Panchangam Reference

Appendix II lunar longitudes: research transcription ​

This is a research-only numeric transcription of all 248 lunar longitude rows. Source: the 1962 scan linked in metadata.json, physical PDF pages 162-171, printed pages 125-134. The opening page's printed numeral is absent; 125 is inferred from the following printed 126 and the book pagination.

Files ​

  • lunar-longitudes.csv: flat machine-readable rows and locators.
  • lunar-longitudes.json: identical structured rows.
  • metadata.json: source SHA256, units, page ranges, method and limitations.
  • ambiguities-and-source-discrepancies.json: unresolved discrepancy and header errors.
  • symmetry-check.json: diagnostic against the source's printed verification rule.

The Sanskrit mnemonic column is outside this numeric transcription's scope. Physical page numbers and row-on-page locators are one-based. The separate pdf_page_index is zero-based. Intermediate unnumbered day rows follow their printed anchors in order.

Findings requiring care ​

  1. Day 100, physical 165 / printed 128 / body row 21 reads 8r1deg17min.
  2. Day 148, physical 167 / printed 130 / body row 16 reads 4r26deg17min.
  3. These sum modulo 360 degrees to 0r27deg34min, ten arcminutes below the printed page 134 verification target 0r27deg44min.
  4. Both minute cells visibly read 17; no corrected value is assigned.
  5. Physical 170's running header says 208-232, but its 27 body rows span 208-234.
  6. Physical 171's header says 233-248, but its 14 body rows explicitly begin 235.

There are no remaining visually unreadable numeric cells in this reading. The source-authority discrepancy above remains unresolved. An initial low-resolution reading of day16's minutes as 16 was rejected after 250dpi inspection: the source visibly reads15; the final artifacts use15. Faint cells in rows 142,148 and155 were checked at 300 dpi.

Validation and limits ​

All 248 sequential day indices are present exactly once. All numeric cells satisfy their column ranges. The book's complement relation passes all pairs except 100/148; this arithmetic diagnostic never supplies or silently corrects a cell. Day 124 is 6r13deg52min and day 248 is 0r27deg44min, consistent with the printed midpoint statement on page 134. Days 153 and154 agree with root's separate readings:7r7deg36min and7r21deg48min.

This is a single-transcriber draft, not an independently certified table. No epoch, ahargana, interpolation, time-of-day or engine model is implemented. It does not validate the current nine-offset engine. Publication of this transcription changes no engine or calculation behavior.

Next evidence steps are independent double-entry transcription, checking an alternate edition or errata for 100/148, and separately reconstructing the source's epoch/argument rules before any proposed model comparison.

Root spot review — 2026-09-21 ​

A separate visual spot review of complete physical pages 165 and 167, plus the higher-resolution crops, confirmed the day100 and day148 readings and their row locators. Both minute cells read17; the ten-arcminute inconsistency remains. This checks the flagged pair only, not independent double-entry of all248 rows.