A crucial experiment
We investigate a possible event in ancient science.
We conclude that it can be useful to teach something that is actually wrong.
We notice an area of science where theories are so numerous a system has been imposed to keep track.
Science is generally self-correcting, but usually not in the same person.