I used a UMAX Vista S6 scanner to scan the book in. Lineart scans with this scanner are very fast, which is very convenient when scanning in nearly 400 pages. I used OmniPage Pro to recognize the text in batches of a chapter or two at a time.
I saved the recognized text in a text file with one line per paragraph, and then moved it to my Linux machine, where I used the jed editor (with a set of macros which I wrote) to add HTML markup.
As a Linux zealot, I would rather do the whole job under Linux, because I dislike having Windows 95 crash on me approximately every 45 minutes, as it did while I scanned this book. To that end, I have written a Linux driver for the UMAX line of scanners. I also hope that eventually I'll be able to run OmniPage Pro under Linux as well.