The smallest negative finite value the float can contain.