Babbage's Analytical Engine
Babbage's autobiography, Passages from the Life of a Philosopher (1864) is a book that exposes both Babbage's eccentricity and his genius. He writes about everything from his involvement in politics to his explorations of the caldera of Vesuvius and his work on early British railways. Along the way, he exposes himself as having an overwhelming hatred of street music.
His description of the analytical engine is interesting. Through most of it you gather (with effort) that the machine could execute sequential programs. The challenge in this paper is to find where fleetingly mentions the possibility of what we now call branch or goto operations, and even more fleetingly mentions the possibility of a conditional branch. Once you recognize that the analytical engine had these, it is clear that it was going to be a real computer.
Babbage's son Henry built a working mill (CPU) for an analytical engine in 1910. You can see a brief writeup on the Scheutz difference engines at the Henry Prevost Babbage web page maintained by The Computer History Museum.