June 7. Woke up at three-something. Lay awake and kept thinking on the problem so I could write up the analysis a little better. Ran four miles steady state shortly after sunrise. Then fell asleep again, started the workday extremely late.