"[N]o service can be rendered to the cause of truth by affecting to deny that, especially in the early history of the sciences, we find long periods of total stagnation, and the tide even ebbing, when by our calculations it should have overflowed."