On small differences in sensation