About

A finding aid for catalogues of Indian manuscripts, the collections they describe, and the manuscripts digitised by the National Mission for Manuscripts (NAMAMI): 1,686 catalogues from 12 sources, 706 collections, 210,545 digitised manuscripts.

Each record lists every source that attests it; records found in several sources are merged. Overlaps and gaps: statistics page.

Sources

• The list of digitised manuscripts comes from the NAMAMI digitisation list (210,545 records). It was imported once from the Mission's spreadsheet.
• Kṛtisampadā: the National Mission for Manuscripts’ national database is now presented as Pandulipi Patala (Bharatiya Kriti Sampada). Government sources describe more than two million manuscript records. It is represented here as an online-database record and linked for searching; its record-level data have not been copied because no public export or stable API has been identified.
• IGNCA CATCAT and manuscript catalogues: CATCAT is the closest institutional predecessor to this catalogue of catalogues. Its official description and IGNCA’s organisation-wise manuscript inventories are represented here. The next task is to recover a catalogue-level export from the legacy database and deduplicate it against Biswas, Janert and the library harvest.
• Most catalogues (1,096) and many collections come from S. C. Biswas, Bibliographic Survey of Indian Manuscript Catalogues, being a union list of manuscript catalogues (Delhi: Eastern Book Linkers, 1998). The records were read by machine from an OCR scan of the book; each says which Biswas number it comes from, and catalogue pages reproduce Biswas's description of the catalogue.
• K. L. Janert, An Annotated Bibliography of the Catalogues of Indian Manuscripts, Part 1 (Wiesbaden: Franz Steiner, 1965) attests 379 records. Its entries were read by machine from the OCR text of the book and matched against the records from Biswas; a match adds Janert to the record's sources, and an entry with no match becomes a record of its own. Entries the matching was not sure of are held back until a person has compared them. Catalogue pages reproduce Janert's annotation.
• Links to scans come from the INDOLOGY archive of scanned catalogues, gathered by Dominik Wujastyk (indology.info/scanned-catalogues). Its folder of catalogues that Biswas does not list is included as separate records.
• Library catalogues: the Library of Congress catalogue (attesting 666 records) and Open Library (680 records) were searched by subject heading ("Manuscripts, Tamil -- Catalogs" and so on for each language of the manuscripts), keeping only catalogues of manuscripts. A library record that matches a record from Biswas or Janert adds the library to its sources, with a link to the library's record; one that matches nothing becomes a record of its own (449 records, 166 of them published after 1998). Reprints are noted on the record of the original. These records repeat the library's cataloguing and are marked "not independently checked".
• Nepal: a search in 2026 for catalogues of manuscripts in and from Nepal (the Library of Congress catalogue, archive.org, the NGMCP pages in Hamburg, the INDOLOGY archive and list, DSBC) attests 42 records. Each research record keeps the pages that attest it; a record that matches one we have adds the research to its sources with any new online copy or identifier, and one that matches nothing becomes a record of its own (15 records, 9 of them online databases such as the NGMCP title lists). Records the research could not confirm are not shown. Contacts for collections in Nepal come from the institutions' own websites; where those did not answer from outside Nepal, from the copy the Internet Archive saved, as each collection page says.
• Texts: the works recorded in the New Catalogus Catalogorum (University of Madras, 1949–), read from its online edition at vmlt.in. For each work the site keeps the facts only — the headword, the place in the NCC, subject, author, cross-references and every manuscript reference as printed (catalogue siglum and number) — not the text of the entry, and links to the entry online. A siglum is linked to a catalogue in this list where the match is clear: by hand, or through the NCC's own list of abbreviations (1984). Manuscript counts follow the item numbers printed. See the statistics for how much is covered.
• Electronic texts: the title catalogues of SARIT and GRETIL Corpus Search and the downloadable IAST catalogue of the Muktabodha Digital Library (1,370 e-text records in all). A readable text is attached to an NCC work only on an exact normalised title match; its page names the source and warns that the edition or recension may differ.
• Manuscripts in Nepal: the NGMPP title list, the index of the 117,272 Indic films of the Nepal-German Manuscript Preservation Project (1970–2001), from the dump of its database published as eGangotri/ngmpp_sql_dump on GitHub (courtesy K. Harimoto); and the NGMCP Descriptive Catalogue (15,832 manuscripts described by the Nepal-German Manuscript Cataloguing Project, 2002–2014), whose wiki the University of Hamburg retired in 2025, from the CC0 mirror by Dominik Wujastyk (INDOLOGY/NGMCP-Descriptive-Catalogue, commit c478b2f; a second mirror by Tyler Neill, tylergneill/ngmcp-wiki-mirror). Facts only: title, reel, NAK accession number, holder, material, script, folios, date, and the ṣaṭka where an entry names it; each described manuscript links to its entry in the mirror. The films go to the NCC work whose headword their title matches; titles the NCC lacks become works of their own, marked "NGMPP only". The NGMCP's own search: catalogue.ngmcp.uni-hamburg.de.
• The list of catalogues used in the New Catalogus Catalogorum: Catalogues, Lists, etc., used in the New Catalogus Catalogorum, with the abbreviations used for them (University of Madras, 1984; scan on archive.org) attests 387 records. Its 396 abbreviations attached to a record were read from the OCR text of the scan, cleaned by hand where the scan had set a page in two columns, and matched like the other sources; each decision on an unsure match records its reason. Many of its hand-lists from private houses, mathas and bhandaras were never printed: their records (20 of the NCC's own) are marked "unpublished", and on a record the source names the abbreviation the NCC uses for it ("NCC 1984 list, “Kotah”").
• archive.org: scans on the Internet Archive whose titles matched no record were looked at one by one from their metadata. A scan of a catalogue we have, or of one of its volumes, is linked from the catalogue's page ("Internet Archive (vol. 4)"); a catalogue we did not have becomes a record of its own (17 records), marked where its year is read only from the title of the scan. Editions, studies, sale catalogues, second scans and catalogues of Arabic and Persian manuscripts only were left out.
• Links to HathiTrust come from the HathiTrust Bibliographic API, asked with the LC and OCLC numbers of each library record: "HathiTrust (full view)" can be read online (in the United States, at least); "HathiTrust (search only)" can only be searched for words and page numbers.
• A few catalogues were researched by hand, following Janert (1965) and the New Catalogus Catalogorum, and checked against a title page or a library record.
• Contact details are taken from each institution's own website.

What "source-only" means

Every catalogue page says which sources attest the record ("Attested in: Biswas 1998 no. 0012 · Janert 1965 no. 3 · Library of Congress (LCCN 45043905)"). "Checked against the title page" means someone compared it with the catalogue itself or a library record. "Not independently checked" (source-only) means the record repeats its source as printed, and the OCR may have damaged names, diacritics or numbers. Places on the map for these collections are the centre of the town, not the building.

What "contacts" means

Institutional contacts only (website, general address, switchboard, office email), with the date they were checked. Contacts change: check the institution's website before you write.

How to contribute

Use the add a catalogue / suggest a correction form. Every catalogue and collection page links to it with the record already filled in. Until a receiving address is configured, the form copies a complete report to the clipboard.

Licence

The code is under the MIT licence. The curated data about catalogues and collections is under CC BY 4.0. The NAMAMI manuscript records are © National Mission for Manuscripts and are reproduced here for reference.