For (roughly) the first half of this volume, Rateliff carries his presentation of Tolkien's manuscripts/typescripts (and notes, and mini-essays commenting on aspects of the manuscripts, and notes on the mini-essays...) to the conclusion of the story, beginning where Part 1 ended (at Lake-town).
Both here and in volume 1, there are interesting differences between the 'script and what eventually was published as the First Edition of _The Hobbit_: to take a simple example, the dwarf we know as Thorin Oakenshield was for the longest name known as Gandalf, while the wizard manipulating events is named Bladorthin.
But more interesting than the variants that Tolkien wrote are those that, in the event, he did _not_ write. He intended, until quite late in the story, to have Bilbo slay Smaug; The Battle of the Five Armies was originally to be the Battle of the Anduin Vale, in which Dwarfs would play no part, and occur during Bilbo's return journey; when Tolkien finally gave the dragon-slaying to Bard (who was invented on the spot for just that purpose), he had Bard die in Smaug's fall, a decision he retconned pretty quickly.
"So what," you may ask, "takes up the rest of the volumes?" Well: there is the story of the Second Edition, and the Revised Edition of the '60s; but there are other revisions, too, which never saw print (until now). Tolkien became semi-obsessed (the way he did) with the phases of the Moon during the story, realizing that they simply didn't seem to work the way the actual Moon does unless some significant changes were made to the timeline. The dates of Bilbo's departure and return are more-or-less fixed in the published text, as is Bilbo's birthday at Lake-town and the Durin's Day discovery of the Back Door. Furthermore, the distances on the map only made things worse.
In the early '60s, he not only did a passel of calculations to figure out how to make it all work together, he began rewriting _The Hobbit_ from the beginning - not only making the necessary changes, but attempting to rewrite the story from page one to bring it more in line with the tone of _The Lord of the Rings_. Fortunately, he abandoned this shortly after beginning the revision of Chapter 3, but what he _did_ write makes fascinating reading. As someone he showed it to said, "It's very good, but it's not _The Hobbit_". As a result, very little of what he was thinking of eventually made it into the Revised Edition - mostly minor corrections. Rateliff observes that Tolkien realized that _The Hobbit_ was a very different type of book than _LR_, and stopped trying to force it into the wrong mold.
The notes are, as implied, extensive and exhaustive, ranging in length from single brief sentences to divigations lasting several pages. The mini-essays are uniformly interesting, on topics ranging from naming to Tolkien's sources to the implications of some of the choices Tolkien made (and didn't make) in writing his little masterpiece. Worth noting: Rateliff is very generous in his thanks to and acknowledgement of the work others have done in this field, especially Taum Santowski, who was to have been the author-editor of this book but died before he could do it. This has opened for me a few doors for future reading...
All in all, the _History of the Hobbit_ fascinated me from beginning to end (with the exception of a brief excursion into Tolkien's alphabets - I _do_ find those interesting, but they were out of tone for this book).