Every mark on this chart is an entry in a hand-built database of star names drawn from primary Sanskrit texts, spanning roughly 1500 BCE to 1150 CE. Positions are modern J2000 coordinates for the star each verse is understood to name; marker size grows with apparent brightness.

How to read the marks. Filled circles and diamonds are single stars. Dashed rings are asterisms — groups of stars, plotted at the group's position. Dashed gold lines are the Purāṇic star-roads (vīthī), each drawn through its own three nakshatras. The soft band is the Milky Way, computed as the galactic equator from the J2000 galactic pole; the three Sanskrit names for it sit on that band.

What is deliberately not plotted. The Purāṇas name many positions on the body of the Śiśumāra, the celestial porpoise, but never say which visible star any of them is. Those names are in the Index rather than pinned to a guess — every limb-to-star chart in circulation is modern reconstruction. The same goes for named sky-regions (Viṣṇupada, the mārgas) and for Bhāskara's collective star-catalog.

What the texts actually give. The Vedic texts name asterisms and deities but never coordinates; the siddhāntas give a position for one junction star (yogatārā) per nakshatra. Modern equations rest on those siddhāntic positions — chiefly Burgess's 1860 analysis of the Sūrya Siddhānta — and on continuous tradition. The confidence badge on each entry records where that chain is firm and where scholars disagree.

Two findings worth stating plainly. The Āryabhaṭīya names no individual stars at all — verified against both W.E. Clark's 1930 translation and the full Sanskrit text — so it contributes nothing here. And the Ṛgveda names almost none: an exhaustive search of the Aufrecht text yields only the Bears (ṛkṣāḥ), the archaic Aghā and Arjunī of the wedding hymn, and Tiṣya twice. Its Agastya is always the sage, never Canopus.

Sources. Sanskrit copied verbatim from Sanskrit Wikisource and GRETIL; translations are public-domain (Burgess 1860, Iyer 1884, Keith 1914/1920, Eggeling 1882, Whitney 1905, Griffith 1896, Oldenberg 1886, Wilson 1840, Colebrooke 1805, Sanyal 1930s, Wilkinson & Sastri 1861). Copyrighted translations are paraphrased and cited, never quoted. Where no published translation exists, the rendering is labelled as the compiler's own.

Full database with every caveat, and the Śiśumāra mapping table: docs/star-names/ in the repository (README.md and star-names.json).