"But a science is exact to the extent that its method measures up to and is adequate to its object."