In the Distributed Proofreaders forum (membership required for forum access), someone recenly asked if on-line images could be substituted for poor-quality scanned images. (Distributed Proofreaders is a feeder project for Project Gutenberg that prepares public-domain books for publication on the Internet.) For example, if a public-domain book contained an old, fuzzy photograph of Millay’s “Fishermen”, could the photograph of the same picture published on the Louvre’s web site be copied and published with the Project Gutenberg version of the book?
Project Gutenberg’s answer:
We just got some news on this issue. It turns
out (based on the most recent legal cases) that pictures
of artwork (at least, paintings) that try to depict
the artwork accurately are public domain, if the artwork
itself is public domain.
Basically, this opens the door for *any* photo of a
painting created prior to 1923 - even if the photo is
more recent.
This decision stems from the Bridgeman vs. Corel case.
To celebrate this news, Jenerator is pleased to present “Fishermen”, a charcoal drawing by Millet.

…and my goodness didn’t this train of enquiry turn up an interesting blog…