a better locking mechanism

Tim Robinson came by to see my prototype and had a number of excellent suggestions. One of the problem areas we discussed was my locking mechanism, which is based on a rotating wedge.

It doesn't work well in moving the digit wheel away from the wedge pivot point. We mused about whether a different wedge shape would help, but I thought about it later and couldn't find one that worked. Babbage sometimes got around this problem by using different and oddly-shaped locking teeth that weren't simultaneously functioning as gear teeth for meshing.

To fix the problem I redesigned the lock to use a wedge that drives along a line through the digit wheel axis, and I pointed the wedge. This works much better. The downside is three parts instead of one, and a second supporting rod. 

I may eventually investigate a bar lock moving in an oblique slot like in Difference Engine #2. That presents its own challenges, but it would use far fewer parts.