HathiTrust is what happens when research institutions pool their scans. A partnership of universities and libraries — seeded largely by the Google Books digitization effort alongside the Internet Archive — it preserves millions of volumes and opens them as a digital reading room with real house rules, rather than as a free-for-all download pile.
Anything in the public domain is generally full view and downloadable whole as PDF, sometimes with the OCR text alongside. In-copyright works are searchable but readable only in fragments, with fuller access for members of partner institutions and separate arrangements for readers with print disabilities. That split is the identity of the place: preservation and lawful access, not lending.
No account is needed to search or read full-view items; institutional affiliation widens the door. On an e-ink reader, the downloaded PDFs are photographs of pages — manageable on a large screen, punishing on six inches, with no reflow to hope for. For out-of-print scholarship and historical material, particularly from the American academy, it is close to unmatched; for anything recent, it is simply the wrong shelf.