Although my digitization work for Project Gutenberg primarily focused on mathematics, I also contributed a smaller number of books more properly situated within physics and astronomy. Prior to 2018, the 95-year limit on copyright in the US limited digitization to works published in 1922 or before. This included books on relativity (including books by Einstein and Eddington), electromagnetism, thermodynamics, mechanics, and general science, but excluded works from the rapid development of quantum mechanics in the 1920s.
In 2016, I digitized Treatise on Thermodynamics by Max Planck, a 1903 translation by Alexander Ogg of the 1897 first edition. At Distributed Proofreaders (DP) in 2010, I had post-processed the 1922 seventh edition of the German original, Vorlesungen über Thermodynamik, which had been scanned and initiated by Laura Wisewell. By the time I arrived, another DP volunteer, happily (for me) a native speaker of German, was serving as project manager. I do not know for certain why this volunteer's name was not formally associated with the project and included in the book's credits, but it is a pleasure to acknowledge their friendly, helpful guidance and technical assistance.
Elementary Principles in Statistical Mechanics (1902) by Josiah Willard Gibbs is a treatise on statistical mechanics. Elsewhere I have written about digitization as a collaboration across space and time. There is an almost-eerie sense of temporal connection when I have occasion to visit the physics department at Yale, where a plaster model of the thermodynamic surface is displayed.
To reiterate, I am not a physicist. These books are possibly of greater historical interest than they are as scientific references. On the other hand, we are in an era where knowledge risks being lost. In this world, as during every such interval of history, redundancy improves the chances of information surviving.