Also, one of his books includes an appendix discussing his part in the investigation of the Challenger shuttle disaster of 1986.
He describes sitting in a meeting, dropping an o-ring in his glass of ice water, pulling it out and breaking it on the table.
He also spends a chapter talking about experiments with ants, and leaves you hanging elsewhere with an experiment at college where a hose hooked to a common revolving sprinkler is submerged in a tank. After pointing out that water shot through the hose into the tank caused the arms to rotate in the expected direction, he challenges you to say which way it would rotate if you instead did something else, it was either sucked water out of the hose, or pressurized the water in the tank to push the water out of the hose.
I'll have to dig up the book again now.
Or, as correctly described in a post above, he flattened it and showed it didn't bounce back, NOT that it broke on the table.
A mind is a terrible thing to lose.
I would imagine you could set it up with a two arm sprinkler and a fitting sealed to and through the bottom of the tank that could be connected and disconnected to a garden hose.
Connect the hose, turn on the water and fill the tank.
Then disconnect the hose and let air pressure force the water out of the tank through the openings in the bent arms, which direction will the sprinkler turn?
Will it turn at all?
What if you seal one end of an arm?