the book improved by an identical amount ⁴.

His lunar account gives radii60,6;35 and1;25, with an apparent resultant between5;10 and8. Page432 uses modern mathematical symbols that Roberts explicitly presents as expressing the text's method; he also distinguishes certain values he calculated from a transmitted one. Assigning all notation to Ibn al-Shatir or resolving sign ambiguities without a manuscript comparison would exceed our reading ⁵.

A published disagreement needs attention: Dizer gives43;50 for the lower conjunction/opposition distance, while Roberts prints54;50. The modern teaching example in the invention record checks sums and differences of the radii, but neither erases this discrepancy nor claims to edit the Arabic text. Normalized numbers are not kilometres or certification against space-age measurements ^{6 5}.

Roberts compares the Moon with Copernicus and leaves transmission unresolved, allowing independent solutions and reporting no evidence of a Latin translation in his investigation. King later argues more strongly for some transmitted details, but supplies no named translator or specific route established by this research. Resemblance and historical movement are related, not identical questions ^{7 8}.

Centuries of circulation of tables and instrument manuals must be distinguished from adoption of every argument in this treatise. King separates later interest in handbooks from interest in non-Ptolemaic models. The invention's accompanying image is a licensed modern distance illustration, not a manuscript photograph or republication of Roberts's copyrighted figure ^{8 9}.

The actual readings are Roberts's complete article and King's and Dizer's biographies. The Arabic original, other witnesses and missing observational book are not readings undertaken for this record ^{4 10 11}.