How to Download Wikisource Books and Check for Missing or Repeated Chapters

To download a Wikisource book, open the main page for the complete work, choose its Download option, and select EPUB. Then compare the downloaded book with the online contents before settling in to read. Printing a chapter page is not the same as exporting the whole book.

The download itself is the easy part. A file can contain every chapter and still have a structural problem: in one of three Wikisource EPUBs we inspected on September 10, 2026, all 95 chapter files were present, but a separate section appeared twice in the package's reading-order list. Here is a practical way to choose the right source, save the book, and check what actually arrived.

Start at the Work's Main Page, Not a Chapter

A search result may take you directly to a chapter. Follow the book-title link back to the page that introduces the work and lists its parts. Wikisource's offline-reading instructions identify this top-level page as the starting point for a multi-page export.

Before downloading, check three things:

• Identity: Is this the work, translator, and edition you want?
• Scope: Does the page collect the whole work, one volume, or only an excerpt?
• Language: Is the actual text in the language you intend to read?

For example, The Book of Tea has seven chapters on its contents page. Its author is Okakura-Kakuzo, but the file we downloaded from English Wikisource is English. An author's nationality is not a substitute for checking the text.

The Oldfather translation of Epictetus is a more complicated case: the main page combines two volumes, four books of the Discourses, and additional material such as the Manual and Fragments. “I downloaded Epictetus” does not describe the intended scope precisely enough.

Save the source URL, translator or edition details, and download date with the file. Review the work's own rights notices before choosing a subsequent use; a readable download does not establish permission for every upload or redistribution. If you are still choosing a source or language edition, begin with our guide to finding public-domain ebooks in the language you need.

Use the Book Export, Not Browser Printing

From the complete work's page:

1. Open Download or the Download EPUB option available in the page's export tools.
2. Choose EPUB for this workflow and save the resulting .epub file.
3. Keep the original download unchanged while you inspect it.
4. Open that local file in your reading app; do not mistake a saved webpage or download-link shortcut for the book.

Wikisource also lists PDF and other export choices. The availability of a button does not establish that a particular export is complete or well formatted, as its offline-reading help notes. Our examples below inspect EPUB only, not PDF export quality.

The separate Printable version command removes webpage furniture for printing the current page. Wikisource's reading guide explicitly distinguishes this from saving an entire multi-page book. A beautiful PDF of chapter one is still only chapter one.

If you need the appearance of a particular printed edition, follow the work's source-scan links where supplied; do not assume an exported reading file is a facsimile. If the destination is Kindle, continue with the Send to Kindle guide after checking your file rather than assuming every transfer method accepts the same formats.

Make a Small Inventory Before You Read

The most useful comparison is not “How many pages did the download produce?” It is “Which parts should this particular edition contain?”

List the major sections from the online contents. For a multi-volume work, expand each volume or book into its chapters. Keep introductions, dedications, appendices, and indexes separate from the main chapter count so they cannot quietly disappear from the check.

Use the blank chapter-inventory worksheet. For each expected section, record its source link, where it appears in your downloaded file, and what you actually checked. Leave uncertain results blank or mark them unverified; copying our example results would defeat the purpose.

Then make three different checks:

An eight-part work need not contain exactly eight internal files. Title pages, notes, and other material may have their own entries. Conversely, several chapters can share one file. Use the contents and actual text to establish scope; use file counts only to investigate discrepancies.

What Three Actual Wikisource Downloads Contained

On September 10, 2026, we downloaded three English EPUBs through the official export links. We inspected the ZIP archives, package metadata, reading-order lists, chapter file names and headings, and navigation destinations. The files were not edited before inspection.

All three passed the ZIP integrity check. Every inspected link in their EPUB navigation documents resolved to an existing file and, where specified, an existing fragment target. These results establish limited structural facts, not complete proofreading, EPUBCheck conformance, or successful rendering in a particular reader. We did not test these books in BookTranslator.

The dated inspection record includes the official source and download URLs, file sizes, SHA-256 hashes, counts, and limitations. We publish our inspection results, not copies of the library's books. Later exports may differ.

Why the Epictetus result matters

An EPUB's package includes a manifest of resources and a “spine,” its reading-order list, as described in the EPUB specification. In our Epictetus download, the Manual had the same manifest identifier twice and the same spine reference twice. We confirmed that in the original package XML, not just in a contents display.

That is different from a missing chapter. The 95 Discourses chapter files were there, and their presence should not be reported as a failure. Nor can a duplicate package reference tell us exactly how every reading app will respond. We did not perform that device test.

For a reader, the implication is simple: check the transition into and out of the Manual as well as whether the section can be opened. More generally, when a contents list looks complete, read through at least one boundary between major parts. Menu navigation and continuous reading exercise different aspects of the file.

If Something Is Missing or Repeated, Find Which Layer Is Wrong

Do not immediately convert the file or delete the suspicious section. First compare the download with the online work:

• The online work is missing the material too: check whether you selected an excerpt, an unfinished transcription, or a different volume. An exporter cannot supply a chapter absent from its source.
• The online contents link to a chapter, but the export lacks it: keep the exact source URL and file. Return to the complete work's page and verify the export scope; a generated-file problem is possible even when online reading works.
• The section exists but the menu skips it or the reading sequence looks wrong: record both the menu result and what happens when reading forward. Avoid treating these as the same symptom.
• The words are present but the page is hard to read: compare the same passage with the online source before blaming missing content. Poetry, narrow columns, notes, and images deserve their own checks.

Wikisource's export-preparation guide explains that the export depends on how the work and its subpage contents are organized; it also identifies formatting that can behave differently outside the website. These are reasons to inspect the actual output, not proof of the cause of every faulty download.

If the current export remains unsuitable, continue reading the online text or choose another clearly identified edition while the issue is investigated. If you report a problem to Wikisource, include the work URL, export format, download date, expected section, and the smallest concrete example. “Book broken” is much harder to act on than “the online Book IV chapter appears here, but I cannot find it in this dated EPUB.”

For a deeper check of an EPUB you intend to use or translate, the EPUB QA checklist separates package structure, navigation, notes, and reading order. It is a next step, not a requirement to become an ebook developer before reading a library book.

Optional: Inspect the File Without Editing It

Readers comfortable with a terminal can download our small Python inspection script. It uses Python's standard library, reads a local EPUB, and prints JSON. It does not upload the book, extract files to disk, change the EPUB, or repair errors.

python3 check-wikisource-epub.py "your-book.epub"

Look for duplicate manifest identifiers, repeated spine references, and unresolved local navigation targets. The output deliberately does not declare a book “complete” or “reader compatible.” It cannot know which edition you meant to obtain, identify every omitted paragraph, or replace a conformance checker. Use the chapter inventory for scope and your reading app for the final reading check.

Translate Only When Language Is the Remaining Obstacle

If the existing Wikisource edition already gives you the language and text you want, you are done: keep its source details and read it. A historical translation may also be the exact edition you need; generating a new one would not reproduce that translator's choices.

When no suitable reading edition is available in your language, first prepare and identify the complete source file. With a usable EPUB and appropriate rights for the intended use, BookTranslator's EPUB translator is an option for creating a reading translation. Review the translated meaning and chapter flow rather than assuming a successful download or translation request proves either.

The order matters: choose the edition, obtain the whole intended work, check the file, and only then address a remaining language barrier. Translation is not a repair for an unidentified edition, absent source material, or an export you have not inspected.