Dikshitar, The Purāṇa Index (PUI)
PUI is V. R. Ramachandra Dikshitar's The Purāṇa Index (1951) — a large reference index to
the names of the major Purāṇas: each name gets a brief identification and precise references
into the texts. The terse, locational complement to the narrative Purāṇic Encyclopedia (PE).
With 17,500+ entries across two printed volumes, it covers exactly the onomastic layer the Purāṇas are richest in — genealogical figures, tribes and peoples, pravara lineages, sacred geography, ritual objects — and 38.6 % of its headwords occur in no other digitized dictionary on CDSL.
At a glance
| Code | PUI (source in csl-orig/v02/pui; no separate repo) |
| Full title | The Purāṇa Index |
| Author | V. R. Ramachandra Dikshitar |
| Year / size | 1951 · ~2232 pages · ~17,500 entries · 2 volumes |
| Type | Name index (Purāṇas) |
| Source | csl-orig/v02/pui/ |
| Open | Basic · List · Advanced · Mobile |
| Data | Downloads |
| csl-doc | pui.rst |
When to use it
Reach for PUI to find where a name occurs in the Purāṇas with a one-line identification.
Concretely:
- Pin a mythological figure to chapter-and-verse: every entry ends in precise textual references, not page numbers of one edition but book.chapter.verse loci that survive translation.
- Disambiguate homonyms: same-named figures get separate entries distinguished by sub-labels (see below).
- Check a claim's spread: several references under one name show whether a story belongs to one Purāṇa or is told across many.
For the fuller narrative use the Purāṇic Encyclopedia (PE); for word meanings use Monier-Williams (MW).
Reading an entry
Each entry is a name (<k1>) with a short gloss and Purāṇa references. The first entry
(csl-orig/v02/pui/pui.txt):
<L>1<pc>1-001<k1>aMkuSA<k2>aMkuSA
{%Aṃkuśā%}¦ — a Śakti on the Drāviṇikā mudrā.
<div n="P"/>Br. IV. 36. 76.
Two more, showing how terse the identifications can be (csl-orig/v02/pui/pui.txt):
<L>501<pc>1-064<k1>antargiri<k2>antargiri
{%Antargiri%}¦ - a tribe.
<div n="P"/>M. 114. 44.
<L>9003<pc>2-478<k1>bAlaqi<k2>bAlaqi
{%Bala?i%}¦ - Pravara (A?giras).
<div n="P"/>M. 196. 15.
| In the source | Meaning |
|---|---|
{%Aṃkuśā%} | the name (search key <k1>) |
— a Śakti on the Drāviṇikā mudrā | the brief identification |
<div n="P"/>Br. IV. 36. 76 | the reference — Brahmāṇḍa Purāṇa IV.36.76 |
- a tribe / - Pravara (Aṅgiras) | genre labels: a people; an ancestral-seer name |
The identification line is deliberately telegraphic — often a single category word ("tribe", "Pravara", "river", "son of X") — because the index's job is location first, story second.
See Data Formats for the markup reference.
The citation system
References follow the pattern sigla + book.chapter.verse. Counted across the file,
references beginning Br. (Brahmāṇḍa Purāṇa) dominate at ~7,100, followed by Vi. at
~3,200 — with the rest of the corpus distributed over one- and two-letter sigla for the other
texts (the digitized front matter, scanned via csl-doc,
carries Dikshitar's own key). Over 15,000 reference lines sit under the <div n="P"/>
marker, so multi-reference entries are the norm: a major figure like Kṛṣṇa or Manu carries
dozens of loci spanning many Purāṇas.
Same name, different entries
Homonyms are kept apart by sub-labels attached to the key — visible in the source as an
<h> marker:
<L>9954<pc>2-601<k1>matsya<k2>matsya<h>IV
{%Matsya%} (IV)¦ - gift of a silver one during the ceremo-
nial relating to the digging of tanks.
<div n="P"/>M. 58. 19.
<L>9955<pc>2-602<k1>matsya<k2>matsya<h>V
{%Matsya%} (V)¦ - the image of.
<div n="P"/>M. 259. 2.
These Roman-numeral senses (~7,300 entries carry one) mean a search can return several distinct numbered entries under one form — each worth reading separately before dismissing.
Search tips
- Expect multiple hits per name. Homonym entries and cross-Purāṇa retellings both multiply results; scan the identifications, not just the headwords.
- Names appear in inflected and compound forms. Try the stem alone.
- Follow the verse loci. Book.chapter.verse references work across editions and even translations, unlike page numbers.
What makes it distinctive
- A locational index to the Purāṇas — name → identification → references.
- Verse-level precision: loci cite book.chapter.verse, stable across editions.
- Homonym discipline: numbered sub-labels keep distinct figures separate.
- Highly independent headword stock — over a third of its names occur nowhere else on CDSL (38.6 % unique share per the site's atlas measurements, see Guides Hypotheses).
Headword key conventions
<k1> = <k2> = SLP1 (Purāṇic name). Page references use vol-page format. Each entry gives a brief identification then Purāṇa text references.
| Convention | Marker / form | Example | Note |
|---|---|---|---|
| Search key | SLP1 (name) | aMkuSA | k1 = k2 |
| Body form | IAST italic {%…%} | {%Āṃkuśā%} | The printed name form |
| Page reference | <pc>vol-page | <pc>1-001 = vol. 1, p. 1 | Two-volume set |
| Homonym label | <h>N after the key | <h>IV, <h>V | Roman-numeral sense split of same-named entries |
| Citation | Purāṇa abbreviation + locus | Br. IV. 36. 76 | Brahmāṇḍapurāṇa book, chapter, verse |
Abbreviations
No machine-readable abbreviation list yet; the legend lives in the scanned front matter. View front matter →
See also
- Purāṇic Encyclopedia (PE) — the narrative companion
- Monier-Williams (MW) · the full catalog
- csl-doc pui.rst