Geometry — a paragon of mathematical deduction?